issue_owner_repo listlengths 2 2 | issue_body stringlengths 0 261k ⌀ | issue_title stringlengths 1 925 | issue_comments_url stringlengths 56 81 | issue_comments_count int64 0 2.5k | issue_created_at stringlengths 20 20 | issue_updated_at stringlengths 20 20 | issue_html_url stringlengths 37 62 | issue_github_id int64 387k 2.46B | issue_number int64 1 127k |
|---|---|---|---|---|---|---|---|---|---|
[
"MonetDB",
"MonetDB"
] | Date: 2017-10-02 14:57:08 +0200
From: Manuel <<manuel>>
To: SQL devs <<bugs-sql>>
Version: 11.25.23 (Dec2016-SP5)
CC: @njnes
Last updated: 2017-11-01 16:41:06 +0100
## Comment 25677
Date: 2017-10-02 14:57:08 +0200
From: Manuel <<manuel>>
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/61.0.3163.100 Safari/537.36
Build Identifier:
If I call sys.epoch function on date beyond 19-01-2038 03:14:00, what I get back is a negative number. I understand that this is an overflow issue, but currently there is no work-around for this problem in monetdb.
Reproducible: Always
### Steps to Reproduce:
select sys.epoch(str_to_timestamp('19-01-2038 03:15', '%d-%m-%Y %H:%M')) AS "timestamp";
### Actual Results:
-2147483596
### Expected Results:
2147483700
I am actually using sys.epoch to evaluate the mean (avg) over a date column: Is there an alternative workaround for that ?.
## Comment 25720
Date: 2017-10-15 11:32:17 +0200
From: @njnes
given the current data type (int) used an error should be given by the epoch function. Somehow we need to switch to BIGINT.
## Comment 25834
Date: 2017-11-01 16:41:06 +0100
From: Manuel <<manuel>>
(In reply to Niels Nes from comment 1)
> given the current data type (int) used an error should be given by the epoch
> function. Somehow we need to switch to BIGINT.
Would it be possible to introduce a new function returning a bigint ?
| Overflow in sys.epoch | https://api.github.com/repos/MonetDB/MonetDB/issues/6412/comments | 0 | 2020-11-30T15:53:08Z | 2024-06-27T13:01:00Z | https://github.com/MonetDB/MonetDB/issues/6412 | 753,593,718 | 6,412 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-28 11:08:37 +0200
From: @PedroTadim
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-10-26 14:01:26 +0200
## Comment 25671
Date: 2017-09-28 11:08:37 +0200
From: @PedroTadim
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/61.0.3163.100 Safari/537.36
Build Identifier:
Mserver5 receives a SIGSEGV while compiling a sqlitelogictest aggregation query.
Reproducible: Always
### Steps to Reproduce:
1. CREATE TABLE tab0 (col0 INTEGER, col1 INTEGER, col2 INTEGER);
2. SELECT DISTINCT + 38 AS col2 FROM tab0 AS cor0 WHERE NOT NULL IS NOT NULL OR NULL IN ( + col1, col0, - 97, + col2 ) GROUP BY col1;
### Actual Results:
Mserver5 receives a SIGSEGV while compiling the query in a exp_name function call.
### Expected Results:
An empty result set.
GDB information:
Thread 7 "mserver5" received signal SIGSEGV, Segmentation fault.
0x00007fffe8b0992c in exp_name (e=0x0) at rel_exp.c:696
696 if (e->name)
Backtrace:
0 0x00007fffe8b0992c in exp_name (e=0x0) at rel_exp.c:696
1 0x00007fffe8ae5d5e in rel_logical_exp (sql=0x7fffb800ab40, rel=0x7fffb80726a0, sc=0x7fffb8070f50, f=1) at rel_select.c:2633
2 0x00007fffe8ae4639 in rel_logical_exp (sql=0x7fffb800ab40, rel=0x7fffb80711e0, sc=0x7fffb8070fd0, f=1) at rel_select.c:2370
3 0x00007fffe8af09de in rel_select_exp (sql=0x7fffb800ab40, rel=0x7fffb80711e0, sn=0x7fffb8071160, ek=...) at rel_select.c:4792
4 0x00007fffe8af1cc6 in rel_query (sql=0x7fffb800ab40, rel=0x0, sq=0x7fffb8071160, toplevel=1, ek=..., apply=8) at rel_select.c:5088
5 0x00007fffe8af35a6 in rel_subquery (sql=0x7fffb800ab40, rel=0x0, sq=0x7fffb8071160, ek=..., apply=8) at rel_select.c:5453
6 0x00007fffe8af36cd in rel_selects (sql=0x7fffb800ab40, s=0x7fffb8071160) at rel_select.c:5474
7 0x00007fffe8ad950f in rel_semantic (sql=0x7fffb800ab40, s=0x7fffb8071160) at rel_semantic.c:225
8 0x00007fffe8a1ab09 in sql_symbol2relation (c=0x7fffb800ab40, sym=0x7fffb8071160) at sql.c:117
9 0x00007fffe8a39940 in SQLparser (c=0x7fffea11f338) at sql_scenario.c:1129
10 0x00007ffff7a15768 in runPhase (c=0x7fffea11f338, phase=1) at mal_scenario.c:506
11 0x00007ffff7a158ac in runScenarioBody (c=0x7fffea11f338) at mal_scenario.c:539
12 0x00007ffff7a15a4f in runScenario (c=0x7fffea11f338) at mal_scenario.c:568
13 0x00007ffff7a17541 in MSserveClient (dummy=0x7fffea11f338) at mal_session.c:467
14 0x00007ffff7a17031 in MSscheduleClient (command=0x7fffb80008d0 "0", challenge=0x7fffcd398deb "STelO9tk0WD", fin=0x7fffb80028f0, fout=0x7fffc4002b70, protocol=PROTOCOL_9, blocksize=8190,
compute_column_widths=0) at mal_session.c:358
15 0x00007ffff7aba460 in doChallenge (data=0x7fffc40008d0) at mal_mapi.c:273
16 0x00007ffff769180c in thread_starter (arg=0x7fffc4004c60) at gdk_system.c:485
17 0x00007ffff4ca536d in start_thread () from /lib64/libpthread.so.0
18 0x00007ffff49ddbbf in clone () from /lib64/libc.so.6
## Comment 25672
Date: 2017-09-28 11:17:22 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [3a21d4ec5b84](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=3a21d4ec5b84) made by Pedro Ferreira <pedro.ferreira@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=3a21d4ec5b84](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=3a21d4ec5b84)
Changeset description:
Added test for bug #6411
## Comment 25682
Date: 2017-10-07 14:08:31 +0200
From: @njnes
fixed crash, reimplemented large part of the in handling.
## Comment 25689
Date: 2017-10-11 16:35:55 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [b7681c154e83](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=b7681c154e83) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=b7681c154e83](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=b7681c154e83)
Changeset description:
reimplemented IN/NOT IN. Fixed bug #6411
| Sqlitelogictest crash in aggregation query | https://api.github.com/repos/MonetDB/MonetDB/issues/6411/comments | 0 | 2020-11-30T15:53:06Z | 2024-06-27T13:00:59Z | https://github.com/MonetDB/MonetDB/issues/6411 | 753,593,669 | 6,411 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-21 10:21:06 +0200
From: @PedroTadim
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-10-26 14:01:27 +0200
## Comment 25665
Date: 2017-09-21 10:21:06 +0200
From: @PedroTadim
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/61.0.3163.91 Safari/537.36
Build Identifier:
Mserver receives a SIGSEV in an aggregation query with IN clause. Tested in the latest development version of the Jul2017 branch.
Reproducible: Always
### Steps to Reproduce:
1. CREATE TABLE tab2(col0 INTEGER, col1 INTEGER, col2 INTEGER);
2. SELECT 9 - - col1 FROM tab2 GROUP BY col1 HAVING - CAST ( NULL AS INTEGER ) IN ( + - SUM ( - + 47 ), + MIN ( col2 ) );
### Actual Results:
Mserver receives a SIGSEGV in a rel_logical_exp call.
### Expected Results:
An empty result set.
GDB output:
Thread 6 "mserver5" received signal SIGSEGV, Segmentation fault.
0x00007fffe8ae5d36 in rel_logical_exp (sql=0x7fffb800ab40, rel=0x7fffb80e98b0, sc=0x7fffb80e7e80, f=3) at rel_select.c:2628
2628 le = exp_column(sql->sa, exp_relname(le), exp_name(le), exp_subtype(le), le->card, has_nil(le), is_intern(le));
Backtrace:
0 0x00007fffe8ae5d36 in rel_logical_exp (sql=0x7fffb800ab40, rel=0x7fffb80e98b0, sc=0x7fffb80e7e80, f=3) at rel_select.c:2628
1 0x00007fffe8af0f39 in rel_select_exp (sql=0x7fffb800ab40, rel=0x7fffb80e8540, sn=0x7fffb80e7f40, ek=...) at rel_select.c:4884
2 0x00007fffe8af1c67 in rel_query (sql=0x7fffb800ab40, rel=0x0, sq=0x7fffb80e7f40, toplevel=1, ek=..., apply=8) at rel_select.c:5082
3 0x00007fffe8af3547 in rel_subquery (sql=0x7fffb800ab40, rel=0x0, sq=0x7fffb80e7f40, ek=..., apply=8) at rel_select.c:5447
4 0x00007fffe8af366e in rel_selects (sql=0x7fffb800ab40, s=0x7fffb80e7f40) at rel_select.c:5468
5 0x00007fffe8ad94e1 in rel_semantic (sql=0x7fffb800ab40, s=0x7fffb80e7f40) at rel_semantic.c:225
6 0x00007fffe8a1ab09 in sql_symbol2relation (c=0x7fffb800ab40, sym=0x7fffb80e7f40) at sql.c:117
7 0x00007fffe8a39940 in SQLparser (c=0x7fffea11f338) at sql_scenario.c:1129
8 0x00007ffff7a15768 in runPhase (c=0x7fffea11f338, phase=1) at mal_scenario.c:506
9 0x00007ffff7a158ac in runScenarioBody (c=0x7fffea11f338) at mal_scenario.c:539
10 0x00007ffff7a15a4f in runScenario (c=0x7fffea11f338) at mal_scenario.c:568
11 0x00007ffff7a17541 in MSserveClient (dummy=0x7fffea11f338) at mal_session.c:467
12 0x00007ffff7a17031 in MSscheduleClient (command=0x7fffb80008d0 "0", challenge=0x7fffcdc79deb "bwX3mxJHt8d", fin=0x7fffb80028f0, fout=0x7fffc4002b70, protocol=PROTOCOL_9, blocksize=8190,
compute_column_widths=0) at mal_session.c:358
13 0x00007ffff7aba460 in doChallenge (data=0x7fffc40008d0) at mal_mapi.c:273
14 0x00007ffff769180c in thread_starter (arg=0x7fffc4004c60) at gdk_system.c:485
15 0x00007ffff4ca536d in start_thread () from /lib64/libpthread.so.0
16 0x00007ffff49ddbbf in clone () from /lib64/libc.so.6
## Comment 25666
Date: 2017-09-21 10:38:21 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [d2a9f17a79ea](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=d2a9f17a79ea) made by Pedro Ferreira <pedro.ferreira@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=d2a9f17a79ea](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=d2a9f17a79ea)
Changeset description:
Added test for bug #6410
## Comment 25669
Date: 2017-09-27 16:36:00 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [7a69de77b1b3](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=7a69de77b1b3) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7a69de77b1b3](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7a69de77b1b3)
Changeset description:
fixes for bugs 6409 and 6410. Only set subquery flags when realy needed.
## Comment 25670
Date: 2017-09-27 16:54:25 +0200
From: @njnes
fixed by improved subquery handing
| Sqlitelogictest crash on aggregation query with IN clause | https://api.github.com/repos/MonetDB/MonetDB/issues/6410/comments | 0 | 2020-11-30T15:53:02Z | 2024-06-27T13:00:58Z | https://github.com/MonetDB/MonetDB/issues/6410 | 753,593,622 | 6,410 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-20 11:00:13 +0200
From: @PedroTadim
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-10-26 14:01:26 +0200
## Comment 25656
Date: 2017-09-20 11:00:13 +0200
From: @PedroTadim
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/61.0.3163.91 Safari/537.36
Build Identifier:
The server receives a SIGABRT signal on an aggregation query with a NOT IN condition in the HAVING clause. Tested on the latest development version of the Jul2017 branch.
Reproducible: Always
### Steps to Reproduce:
1. CREATE TABLE tab1(col0 INTEGER, col1 INTEGER, col2 INTEGER);
2. SELECT cor0.col1 FROM tab1 AS cor0 GROUP BY cor0.col1 HAVING NULL NOT IN ( - cor0.col1 );
### Actual Results:
The server receives a SIGABRT signal from an assert(0) call in exp_bin.
### Expected Results:
An empty result set.
Backtrace of the assertion error:
0 0x00007ffff490369b in raise () from /lib64/libc.so.6
1 0x00007ffff49054a0 in abort () from /lib64/libc.so.6
2 0x00007ffff48fbd5a in __assert_fail_base () from /lib64/libc.so.6
3 0x00007ffff48fbdd2 in __assert_fail () from /lib64/libc.so.6
4 0x00007fffe8a46f1c in exp_bin (be=0x7fffb80cb0c0, e=0x7fffb80e27f0, left=0x7fffb80e4f90, right=0x7fffb80e5660, grp=0x0, ext=0x0, cnt=0x0, sel=0x0) at rel_bin.c:721
5 0x00007fffe8a4c918 in rel2bin_semijoin (be=0x7fffb80cb0c0, rel=0x7fffb80e2750, refs=0x7fffb80e4440) at rel_bin.c:1943
6 0x00007fffe8a59c71 in subrel_bin (be=0x7fffb80cb0c0, rel=0x7fffb80e2750, refs=0x7fffb80e4440) at rel_bin.c:4802
7 0x00007fffe8a4ee9e in rel2bin_project (be=0x7fffb80cb0c0, rel=0x7fffb80e16c0, refs=0x7fffb80e4440, topn=0x0) at rel_bin.c:2437
8 0x00007fffe8a59d2b in subrel_bin (be=0x7fffb80cb0c0, rel=0x7fffb80e16c0, refs=0x7fffb80e4440) at rel_bin.c:4818
9 0x00007fffe8a5a022 in output_rel_bin (be=0x7fffb80cb0c0, rel=0x7fffb80e16c0) at rel_bin.c:4883
10 0x00007fffe8a751fd in sql_relation2stmt (be=0x7fffb80cb0c0, r=0x7fffb80e16c0) at sql_gencode.c:529
11 0x00007fffe8a7557c in backend_dumpstmt (be=0x7fffb80cb0c0, mb=0x7fffb8112c90, r=0x7fffb80e16c0, top=1, add_end=1,
query=0x7fffb8109c70 "select cor0.col1 from tab1 as cor0 group by cor0.col1 having null not in ( - cor0.col1 );") at sql_gencode.c:576
12 0x00007fffe8a75de1 in backend_dumpproc (be=0x7fffb80cb0c0, c=0x7fffea11f338, cq=0x7fffb810fb30, r=0x7fffb80e16c0) at sql_gencode.c:703
13 0x00007fffe8a39e8a in SQLparser (c=0x7fffea11f338) at sql_scenario.c:1182
14 0x00007ffff7a15768 in runPhase (c=0x7fffea11f338, phase=1) at mal_scenario.c:506
15 0x00007ffff7a158ac in runScenarioBody (c=0x7fffea11f338) at mal_scenario.c:539
16 0x00007ffff7a15a4f in runScenario (c=0x7fffea11f338) at mal_scenario.c:568
17 0x00007ffff7a17541 in MSserveClient (dummy=0x7fffea11f338) at mal_session.c:467
18 0x00007ffff7a17031 in MSscheduleClient (command=0x7fffb80008d0 "0", challenge=0x7fffcdc79deb "w2acK2pZ", fin=0x7fffb80028f0, fout=0x7fffc4002b70, protocol=PROTOCOL_9, blocksize=8190,
compute_column_widths=0) at mal_session.c:358
19 0x00007ffff7aba3c1 in doChallenge (data=0x7fffc40008d0) at mal_mapi.c:273
20 0x00007ffff769180c in thread_starter (arg=0x7fffc4004c60) at gdk_system.c:485
21 0x00007ffff4ca536d in start_thread () from /lib64/libpthread.so.0
22 0x00007ffff49ddbbf in clone () from /lib64/libc.so.6
## Comment 25661
Date: 2017-09-20 12:11:52 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [ca1758c8b9b8](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=ca1758c8b9b8) made by Pedro Ferreira <pedro.ferreira@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=ca1758c8b9b8](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=ca1758c8b9b8)
Changeset description:
Added test for bug #6409
## Comment 25663
Date: 2017-09-21 09:39:57 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [86be4d50523b](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=86be4d50523b) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=86be4d50523b](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=86be4d50523b)
Changeset description:
fixes for bug #6409
## Comment 25664
Date: 2017-09-21 09:41:36 +0200
From: @njnes
fixed. Make sure c-expressions structures are unique, ie create aliases where needed
## Comment 25668
Date: 2017-09-27 16:35:24 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [7a69de77b1b3](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=7a69de77b1b3) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7a69de77b1b3](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7a69de77b1b3)
Changeset description:
fixes for bugs 6409 and 6410. Only set subquery flags when realy needed.
| sqllogictest crash on aggregation query with NOT IN clause in HAVING clause | https://api.github.com/repos/MonetDB/MonetDB/issues/6409/comments | 0 | 2020-11-30T15:52:58Z | 2024-06-27T13:00:57Z | https://github.com/MonetDB/MonetDB/issues/6409 | 753,593,573 | 6,409 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-19 19:02:48 +0200
From: Manuel <<manuel>>
To: SQL devs <<bugs-sql>>
Version: 11.25.23 (Dec2016-SP5)
CC: @njnes
Last updated: 2017-09-20 11:31:25 +0200
## Comment 25648
Date: 2017-09-19 19:02:48 +0200
From: Manuel <<manuel>>
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.113 Safari/537.36
Build Identifier:
I get insistent behaviour using the Floor function:
select Floor( -1 / 1001.0 ); /* 0 */
select Floor( -1 / 1000.0 ); /* -1 */
I would expect both of them to return -1
Reproducible: Always
### Steps to Reproduce:
Just query
select Floor(( -1 / 1001.0 ));
### Expected Results:
-1
## Comment 25649
Date: 2017-09-19 22:09:24 +0200
From: @sjoerdmullender
-1 / 1000.0 = -0.001
-1 / 1001.0 = -0.000999000999000999
If you truncate those two numbers to 3 decimals, you get -0.001 and 0.000. If you then apply the floor function to those, you get -1 and 0.
Your problem, I think, is that you expect the number in the query to behave like floating point numbers. They are not floating point numbers, they are decimals.
The query behaves the way you expect if you force the numbers to be floating point:
select Floor( cast(-1 as real) / 1000.0 ); -- -1
select Floor( cast(-1 as real) / 1001.0 ); -- -1
Try all combinations without calling floor() and see the results. E.g.:
select -1 / 1001.0; -- 0.000
select -1 / 1000.0; -- -0.001
I'd say, this is not a bug.
## Comment 25659
Date: 2017-09-20 11:31:25 +0200
From: Manuel <<manuel>>
(In reply to Sjoerd Mullender from comment 1)
> -1 / 1000.0 = -0.001
> -1 / 1001.0 = -0.000999000999000999
>
> If you truncate those two numbers to 3 decimals, you get -0.001 and 0.000.
> If you then apply the floor function to those, you get -1 and 0.
>
> Your problem, I think, is that you expect the number in the query to behave
> like floating point numbers. They are not floating point numbers, they are
> decimals.
>
> The query behaves the way you expect if you force the numbers to be floating
> point:
>
> select Floor( cast(-1 as real) / 1000.0 ); -- -1
> select Floor( cast(-1 as real) / 1001.0 ); -- -1
>
> Try all combinations without calling floor() and see the results. E.g.:
> select -1 / 1001.0; -- 0.000
> select -1 / 1000.0; -- -0.001
>
> I'd say, this is not a bug.
Thank you Sjoerd, I now understand the issue. Am I correct in saying that every non integer literal number is then treated as a decimal with 3 decimal places?
| Rounding error on Floor function | https://api.github.com/repos/MonetDB/MonetDB/issues/6408/comments | 0 | 2020-11-30T15:52:56Z | 2024-06-28T13:10:55Z | https://github.com/MonetDB/MonetDB/issues/6408 | 753,593,541 | 6,408 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-12 14:57:37 +0200
From: @gijzelaerr
To: MonetDB5 devs <<bugs-monetdb5>>
Version: 11.27.5 (Jul2017-SP1)
CC: @mlkersten
Last updated: 2018-08-28 09:48:33 +0200
## Comment 25628
Date: 2017-09-12 14:57:37 +0200
From: @gijzelaerr
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.113 Safari/537.36
Build Identifier:
If i do a consecutive create and destroy of a database the client (or server) gets stuck.
for example:
DATABASE="test"
while true; do
monetdb stop $DATABASE || true
monetdb destroy -f $DATABASE || true
monetdb create $DATABASE
monetdb release $DATABASE
monetdb start $DATABASE
done
output:
stop: no such database: test
destroy: no such database: test
created database in maintenance mode: test
taken database out of maintenance mode: test
starting database 'test'... done
stopping database 'test'... done
destroyed database: test
created database in maintenance mode: test
taken database out of maintenance mode: test
starting database 'test'...
monetdb hangs after the second stop or start. I can't start, stop or connect anymore after hits. I have to kill -9 monetdbd to get control back.
OS: OSX 10.12.6
Reproducible: Always
## Comment 25629
Date: 2017-09-12 15:13:21 +0200
From: @gijzelaerr
also on Debian.
## Comment 25630
Date: 2017-09-12 16:33:13 +0200
From: @sjoerdmullender
Is there anything in the merovingian log file?
## Comment 25631
Date: 2017-09-12 16:56:52 +0200
From: @gijzelaerr
Created attachment 568
merovingian log file
> Attached file: [merovingian.log](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6407_merovingian.log_568) (text/x-log, 431469 bytes)
> Description: merovingian log file
## Comment 25632
Date: 2017-09-12 16:58:05 +0200
From: @gijzelaerr
looks like the stop times out and merovingian kills it after a minute
ERR merovingian[1207]: timeout of 60 seconds expired, sending process 11488 (database 'test') the KILL signal
## Comment 25633
Date: 2017-09-12 17:05:23 +0200
From: @sjoerdmullender
I see something interesting in the log: The very last line is
2017-09-12 16:55:28 MSG merovingian[1207]: sending process 11607 (database 'test') the TERM signal
The thing is, the latest installment of the database server has PID 11633. PID 11607 was the PID of the previous installment of the database server, which was stopper (without the timeout you noticed) only moments before.
It looks like the internal data structures of monetdbd are messed up.
## Comment 25667
Date: 2017-09-23 21:45:32 +0200
From: @mlkersten
On my Mac I don't see any problems. On my Linux box I have to wait for time outs, but it continuous.
## Comment 25905
Date: 2017-11-23 16:40:42 +0100
From: @sjoerdmullender
I fixed a race condition in monetdbd that can happen when a mserver is forked off. The race is basically, monetdbd forks, the child exits quickly (or the parent thread just takes its time), and in the mean time, another monetdbd thread notices that the child died. This other thread tries to remove this child from the administration (and fails since it isn't there yet). Then the forking thread wakes up and adds the child to the administration.
This could perhaps have something to do with this bug.
The fix is in changeset [805d8e0173c8](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=805d8e0173c8)
Can you try, since we don't seem to be able to reproduce.
## Comment 26607
Date: 2018-08-28 09:48:33 +0200
From: @sjoerdmullender
Assuming fixed. If not, please reopen.
| looping monetdb create release start stop destroy a couple of times hangs on stop or start | https://api.github.com/repos/MonetDB/MonetDB/issues/6407/comments | 0 | 2020-11-30T15:52:52Z | 2024-06-27T13:00:55Z | https://github.com/MonetDB/MonetDB/issues/6407 | 753,593,496 | 6,407 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-11 22:47:01 +0200
From: Moba <<mobashhir1>>
To: Embedded devs <<bugs-embedded>>
Version: unspecified
CC: mobashhir1, @njnes, @PedroTadim
Last updated: 2020-07-10 21:36:22 +0200
## Comment 25627
Date: 2017-09-11 22:47:01 +0200
From: Moba <<mobashhir1>>
User-Agent: Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/61.0.3163.79 Safari/537.36
Build Identifier:
Copy (table) into (csv) works fine in mclient as a standalone query.
When using embedded python, the exact same query when executed in the _conn.execute() command creates only an empty csv without any data.
Using raw python and creating an engine and a connection to execute the query will produce the csv WITH the data.
Only _conn.execute produces the csv WITHOUT the data.
Reproducible: Always
### Steps to Reproduce:
1.CREATE TABLE TEST_T1(TESTING_COL INT);
2.INSERT INTO TEST_T1 VALUES(100);
3.INSERT INTO TEST_T1 VALUES(200);
4.INSERT INTO TEST_T1 VALUES(300);
5.CREATE FUNCTION COPY_RESULTS()
RETURNS VARCHAR(32)
LANGUAGE PYTHON {
try:
copy_query = 'ok'
_conn.execute("COPY SELECT * FROM test_t1 INTO \'C:\\\\Users\\\\{username}\\\\test_t1.csv\' DELIMITERS \',\' ; ")
except Exception as ex:
return str(ex)
return copy_query
};
7. Please put another location or your username in the above path.
6.select COPY_RESULTS();
### Actual Results:
Run the above steps.
Sometimes there are issues with the python indentation but that is easy to resolve.
Once Step6. is successfully run, you'll be able to see the csv file WITHOUT the data in the mentioned location but it will be empty.
Delete that file and run this simple query
COPY SELECT * FROM test_t1 INTO 'C:\\\\Users\\\\{username}\\\\test_t1.csv\'
this will produce the csv WITH the data.
### Expected Results:
_conn.execute('query') should produce the csv with the results, but it is empty.
## Comment 25653
Date: 2017-09-20 10:21:13 +0200
From: @njnes
seems specific to the embedded part of monetdb
## Comment 25654
Date: 2017-09-20 10:22:56 +0200
From: @njnes
seems specific to the embedded part of monetdb
## Comment 25655
Date: 2017-09-20 10:26:10 +0200
From: @njnes
seems specific to the embedded part of monetdb
## Comment 25660
Date: 2017-09-20 11:54:33 +0200
From: @PedroTadim
I made a fix for this issue in the Java version of MonetDBLite. We are discussion now some compilation issues, before merging the fixe.
## Comment 27902
Date: 2020-07-10 21:36:22 +0200
From: @sjoerdmullender
The old MonetDB-Light project has been abandoned because it was unmaintainable.
There is now a new monetdbe (MonetDB-Embedded) in the works.
| COPY TABLE INTO csv creates an empty csv if executed using the _conn.execute() command. | https://api.github.com/repos/MonetDB/MonetDB/issues/6406/comments | 0 | 2020-11-30T15:52:49Z | 2024-06-28T13:10:55Z | https://github.com/MonetDB/MonetDB/issues/6406 | 753,593,452 | 6,406 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-11 11:15:12 +0200
From: @swingbit
To: GDK devs <<bugs-common>>
Version: 11.27.5 (Jul2017-SP1)
Last updated: 2017-10-26 14:01:34 +0200
## Comment 25625
Date: 2017-09-11 11:15:12 +0200
From: @swingbit
User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.101 Safari/537.36
Build Identifier:
In gdk_batop.c:BATkeyed()
ifndef DISABLE_PARENT_HASH
|| ((parent = VIEWtparent(b)) != 0 &&
BATcheckhash(BBPdescriptor(parent)))
endif
) {
Variable parent is not declared when DISABLE_PARENT_HASH is not defined:
src/gdk/gdk_batop.c: In function ‘BATkeyed’:
src/gdk/gdk_batop.c:960:12: error: ‘parent’ undeclared (first use in this function)
|| ((parent = VIEWtparent(b)) != 0 &&
^~~~~~
Reproducible: Always
## Comment 25626
Date: 2017-09-11 11:24:29 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [02cbcb898913](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=02cbcb898913) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=02cbcb898913](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=02cbcb898913)
Changeset description:
'parent' undeclared.
Fixes bug #6405.
| Compilation error if DISABLE_PARENT_HASH not defined | https://api.github.com/repos/MonetDB/MonetDB/issues/6405/comments | 0 | 2020-11-30T15:52:45Z | 2024-06-27T13:00:53Z | https://github.com/MonetDB/MonetDB/issues/6405 | 753,593,400 | 6,405 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-10 12:30:57 +0200
From: Trung Le <<tle211212>>
To: SQL devs <<bugs-sql>>
Version: 11.27.5 (Jul2017-SP1)
CC: @njnes
Last updated: 2017-10-26 14:01:39 +0200
## Comment 25621
Date: 2017-09-10 12:30:57 +0200
From: Trung Le <<tle211212>>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Firefox/52.0
Build Identifier:
When I use COPY INTO feature to import data for a table that has primary key or foreign key constraint, the monetdb db crashes. I expect the case should work or return an not-supported message.
Reproducible: Always
If it is not supported, monetdb should response such information instead of crash.
## Comment 25623
Date: 2017-09-11 10:00:45 +0200
From: @sjoerdmullender
Can you give an example for this problem, please?
## Comment 25624
Date: 2017-09-11 10:17:25 +0200
From: @kutsurak
Created attachment 567
COPY INTO crash reproduction
I am attaching a minimal example that reproduces the problem for me.
As far as I can tell it crashes if the table has a primary key constraint AND we are not loading all the columns.
Panos.
> Attached file: [copy_into_crash.sql](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6404_copy_into_crash.sql_567) (application/sql, 112 bytes)
> Description: COPY INTO crash reproduction
## Comment 25639
Date: 2017-09-17 10:16:02 +0200
From: @njnes
non of the input columns were used and the table code requires atleast one
valid output bat. We now give an error in this case.
## Comment 25643
Date: 2017-09-17 18:02:24 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [261a59262ad6](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=261a59262ad6) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=261a59262ad6](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=261a59262ad6)
Changeset description:
add test and output for bug #6404
## Comment 25644
Date: 2017-09-17 18:03:59 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [d1128cc44188](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=d1128cc44188) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=d1128cc44188](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=d1128cc44188)
Changeset description:
make sure we reference expressions in a copy into statement (with column list).
This solves bug #6404.
| COPY INTO crashes if table has primary key or foreign key constraint | https://api.github.com/repos/MonetDB/MonetDB/issues/6404/comments | 0 | 2020-11-30T15:52:42Z | 2024-06-27T13:00:52Z | https://github.com/MonetDB/MonetDB/issues/6404 | 753,593,350 | 6,404 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-10 12:27:31 +0200
From: Trung Le <<tle211212>>
To: SQL devs <<bugs-sql>>
Version: 11.27.5 (Jul2017-SP1)
CC: @kutsurak, @njnes
Last updated: 2017-09-18 03:06:51 +0200
## Comment 25620
Date: 2017-09-10 12:27:31 +0200
From: Trung Le <<tle211212>>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Firefox/52.0
Build Identifier:
When I use COPY INTO, although I specify the list of columns matching with the CSV file, MonetDB wants the columns to be in the same order as in the CREATE TABLE statement. This might be a regression.
Reproducible: Always
### Steps to Reproduce:
1.CREATE TABLE t (i1 int, i2 int, s2 string)
2.CSV content: 1,"abc" (1 line is enough)
3.COPY INTO t (i1, s2) FROM a.csv USING delimiters ',','\n','"'
### Actual Results:
MonetDB returns:
Failed to import table 't', Column value 2 missing
### Expected Results:
MonetDB should say 1 record imported.
## Comment 25622
Date: 2017-09-11 09:48:50 +0200
From: @kutsurak
Hi,
In this case you need to specify, the columns and their order after the filename as well:
COPY INTO t(i1, s2) FROM '/tmp/lala.csv'(i1, s2) using USING DELIMITERS ',','\n','"';
You may also omit the columns after the table name:
COPY INTO t FROM '/tmp/lala.csv' (i1, s2) using USING DELIMITERS ',','\n','"';
Some more documentation about COPY INTO can be found here: https://www.monetdb.org/Documentation/Cookbooks/SQLrecipes/CSV_bulk_loads
Please let us know if this works for you.
Best regards,
Panos.
## Comment 25647
Date: 2017-09-18 03:06:51 +0200
From: Trung Le <<tle211212>>
Hi,
I am sorry that I am late to reply. Yes, your suggestion works. Thanks alot.
| COPY INTO seems to require table column order stated in CREATE TABLE statement. | https://api.github.com/repos/MonetDB/MonetDB/issues/6403/comments | 0 | 2020-11-30T15:52:39Z | 2024-06-28T13:10:54Z | https://github.com/MonetDB/MonetDB/issues/6403 | 753,593,317 | 6,403 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-07 20:34:59 +0200
From: Rodrigo Schmidt Nurmberg <<rsn>>
To: clients devs <<bugs-clients>>
Version: 11.27.5 (Jul2017-SP1)
CC: martin.van.dinther
Last updated: 2019-09-26 17:16:18 +0200
## Comment 25619
Date: 2017-09-07 20:34:59 +0200
From: Rodrigo Schmidt Nurmberg <<rsn>>
User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/61.0.3163.79 Safari/537.36
Build Identifier:
I'm using prepareCall to call UDF across a range of databases, would be nice if MongoDB support this as well.
Reproducible: Always
### Steps to Reproduce:
Connection.prepareCall("{ call SomeFunction(?) }");
Connection.prepareCall returns a CallableStatement that isn't implemented yet, thus [MonetConnection.prepareCall](https://github.com/MonetDB/monetdb-java/blob/master/src/main/java/nl/cwi/monetdb/jdbc/MonetConnection.javaL629) returns null.
## Comment 25634
Date: 2017-09-14 17:42:35 +0200
From: Martin van Dinther <<martin.van.dinther>>
FYI: this request is on our ToDo-list.
As a workaround you can use:
Statement.execute("call SomeProcedure( args_as_literals )")
or
PreparedStatement.executeQuery("select SomeFunction(?)")
Note: this is a MonetDB bugzilla, not MongoDB! I assume it is a typo.
## Comment 25635
Date: 2017-09-14 17:46:34 +0200
From: Martin van Dinther <<martin.van.dinther>>
Correction, instead of
PreparedStatement.executeQuery("select SomeFunction(?)")
use:
pstmt = con.prepareStatement("select SomeFunction(?)");
pstmt.executeQuery();
## Comment 26942
Date: 2019-03-21 19:32:48 +0100
From: Martin van Dinther <<martin.van.dinther>>
Added implementation of java.sql.CallableStatement interface such that Connection.prepareCall("{ call SomeProcedure(?) }");
is now supported.
See also: https://dev.monetdb.org/hg/monetdb-java?cmd=changeset;node=4880267d0fe1
## Comment 27325
Date: 2019-09-26 17:16:18 +0200
From: Martin van Dinther <<martin.van.dinther>>
This is now implemented in the new: monetdb-jdbc-2.29.jre7.jar release.
It is downloadable from https://www.monetdb.org/downloads/Java/
| JDBC: Support for Connection.prepareCall | https://api.github.com/repos/MonetDB/MonetDB/issues/6402/comments | 0 | 2020-11-30T15:52:37Z | 2024-06-27T13:00:50Z | https://github.com/MonetDB/MonetDB/issues/6402 | 753,593,274 | 6,402 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-06 08:08:59 +0200
From: Edvinas <<edvinas.rasys>>
To: MonetDB5 devs <<bugs-monetdb5>>
Version: 11.33.3 (Apr2019)
CC: arkr17997, edvinas.rasys, kravchenko.anton86, @mlkersten, @njnes
Last updated: 2020-06-03 16:58:53 +0200
## Comment 25614
Date: 2017-09-06 08:08:59 +0200
From: Edvinas <<edvinas.rasys>>
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.113 Safari/537.36
Build Identifier:
Our use of monetdb involves creating and dropping a lot of tables in a quick sequence. We've observed that our servers continuously run out of memory and hang, most often during such drop/create/import sequence.
We managed to narrow down the scenario to the create/drop table sequence. This has been tried with the monetdb nodejs driver and mclient, both produce the same results.
Reproducible: Always
### Steps to Reproduce:
1.monetdb create test
2.monetdb release test
3. run the following sh script
!/bin/sh
DB="test"
for i in `seq 1 100`;
do
mclient -d $DB <<EOF
create table test_$i ( PersonID int,
LastName varchar(255),
FirstName varchar(255),
Address varchar(255),
City varchar(255)
);
drop table test_$i;
EOF
echo table_$i processed
done
### Actual Results:
When observed with htop, the RES value of mserver5 keeps increasing when step 3 is repeated several times
### Expected Results:
Since there is no data in the tables, and the tables are dropped after creating, the expected result is that the RES value should be the same after the script as it was before the script.
This issue has been observed on monetdb versions 11.23.13, 11.25.23 and 11.27.5
Run on Ubuntu Linux 14.04
Currently tested on:
mserver5 --version
MonetDB 5 server v11.27.5 "Jul2017-SP1" (64-bit, 128-bit integers)
Copyright (c) 1993-July 2008 CWI
Copyright (c) August 2008-2017 MonetDB B.V., all rights reserved
Visit https://www.monetdb.org/ for further information
Found 2.0GiB available memory, 2 available cpu cores
Libraries:
libpcre: 8.31 2012-07-06 (compiled with 8.31)
openssl: OpenSSL 1.0.1f 6 Jan 2014 (compiled with )
libxml2: 2.9.1 (compiled with 2.9.1)
Compiled by: root@dev.monetdb.org (x86_64-pc-linux-gnu)
Compilation: gcc -O3 -fomit-frame-pointer -pipe -g -D_FORTIFY_SOURCE=2
Linking : /usr/bin/ld -m elf_x86_64
Not sure if this is useful, but here is some debugger info from our servers, when the memory is exhausted:
(gdb) where
0 0x00007f6819301c53 in gethostid () at ../sysdeps/unix/sysv/linux/gethostid.c:101
Cannot access memory at address 0x8
(gdb) thr app all bt
Thread 374 (Thread 0x7f680f7d4700 (LWP 6381)):
0 0x00007f6819301c53 in gethostid () at ../sysdeps/unix/sysv/linux/gethostid.c:101
1 0x0000000000000000 in ?? ()
Thread 373 (Thread 0x7f680f5d3700 (LWP 6363)):
0 sem_wait () at ../nptl/sysdeps/unix/sysv/linux/x86_64/sem_wait.S:85
1 0x00007f681a074600 in ?? () from /usr/lib/libmonetdb5.so.23
2 0x00007f68195dd184 in start_thread (arg=0x7f680f5d3700) at pthread_create.c:312
3 0x00007f681930a37d in eventfd (count=-129114208, flags=0) at ../sysdeps/unix/sysv/linux/eventfd.c:42
4 0x0000000000000000 in ?? ()
Thread 7 (Thread 0x7f68130ef700 (LWP 1151)):
Python Exception <class 'gdb.MemoryError'> Cannot access memory at address 0x21:
0 0x00007f6819301c53 in gethostid () at ../sysdeps/unix/sysv/linux/gethostid.c:101
Cannot access memory at address 0x21
## Comment 25920
Date: 2017-11-27 12:57:02 +0100
From: @sjoerdmullender
The problem is most likely in the storage allocator for the gtrans structure that just keeps on growing.
## Comment 26139
Date: 2018-01-31 08:46:15 +0100
From: Edvinas <<edvinas.rasys>>
Is this issue being looked at? I'm not able to see any reference to it in the recent commits or release notes.
## Comment 26199
Date: 2018-02-15 10:43:08 +0100
From: Edvinas <<edvinas.rasys>>
Retested with 11.27.13 and this is still an issue for us. It's difficult to observe with 100 table rebuilds in the proposed test case, but with 10 000 rebuilds (for i in `seq 1 10000`;) it appears to be leaking around 30-40MB. It does not appear to be dependent on the length of the columns, but it does depend on the number of columns:
10k table rebuilds - 5 columns (varchar 255) ~ 30-40MB
10k table rebuilds - 10 columns (varchar 255) ~ 60MB
10k table rebuilds - 10 columns (varchar 500) ~ 60MB
10k table rebuilds - 20 columns (varchar 500) ~ 110MB
this averages down to around 0.5-0.6KB on 1 column rebuild
our typical single production rebuild scenario involves around 40 tables each with 2000 columns, so extrapolating that would be around 40-50MB leak. Our real world observations appear to be higher than that, but this issue appears to be the main cause.
## Comment 26468
Date: 2018-05-25 21:13:58 +0200
From: Anton Kravchenko <<kravchenko.anton86>>
(In reply to Sjoerd Mullender from comment 1)
> The problem is most likely in the storage allocator for the gtrans structure
> that just keeps on growing.
I hit this memory leak too. Have you tried to test your hypothesis (by printing the size of the storage allocator for the gtrans while creating/droping tables)?
## Comment 26748
Date: 2018-12-24 11:24:52 +0100
From: Thanos <<arkr17997>>
Must be like to visit here this amazing blog it is the nice way to look here http://addprinterwindows10.com/ and install the add wireless printer windows 10 use the usb cable and support the all system easily.
## Comment 26928
Date: 2019-03-19 00:05:34 +0100
From: @mlkersten
Any progress on this issue with the upcoming Apr19 release>
## Comment 26935
Date: 2019-03-19 16:25:28 +0100
From: @mlkersten
Running the script on default branch in March 2019 with sequence 10K does not show any increase in RES size. Running on Fedora. Closing the report for now.
## Comment 27116
Date: 2019-07-09 14:30:25 +0200
From: Edvinas <<edvinas.rasys>>
I've run the script with the latest (11.33.3) version on
Ubuntu 16.04
Ubuntu 18.04
CentOS 7.6.1810
and the issue is still present. When observed in htop both VIRT and RES sizes just keep growing...
## Comment 27183
Date: 2019-07-28 14:57:45 +0200
From: @mlkersten
Running the script against the default on July 27.
It created 10K connections and dropped it.
23297 mk 20 0 1211360 107192 45752 S 0.0 0.3 0:13.88 mserver5
...
23297 mk 20 0 1211360 145036 45752 S 0.5 0.4 0:50.81 mserver5
It shows that the RES size increased with
145036 - 107192 = 37844 KB -> 3.7KB/test
## Comment 27587
Date: 2020-03-14 22:41:39 +0100
From: @mlkersten
Run script with 1000 connections against Default on March 14 2020
22065 mk 20 0 1139068 86196 41416 S 21.5 0.3 0:01.78 mserver5
22065 mk 20 0 1139068 89556 41416 S 0.3 0.3 0:05.14 mserver5
3360 K for 1000 connections.
It shows that the RES size increased with
89556 - 86196 = 3360 KB -> 3.3KB/test
## Comment 27588
Date: 2020-03-14 22:48:52 +0100
From: @mlkersten
And for a 10.000 run we have
25546 mk 20 0 1136608 84416 41888 S 20.8 0.3 0:00.87 mserver5
25546 mk 20 0 1138660 130988 41888 S 14.6 0.4 0:50.21 mserver5
130988 - 84416 = 46572 ->4.6K
This could be the basis for a torture test.
## Comment 27634
Date: 2020-03-29 19:08:21 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [90bdae54a2ef](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=90bdae54a2ef) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=90bdae54a2ef](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=90bdae54a2ef)
Changeset description:
do a deep copy of the global transaction structure, within the store_apply_delta's, when no other 'sessions' are running.
This frees any data leftover from previous drops solving bug #6401.
Current setting is extreme (ie always if possible). We may need to have a tunable for this.
## Comment 27635
Date: 2020-03-29 19:09:04 +0200
From: @njnes
the leak is now fully fixed by doing a deepcopy of the central structures.
| Suspected memory leak in mserver5 when creating/dropping tables | https://api.github.com/repos/MonetDB/MonetDB/issues/6401/comments | 0 | 2020-11-30T15:52:33Z | 2024-06-27T13:00:49Z | https://github.com/MonetDB/MonetDB/issues/6401 | 753,593,233 | 6,401 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-09-01 14:09:30 +0200
From: Fabio <<palmaresk8>>
To: clients devs <<bugs-clients>>
Version: 11.27.5 (Jul2017-SP1)
CC: martin.van.dinther, @PedroTadim
Last updated: 2017-10-26 14:01:40 +0200
## Comment 25607
Date: 2017-09-01 14:09:30 +0200
From: Fabio <<palmaresk8>>
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.113 Safari/537.36
Build Identifier:
I use DBeaver as a GUI (https://dbeaver.jkiss.org/) for MonetDB. The connection to the database is done with the MonetDB JDBC driver. Because some tables where created by R, string columns where defined as 'TEXT' (which MonetDB translate to SQL type 'CLOB'). Problem is: when I try to use DBeaver to visualize data, it gives an error: getCharacterStream() currently not supported. The message also shows the exception: java.sql.SQLFeatureNotSupportedException: Operation getCharacterStream() currently not supported at nl.cwi.monetdb.jdbc.MonetClob.getCharacterStream. This only happens with MonetDB JDBC driver. The same does not occurs with PostgreSQL JDBC driver in the same conditions, for example.
Reproducible: Always
### Steps to Reproduce:
1. Configure MonetDB connection in DBeaver using JDBC driver. Select option 'Cache CLOB values' in Preferences > Database > Result Sets > Binaries to see the content of CLOB columns
2. Create a test table with one column of type 'TEXT' and insert a single string in it: `CREATE TABLE foo (name TEXT); INSERT INTO foo VALUES ('Manolo');`
3. Try to vizualize table contents. Column 'name' will show as 'NULL' (which is not). Double click the cell to watch your content and you will see one exception instead.
### Actual Results:
exception java.sql.SQLFeatureNotSupportedException: Operation getCharacterStream() currently not supported at nl.cwi.monetdb.jdbc.MonetClob.getCharacterStream.
### Expected Results:
String content
If I cast column type from CLOB to VARCHAR(1024), for example, I'm able to see the column content in DBeaver. But this is an expensive operation, because column types update are not implemented in MonetDB yet (that is: I have to create a temporary column, copy the values of the old column to the temporary one, drop the old column, create a new column with the same name as the old one, copy the values of the temporary column to the new one, and finally drop the temporary column). The ideal scenario is MonetDB JDBC driver to implement CLOB cache values function.
## Comment 25608
Date: 2017-09-01 14:36:21 +0200
From: @PedroTadim
We got a commit fixing this issue about two weeks ago: https://dev.monetdb.org/hg/monetdb-java/rev/1e49fc74dba4 It should be fixed in the next release.
## Comment 25609
Date: 2017-09-01 14:36:53 +0200
From: @PedroTadim
We got a commit fixing this issue about two weeks ago: https://dev.monetdb.org/hg/monetdb-java/rev/1e49fc74dba4 It should be available in the next release.
## Comment 25618
Date: 2017-09-07 13:32:49 +0200
From: Martin van Dinther <<martin.van.dinther>>
Fix will be included in the next release (v2.27)
| getCharacterStream() currently not supported | https://api.github.com/repos/MonetDB/MonetDB/issues/6400/comments | 0 | 2020-11-30T15:52:31Z | 2024-06-27T13:00:48Z | https://github.com/MonetDB/MonetDB/issues/6400 | 753,593,205 | 6,400 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-31 17:43:37 +0200
From: @yzchang
To: SQL devs <<bugs-sql>>
Version: 11.27.5 (Jul2017-SP1)
CC: @njnes, @yzchang
Last updated: 2017-10-26 14:01:34 +0200
## Comment 25601
Date: 2017-08-31 17:43:37 +0200
From: @yzchang
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_6) AppleWebKit/603.3.8 (KHTML, like Gecko) Version/10.1.2 Safari/603.3.8
Build Identifier:
When a mixture of subquery and scalar values are passed (in some specific ways, or a specific number of them) as the parameter values to a UDF, mserver5 crashes
Reproducible: Always
### Steps to Reproduce:
Run the following queries to get the crash:
=================
create table data (idx int);
insert into data values (13), (17);
create function crash2(i int, j int) returns int begin
return i + j;
END;
select * from test(1, (select idx from data));
create function crash3(i int, j int, k int) returns int begin
return i + j + k;
END;
select * from test((select idx from data), 2, 3);
=================
### Actual Results:
Crash
### Expected Results:
The error message:
SELECT: 'test' requires a single sub query
1. The query "select * from test((select idx from data), 2);" correctly results in the error message.
2. On Mac OS (Sierra) mserver5 doesn't crash, but runs forever.
## Comment 25602
Date: 2017-08-31 18:07:03 +0200
From: @yzchang
Sorry, forgot to change the names of the functions in the SELECT queries. They should be replaced with "crash2" or "crash3"
## Comment 25603
Date: 2017-08-31 18:21:19 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [69732824cc07](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=69732824cc07) made by Jennie Zhang <y.zhang@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=69732824cc07](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=69732824cc07)
Changeset description:
Added test for Bug #6399
## Comment 25642
Date: 2017-09-17 18:01:51 +0200
From: @njnes
crash is fixed
| UDF crashes when subquery and scalar values are passed as pameters | https://api.github.com/repos/MonetDB/MonetDB/issues/6399/comments | 0 | 2020-11-30T15:52:28Z | 2024-06-27T13:00:46Z | https://github.com/MonetDB/MonetDB/issues/6399 | 753,593,171 | 6,399 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-30 14:00:46 +0200
From: Manuel <<manuel>>
To: SQL devs <<bugs-sql>>
Version: 11.27.5 (Jul2017-SP1)
CC: @njnes
Last updated: 2017-10-26 14:01:37 +0200
## Comment 25600
Date: 2017-08-30 14:00:46 +0200
From: Manuel <<manuel>>
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.113 Safari/537.36
Build Identifier:
Given two tables T1, and T2, and a field f present in both T1, and T2, doing an outer join like
select * from T1 left outer join T2 on T1.f = T2.f
does not account for rows where both T1.f and T2.f are NULL. This is correct, according to the operation semantics.
However, if I want to have matches on null values, the appropriate solution would be a query like
select * from T1 left outer join T2 on (T1.f = T2.f OR (T1.f is NULL AND T2.f is NULL)).
Unfortunately this doesn't work: in fact, this query is the same as:
select * from T1 join T2 on (T1.f = T2.f OR (T1.f is NULL AND T2.f is NULL)).
ON conditions containing clauses on NULL values are supported in INNER joins, but not on LEFT/RIGHT OUTER joins.
I tried to overcame this issue by appending an inner join with a left anti join, but the performance is much worse, and it is a way less cleaner approach.
Reproducible: Always
### Steps to Reproduce:
Create two different tables, for example:
create table "E" (
"intCol" bigint,
"stringCol" string
);
insert into "E" values (0, 'zero');
insert into "E" values (1, 'one');
insert into "E" values (2, 'two');
insert into "E" values (null, null);
create table "I" (
"intCol" bigint,
"stringCol" string
);
insert into "I" values (2, 'due');
insert into "I" values (4, 'quattro');
insert into "I" values (null, 'this is not null');
select * from "E" left outer join "I" on "E"."intCol" = "I"."intCol" or ("E"."intCol" is null and "I"."intCol" is null);
### Actual Results:
intCol stringCol intCol stringCol
---------------------------------------------
2 two 2 due
<null> <null> <null> this is not null
### Expected Results:
intCol stringCol intCol stringCol
--------------------------------------
2 two 2 due
0 zero <null> <null>
1 one <null> <null>
<null> <null> <null> this is not null
## Comment 25645
Date: 2017-09-17 21:33:30 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [0260b270032c](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=0260b270032c) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=0260b270032c](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=0260b270032c)
Changeset description:
fixed bug #6398, ie handle null matches in outer joins
## Comment 25646
Date: 2017-09-17 21:34:10 +0200
From: @njnes
fixed both a crash and broken rewrites.
| Null Matches in outer join are not supported | https://api.github.com/repos/MonetDB/MonetDB/issues/6398/comments | 0 | 2020-11-30T15:52:25Z | 2024-06-27T13:00:45Z | https://github.com/MonetDB/MonetDB/issues/6398 | 753,593,135 | 6,398 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-29 11:17:33 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-10-26 14:01:41 +0200
## Comment 25598
Date: 2017-08-29 11:17:33 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:55.0) Gecko/20100101 Firefox/55.0
Build Identifier:
Functions that declare and return local tables are not isolated properly.
The DECLARE local table behaves as being a global DDL, which means that
the actual state becomes unpredictable.
A test case is added.
Reproducible: Always
### Steps to Reproduce:
CREATE FUNCTION my_generate_series(start INT, finish INT)
RETURNS TABLE (value INT)
BEGIN
DECLARE TABLE tmp_generate_series(value INT);
DECLARE i INT;
SET i = start;
WHILE (i <= finish) DO
INSERT INTO tmp_generate_series VALUES(i);
SET i = i + 1;
END WHILE;
RETURN tmp_generate_series;
END;
explain select count(*) from my_generate_series(1,5) as t1,
my_generate_series(1,100) as t2;
select count(*) from my_generate_series(1,5) as t1,
my_generate_series(1,100) as t2;
select count(*) from my_generate_series(1,5) as t1,
my_generate_series(1,100) as t2;
select count(*) from my_generate_series(1,5) as t1,
my_generate_series(1,100) as t2;
## Comment 25599
Date: 2017-08-29 11:22:51 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [dc63791e4dc6](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=dc63791e4dc6) made by Martin Kersten <mk@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=dc63791e4dc6](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=dc63791e4dc6)
Changeset description:
Added bug #6397
## Comment 25638
Date: 2017-09-17 09:45:27 +0200
From: @njnes
reimplementated declared tables. Each call now uses the mal call stack, solving the isolation problem.
## Comment 25640
Date: 2017-09-17 18:01:00 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [9ccae5c8d830](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=9ccae5c8d830) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=9ccae5c8d830](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=9ccae5c8d830)
Changeset description:
re-implemented declared tables, to solve the side-effect bug #6397.
also lots of fixes for the sqllitelogistest_case bug #6335 (doesn't
crash (or run out of memory) anymore).
| Isolation of generating functions not correct | https://api.github.com/repos/MonetDB/MonetDB/issues/6397/comments | 0 | 2020-11-30T15:52:23Z | 2024-06-27T13:00:44Z | https://github.com/MonetDB/MonetDB/issues/6397 | 753,593,103 | 6,397 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-22 18:58:47 +0200
From: Sreejith Sharma <<sreejith.sharma>>
To: GDK devs <<bugs-common>>
Version: 11.25.23 (Dec2016-SP5)
Last updated: 2017-08-24 07:49:04 +0200
## Comment 25592
Date: 2017-08-22 18:58:47 +0200
From: Sreejith Sharma <<sreejith.sharma>>
User-Agent: Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.103 Safari/537.36
Build Identifier:
I have been running 1000+ SQLs with concurrency of 10/8/5/3 and run in a loop. Noticed memory bumps up even the same set of SQL's has been rerun. Post SQL completion the memory still stays on peak and never comes down unless restarted. Is this nature expected? Tried multiple optimizer options.
Reproducible: Always
### Steps to Reproduce:
1. Running series of SQL with concurrency
2. SQL can consume columns ranging from 2-15 with approx 3M rows
3. Even rerunning of the same set of SQL in shows memory increasing
### Actual Results:
Memory keeps increasing even the same set of SQL has rerun on high concurrency
### Expected Results:
Memory should release once each set of SQL is completed or if the same set of SQL has rerun the memory should not increase furthermore
| Noticed "Memory Leak" with high concurrency and same SQL run in loop | https://api.github.com/repos/MonetDB/MonetDB/issues/6396/comments | 0 | 2020-11-30T15:52:21Z | 2024-06-28T13:10:53Z | https://github.com/MonetDB/MonetDB/issues/6396 | 753,593,072 | 6,396 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-22 14:42:40 +0200
From: Richard Hughes <<richard.monetdb>>
To: SQL devs <<bugs-sql>>
Version: 11.27.5 (Jul2017-SP1)
Last updated: 2017-10-26 14:01:34 +0200
## Comment 25590
Date: 2017-08-22 14:42:40 +0200
From: Richard Hughes <<richard.monetdb>>
Build is Jul2017 [a55ec984eb42](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=a55ec984eb42)
To reproduce:
sql>select count(*) from sys.bbp() where kind='transient';
+------+
| L4 |
+======+
| 161 |
+------+
1 tuple (11.032ms)
sql>select count(*) from sys.bbp() where kind='transient';
+------+
| L4 |
+======+
| 163 |
+------+
1 tuple (9.218ms)
sql>select count(*) from sys.bbp() where kind='transient';
+------+
| L4 |
+======+
| 165 |
+------+
1 tuple (9.192ms)
sql>select count(*) from sys.bbp() where kind='transient';
+------+
| L4 |
+======+
| 167 |
+------+
1 tuple (9.083ms)
sql>select count(*) from sys.bbp() where kind='transient';
+------+
| L4 |
+======+
| 169 |
+------+
1 tuple (8.975ms)
Expected: that number should not be increasing
The issue can be reproduced with any query which is guaranteed to return one row, e.g. "select 1;". I used the example above because it both reproduces and demonstrates the issue in a single query.
Each run of the query leaks 1+<number_of_result_columns> BATs.
The issue appears to be in mvc_scalar_value_wrap(), which is calling mvc_result_table() and then discarding the result so it can never call res_tables_remove(res_tables_find()) on it. The equivalent code in mvc_row_result_wrap() (i.e. for multi-row results, which don't leak) is rather different so I'm not sure what's the neatest patch.
## Comment 25591
Date: 2017-08-22 17:45:39 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [2167d80ea078](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=2167d80ea078) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=2167d80ea078](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=2167d80ea078)
Changeset description:
When creating BATs to wrap scalar results, unfix them.
This fixes bug #6395.
## Comment 25593
Date: 2017-08-23 12:03:08 +0200
From: Richard Hughes <<richard.monetdb>>
Confirmed.
Sorry for the misleading commentary above - I did most of my debugging on Jun2016 then only tested its existence on Jul2017.
## Comment 25595
Date: 2017-08-23 13:34:06 +0200
From: @sjoerdmullender
Your comments did help in locating the problem.
I'm glad the fix is confirmed.
| BAT leak of scalar result sets | https://api.github.com/repos/MonetDB/MonetDB/issues/6395/comments | 0 | 2020-11-30T15:52:18Z | 2024-06-27T13:00:42Z | https://github.com/MonetDB/MonetDB/issues/6395 | 753,593,040 | 6,395 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-17 11:16:32 +0200
From: @mlkersten
To: MonetDB5 devs <<bugs-monetdb5>>
Version: 11.27.5 (Jul2017-SP1)
Last updated: 2017-08-17 11:16:32 +0200
## Comment 25586
Date: 2017-08-17 11:16:32 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:54.0) Gecko/20100101 Firefox/54.0
Build Identifier:
The current implementation of MAL remote allows sending a nil-bat, but the protocol is not strong enough to also detect receiving them. The code too quickly
jumps into reading a bat.
Reproducible: Always
| Remote MAL does not accept nil:bat[:oid] | https://api.github.com/repos/MonetDB/MonetDB/issues/6394/comments | 0 | 2020-11-30T15:52:15Z | 2024-06-28T13:10:52Z | https://github.com/MonetDB/MonetDB/issues/6394 | 753,592,998 | 6,394 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6393/comments | 0 | 2020-11-30T15:52:13Z | 2020-11-30T15:52:14Z | https://github.com/MonetDB/MonetDB/issues/6393 | 753,592,953 | 6,393 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-14 11:22:50 +0200
From: @joerivanruth
To: SQL devs <<bugs-sql>>
Version: 11.27.5 (Jul2017-SP1)
CC: @njnes
Last updated: 2017-10-26 14:01:37 +0200
## Comment 25578
Date: 2017-08-14 11:22:50 +0200
From: @joerivanruth
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.90 Safari/537.36
Build Identifier:
I would expect EXISTS to return false when applied to an empty query result.
Reproducible: Always
### Steps to Reproduce:
CREATE TABLE foo(i INT);
SELECT EXISTS (SELECT i FROM foo WHERE FALSE);
### Actual Results:
true
### Expected Results:
false, because the result set of the inner selection is empty
## Comment 25581
Date: 2017-08-16 18:20:58 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [4a7ab58ad31a](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=4a7ab58ad31a) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=4a7ab58ad31a](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=4a7ab58ad31a)
Changeset description:
fixed bugs 6386 and 6392
## Comment 25583
Date: 2017-08-16 18:21:45 +0200
From: @njnes
fixed the implementation of exist_val.
| SELECT EXISTS (empty table) returns 'true' | https://api.github.com/repos/MonetDB/MonetDB/issues/6392/comments | 0 | 2020-11-30T15:52:10Z | 2024-06-27T13:00:40Z | https://github.com/MonetDB/MonetDB/issues/6392 | 753,592,914 | 6,392 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6391/comments | 0 | 2020-11-30T15:52:06Z | 2020-11-30T15:52:09Z | https://github.com/MonetDB/MonetDB/issues/6391 | 753,592,872 | 6,391 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6390/comments | 0 | 2020-11-30T15:52:04Z | 2020-11-30T15:52:05Z | https://github.com/MonetDB/MonetDB/issues/6390 | 753,592,841 | 6,390 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6389/comments | 0 | 2020-11-30T15:52:02Z | 2020-11-30T15:52:03Z | https://github.com/MonetDB/MonetDB/issues/6389 | 753,592,800 | 6,389 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-10 17:43:53 +0200
From: Martin van Dinther <<martin.van.dinther>>
To: Martin van Dinther <<martin.van.dinther>>
Version: 11.27.5 (Jul2017-SP1)
Last updated: 2017-10-26 15:44:56 +0200
## Comment 25564
Date: 2017-08-10 17:43:53 +0200
From: Martin van Dinther <<martin.van.dinther>>
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:54.0) Gecko/20100101 Firefox/54.0
Build Identifier:
Fetching meta data from a JDBC connection which is not from user monetdb (or has monetdb privileges granted) fails for following MonetDatabaseMetaData methods:
getUserName()
getMaxConnections()
getDatabaseProductVersion()
getDatabaseMajorVersion()
getDatabaseMinorVersion()
with eror:
SELECT: access denied for voc to table 'sys.environment'
and for:
getTables()
getTableTypes()
with eror:
SELECT: access denied for voc to table 'sys.table_types'
Reproducible: Always
### Steps to Reproduce:
1. connect as monetdb user
2. create voc user and schema as listed at https://www.monetdb.org/Documentation/UserGuide/Tutorial
CREATE USER "voc" WITH PASSWORD 'voc' NAME 'VOC Explorer' SCHEMA "sys";
CREATE SCHEMA "voc" AUTHORIZATION "voc";
ALTER USER "voc" SET SCHEMA "voc";
3. connect as voc user
4. call DatabaseMetaData methods:
getUserName()
getMaxConnections()
getDatabaseProductVersion()
getDatabaseMajorVersion()
getDatabaseMinorVersion()
getTables()
getTableTypes()
### Actual Results:
SQLException errors:
SELECT: access denied for voc to table 'sys.environment'
and
SELECT: access denied for voc to table 'sys.table_types'
### Expected Results:
no errors
This issue was reported by Sandy on MonetDB users-list on 24-06-2017 01:46 with subject "Setting Up Tools"
See: https://www.monetdb.org/pipermail/users-list/2017-June/009837.html
## Comment 25565
Date: 2017-08-10 18:10:52 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset 8b40a845240c, made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the monetdb-java repo, refers to this bug.
For complete details, see https://dev.monetdb.org/hg/monetdb-java?cmd=changeset;node=8b40a845240c
Changeset description:
Added test program for Bug #6388
and a fix for calling metadata methods:
getUserName()
getMaxConnections()
getDatabaseProductVersion()
getDatabaseMajorVersion()
getDatabaseMinorVersion()
ToDo: for calling
getTables()
getTableTypes()
we need to do changes in monetdb server code.
## Comment 25566
Date: 2017-08-10 18:12:49 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [f9f9e4953d89](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=f9f9e4953d89) made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=f9f9e4953d89](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=f9f9e4953d89)
Changeset description:
Activating JDBC test for bug #6388.
## Comment 25588
Date: 2017-08-17 18:18:06 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [49d4903a4dff](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=49d4903a4dff) made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=49d4903a4dff](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=49d4903a4dff)
Changeset description:
Added GRANT SELECT TO PUBLIC for 8 system tables and 1 system view
and GRANT EXECUTE for 1 system function.
This solves issue reported in bug #6388.
## Comment 25812
Date: 2017-10-26 15:44:56 +0200
From: Martin van Dinther <<martin.van.dinther>>
New JDBC driver monetdb-jdbc-2.27.jar has been released
| JDBC Connection via user voc produces errors when fetching certain meta data information | https://api.github.com/repos/MonetDB/MonetDB/issues/6388/comments | 0 | 2020-11-30T15:51:58Z | 2024-06-27T13:00:38Z | https://github.com/MonetDB/MonetDB/issues/6388 | 753,592,750 | 6,388 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-07 19:34:42 +0200
From: Manuel <<manuel>>
To: SQL devs <<bugs-sql>>
Version: 11.27.5 (Jul2017-SP1)
Last updated: 2017-10-26 14:01:44 +0200
## Comment 25557
Date: 2017-08-07 19:34:42 +0200
From: Manuel <<manuel>>
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36
Build Identifier:
I noticed a quite drastic performance degradation, once I added more than one order by clauses.
For example the query
select * from "fdb6d859a24DM_FULFile_defau_" order by "name_" limit 500;
takes 0.826 seconds on my machine
whereas the query
select "id_", "name_" from "fdb6d859a24DM_FULFile_defau_" order by "name_", "id_" limit 500
takes over 12 seconds on the same machine.
Am I doing anything wrong ? Is there a way to improve performance on these kinds of queries ?
Reproducible: Always
### Steps to Reproduce:
1.Import the data from hr.sql
2. Compare the performance between the two queries
### Actual Results:
An order by clause with multiple columns cause a very significant performance drop
## Comment 25558
Date: 2017-08-07 19:46:01 +0200
From: Manuel <<manuel>>
Created attachment 566
Sample data
I am not able to attach the sql file as the size is too large. This is the schema of the table
CREATE TABLE "sys"."fdb6d859a24DM_FULFile_defau_" (
"id_" BIGINT,
"name_" VARCHAR(255),
"sex_" VARCHAR(255),
"race_" VARCHAR(255),
"location_" VARCHAR(255),
"department_" VARCHAR(255),
"level_" BIGINT,
"date_of_birth_" TIMESTAMP
);
id_ is assumes value 1,2,3 ....
The table contains 20 million records. Attached a sample of the records.
> Attached file: [human resources 10k.csv](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6387_human_resources_10k.csv_566) (application/vnd.ms-excel, 617844 bytes)
> Description: Sample data
## Comment 25594
Date: 2017-08-23 13:15:50 +0200
From: @sjoerdmullender
The implementation seems to be inefficient.
If you can live with a smaller limit, try that. For instance, compare the two queries with limit 5 instead of limit 500. It should make a huge difference.
## Comment 25596
Date: 2017-08-25 14:48:54 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [f65ce66374ad](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=f65ce66374ad) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=f65ce66374ad](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=f65ce66374ad)
Changeset description:
Reimplemented BATfirstn_grouped to use higher level functions.
This speeds up two column order bys with largish limits by several
orders.
Fix for bug #6387.
## Comment 25597
Date: 2017-08-25 15:27:39 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [7a7283729ef0](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=7a7283729ef0) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7a7283729ef0](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7a7283729ef0)
Changeset description:
Reimplemented some versions of BATfirstn to not use a nested loop.
This should fix the performance problem of bug #6387.
## Comment 25610
Date: 2017-09-05 10:59:52 +0200
From: @sjoerdmullender
In my own test using the part table from TPC-H scale factor 100, time went down from about 47.5 seconds to 3.2 seconds.
| Performance degradation on multi column sort | https://api.github.com/repos/MonetDB/MonetDB/issues/6387/comments | 0 | 2020-11-30T15:51:55Z | 2024-06-27T13:00:37Z | https://github.com/MonetDB/MonetDB/issues/6387 | 753,592,706 | 6,387 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-07 10:12:25 +0200
From: Roman <<kislenok.roman>>
To: SQL devs <<bugs-sql>>
Version: 11.25.23 (Dec2016-SP5)
CC: @njnes
Last updated: 2017-10-26 14:01:30 +0200
## Comment 25554
Date: 2017-08-07 10:12:25 +0200
From: Roman <<kislenok.roman>>
User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.36
Build Identifier:
After I updated my database from v11.23.13 to v11.25.23 I began get some strange error in a simple query. I've reduced original query to reproducible one:
select case name when 'types' then 1 when 'args' then 60.3281 when 'keys' then 0.8415381117315 else 0 end as t1 from tables limit 1;
I expect some numbers from this query, but server (in mclient) return one of this error:
* value (8415381117315) exceeds limits of type int, or
* too many digits (10 > 6)
For most time I got the first one.
Reproducible: Always
### Steps to Reproduce:
1. start mclient
2. run query: "select case name when 'types' then 1 when 'args' then 60.3281 when 'keys' then 0.8415381117315 else 0 end as t1 from tables limit 1;"
### Actual Results:
"value (8415381117315) exceeds limits of type int" or "too many digits (10 > 6)"
## Comment 25555
Date: 2017-08-07 10:13:30 +0200
From: Roman <<kislenok.roman>>
As an example:
sql>select case name when 'types' then 1 when 'args' then 60.3281 when 'keys' then 0.8415381117315 else 0 end as t1 from tables;
value (8415381117315) exceeds limits of type int
sql>select case name when 'types' then 1 when 'args' then 60.3281 when 'keys' then 0.8415381117315 else 0 end as t1 from tables;
value (8415381117315) exceeds limits of type int
sql>select case name when 'types' then 1 when 'args' then 60.3281 when 'keys' then 0.8415381117315 else 0 end as t1 from tables;
value (8415381117315) exceeds limits of type int
sql>select case name when 'types' then 1 when 'args' then 60.3281 when 'keys' then 0.8415381117315 else 0 end as t1 from tables;
value (8415381117315) exceeds limits of type int
sql>select case name when 'types' then 1 when 'args' then 60.3281 when 'keys' then 0.8415381117315 else 0 end as t1 from tables;
too many digits (10 > 6)
sql>select case name when 'types' then 1 when 'args' then 60.3281 when 'keys' then 0.8415381117315 else 0 end as t1 from tables limit 1;
value (8415381117315) exceeds limits of type int
## Comment 25580
Date: 2017-08-16 15:22:03 +0200
From: @njnes
fixed a problem in the supertype function (super and one of the inputs can point
to the same structure)
## Comment 25582
Date: 2017-08-16 18:21:01 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [4a7ab58ad31a](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=4a7ab58ad31a) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=4a7ab58ad31a](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=4a7ab58ad31a)
Changeset description:
fixed bugs 6386 and 6392
| Unexpected error from server for query with long floats | https://api.github.com/repos/MonetDB/MonetDB/issues/6386/comments | 0 | 2020-11-30T15:51:51Z | 2024-06-27T13:00:36Z | https://github.com/MonetDB/MonetDB/issues/6386 | 753,592,659 | 6,386 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-04 16:01:27 +0200
From: @yzchang
To: SQL devs <<bugs-sql>>
Version: 11.27.5 (Jul2017-SP1)
CC: @yzchang
Last updated: 2017-10-26 14:01:44 +0200
## Comment 25551
Date: 2017-08-04 16:01:27 +0200
From: @yzchang
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_6) AppleWebKit/603.3.8 (KHTML, like Gecko) Version/10.1.2 Safari/603.3.8
Build Identifier:
MonetDB seems to be not able to handle user defined AGGREGATE functions with more than 2 parameters. It doesn't recognise such functions as aggregate functions, therefore, it always gives the error "cannot use non GROUP BY column ... in query results without an aggregate function". This is demonstrated by the calls to "aggr3" and "aggr4" in the SELECT queries below.
In addition, the error messages involving AGGR UDFs can also be improved. For instance, if I call an AGGR UDF with the wrong number of parameters (e.g. the queries below calling "aggr2"), I also get the error "cannot use non GROUP BY column ... in query results without an aggregate function". It would be clearer to say something like "Aggregate function <...> not found".
Reproducible: Always
### Steps to Reproduce:
Run the following queries:
CREATE AGGREGATE aggr2(val INTEGER, val2 INTEGER)
RETURNS INTEGER
LANGUAGE PYTHON {
unique = numpy.unique(aggr_group)
x = numpy.zeros(shape=(unique.size))
for i in range(0, unique.size):
x[i] = numpy.sum(val[aggr_group==unique[i]])
return(x)
};
CREATE AGGREGATE aggr3(val INTEGER, val2 INTEGER, val3 INTEGER)
RETURNS DOUBLE
LANGUAGE PYTHON {
unique = numpy.unique(aggr_group)
x = numpy.zeros(shape=(unique.size))
for i in range(0, unique.size):
x[i] = numpy.sum(val[aggr_group==unique[i]])
return(x)
};
CREATE AGGREGATE aggr4(val INTEGER, val2 INTEGER, val3 INTEGER, val4 INTEGER)
RETURNS DOUBLE
LANGUAGE PYTHON {
unique = numpy.unique(aggr_group)
x = numpy.zeros(shape=(unique.size))
for i in range(0, unique.size):
x[i] = numpy.sum(val[aggr_group==unique[i]])
return(x)
};
CREATE TABLE grouped_ints (value INTEGER, groupnr INTEGER, dbvalue double);
INSERT INTO grouped_ints VALUES (1, 0, 0.11), (2, 1, 0.22), (3, 0, 0.33), (4,1, 0.44), (5,0, 0.55);
SELECT groupnr, aggr3(value, value, value) FROM grouped_ints GROUP BY groupnr;
SELECT groupnr, aggr4(value, value, value, value) FROM grouped_ints GROUP BY groupnr;
SELECT groupnr, aggr2(value) FROM grouped_ints GROUP BY groupnr;
SELECT groupnr, aggr2(value, value, value) FROM grouped_ints GROUP BY groupnr;
DROP TABLE grouped_ints;
DROP AGGREGATE aggr2;
DROP AGGREGATE aggr3;
DROP AGGREGATE aggr4;
### Actual Results:
All 4 SELECT queries give the error "cannot use non GROUP BY column ... in query results without an aggregate function"
### Expected Results:
The first 2 SELECT queries using "aggr3" and "aggr4" should give the following results:
+---------+------+
| groupnr | L3 |
+=========+======+
| 0 | 9 |
| 1 | 6 |
+---------+------+
The last 2 SELECT queries should give errors such as:
no such aggregate function 'aggr2(integer)'
no such aggregate function 'aggr2(integer, integer, integer)'
## Comment 25552
Date: 2017-08-04 16:41:05 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [078c96c28337](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=078c96c28337) made by Jennie Zhang <y.zhang@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=078c96c28337](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=078c96c28337)
Changeset description:
Added tests for Bug #6385
## Comment 25553
Date: 2017-08-07 08:57:45 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [517019db709e](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=517019db709e) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=517019db709e](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=517019db709e)
Changeset description:
fixes for bug #6385, ie check for a aggregate earlier (also improved the
error in case of a missing aggregate function)
## Comment 25556
Date: 2017-08-07 16:06:13 +0200
From: @yzchang
Works for me now, closing
| AGGREGATE UDFs with more than 2 parameters incorrectly processed | https://api.github.com/repos/MonetDB/MonetDB/issues/6385/comments | 0 | 2020-11-30T15:51:48Z | 2024-06-27T13:00:35Z | https://github.com/MonetDB/MonetDB/issues/6385 | 753,592,613 | 6,385 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-04 09:57:47 +0200
From: Zhouyang <<jiazhouyang09>>
To: Merovingian devs <<bugs-merovingian>>
Version: 11.27.5 (Jul2017-SP1)
Last updated: 2017-10-26 14:01:28 +0200
## Comment 25547
Date: 2017-08-04 09:57:47 +0200
From: Zhouyang <<jiazhouyang09>>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_3) AppleWebKit/602.4.8 (KHTML, like Gecko) Version/10.0.3 Safari/602.4.8
Build Identifier:
After setting a wrong listenaddr, monetdbd will crash when starting.
Reproducible: Always
### Steps to Reproduce:
1. monetdbd create /a/new/dbfarm/path
2. monetdbd set listenaddr=wronghost /a/new/dbfarm/path
3. monetdbd start -n /a/new/dbfarm/path
### Actual Results:
Aborted
### Expected Results:
Exit and output proper log messages.
The reason is that there is no error handling code after library function "gethostbyname".
Attached please find the patch against version MonetDB-11.27.5.
## Comment 25548
Date: 2017-08-04 10:00:25 +0200
From: Zhouyang <<jiazhouyang09>>
Created attachment 565
fix the crash by adding error handling code to gethostbyname
> Attached file: [monetdb-11.27.5-fix-crash.patch](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6384_monetdb-11.27.5-fix-crash.patch_565) (text/plain, 575 bytes)
> Description: fix the crash by adding error handling code to gethostbyname
## Comment 25549
Date: 2017-08-04 10:41:59 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [fedb63ea4b33](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=fedb63ea4b33) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=fedb63ea4b33](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=fedb63ea4b33)
Changeset description:
Add error handling.
Fixes bug #6384.
## Comment 25550
Date: 2017-08-04 10:42:22 +0200
From: @sjoerdmullender
Thanks for the patch. I added cleanup as well.
| crash when setting a wrong listenaddr | https://api.github.com/repos/MonetDB/MonetDB/issues/6384/comments | 0 | 2020-11-30T15:51:44Z | 2024-06-27T13:00:34Z | https://github.com/MonetDB/MonetDB/issues/6384 | 753,592,569 | 6,384 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-08-01 14:50:33 +0200
From: Ramiro Batista da Luz <<ramiroluz>>
To: buildtools devs <<bugs-buildtools>>
Version: 11.27.1 (Jul2017)
Last updated: 2017-08-03 13:19:50 +0200
## Comment 25531
Date: 2017-08-01 14:50:33 +0200
From: Ramiro Batista da Luz <<ramiroluz>>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Firefox/52.0
Build Identifier:
Following the documented instructions[1] to install monetdb on debian Stretch I got the following error:
W: The repository 'http://dev.monetdb.org/downloads/deb stretch Release' does not have a Release file.
N: Data from such a repository can't be authenticated and is therefore potentially dangerous to use.
Followed by:
https://dev.monetdb.org/downloads/deb/dists/stretch/monetdb/source/Sources gnutls_handshake() failed: Public key signature verification has failed.
It might be related to: https://wiki.debian.org/Teams/Apt/Sha1Removal
[1] - https://www.monetdb.org/downloads/deb/
Reproducible: Always
### Steps to Reproduce:
1. On a debian stretch
2. Follow the steps in https://www.monetdb.org/downloads/deb/
2.1: Create a file /etc/apt/sources.list.d/monetdb.list with:
deb http://dev.monetdb.org/downloads/deb/ stretch monetdb
deb-src http://dev.monetdb.org/downloads/deb/ stretch monetdb
2.2: Install the MonetDB GPG public key:
wget --output-document=- https://www.monetdb.org/downloads/MonetDB-GPG-KEY | sudo apt-key add -
2.3: Run apt update.
sudo apt update
3. The error appears at the end of the update messages.
### Actual Results:
It is not possible to install monetdb.
### Expected Results:
Install monetdb
The monetdb repository is listed as half broken on:
https://wiki.debian.org/Teams/Apt/Sha1Removal
There are instructions to fix half broken repositories.
Fixing half-broken repositories
The repository owner needs to pass --digest-algo SHA512 or --digest-algo SHA256 (or another SHA2 algorithm) to gpg when signing the file. Repositories with DSA keys need to be migrated to RSA first.
Migrating from DSA to RSA is best done by signing the repository with two keys (old and new one) and shipping the new one to the users.
To sign with more than one key, if using reprepro, use a space-separated list of key IDs in the conf/distributions file on the SignWith line.
A relatively safe way to ship the key would be to embed it in the package. To embed the key in the package, export it into its own keyring, like so
gpg --export -a YOURNEWKEYID | gpg --no-default-keyring --keyring newkeyring.gpg --import -
The keyring file will be created in your ~/.gnupg directory. Place it into /etc/apt/trusted.gpg.d/ directory in your package, and it will automatically be picked up by apt-key once installed by your users.
Some months after those changes, it is OK to drop the old key from the repository and the users machines (if shipped with a package).
## Comment 25532
Date: 2017-08-02 10:11:25 +0200
From: @sjoerdmullender
I can reproduce the problem with installing MonetDB. We had tried on various Ubuntu systems, but not on Debian stretch.
I'll take a good look at the pointers you gave.
## Comment 25542
Date: 2017-08-02 16:31:36 +0200
From: @sjoerdmullender
I have updated the MONETDB-GPG-KEY and resigned the Release files. They currently have two signatures (old and new).
Can you please try again and report back?
You need to reimport the GPG key (there is a new one together with the old one), run apt update again, and then you should (hopefully) be able to install the packages.
## Comment 25544
Date: 2017-08-02 17:20:44 +0200
From: Ramiro Batista da Luz <<ramiroluz>>
It didn't worked.
Then I found a message in a debian list.
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=83472415
The problem was with the libgnutls. I am sorry.
Removed the lib: libgnutls-deb0-28
Then it worked.
Thank you for your attention.
I think that monetdb repository can now be on the Compliant List on the Debian wiki.
## Comment 25546
Date: 2017-08-03 13:19:50 +0200
From: @sjoerdmullender
Thanks for the feedback. Closing the issue.
| Debian apt failling to install | https://api.github.com/repos/MonetDB/MonetDB/issues/6383/comments | 0 | 2020-11-30T15:51:41Z | 2024-06-27T13:00:33Z | https://github.com/MonetDB/MonetDB/issues/6383 | 753,592,510 | 6,383 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-30 23:48:37 +0200
From: Brian Ploetz <<bploetz>>
To: Martin van Dinther <<martin.van.dinther>>
Version: 11.27.1 (Jul2017)
CC: martin.van.dinther
Last updated: 2017-10-26 15:44:10 +0200
## Comment 25527
Date: 2017-07-30 23:48:37 +0200
From: Brian Ploetz <<bploetz>>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36
Build Identifier:
I have a JSON column in my table, and when trying to insert into that table using a PreparedStatement...
```
statement.setString(6, "{}");
```
...I get the following error:
```
wrong type for argument 6 of function call: char, expected json
```
I also tried setObject to no avail. My assumption was that I could pass a String containing valid JSON, and the driver would convert it to the correct type under the hood, but it doesn't seem to be supported at all.
Reproducible: Always
## Comment 25604
Date: 2017-08-31 18:51:24 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset 8ea360b612de, made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the monetdb-java repo, refers to this bug.
For complete details, see https://dev.monetdb.org/hg/monetdb-java?cmd=changeset;node=8ea360b612de
Changeset description:
Corrected PreparedStatement methods setString(int, String) and setObject(int, Object, ...) in case the target parameter data type was json or inet or url or uuid.
This fixes bug #6382.
Also added java test program.
## Comment 25606
Date: 2017-08-31 19:12:21 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [3ef07832d018](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=3ef07832d018) made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=3ef07832d018](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=3ef07832d018)
Changeset description:
Activating java test program Bug_PrepStmtSetString_6382 for bug #6382
## Comment 25811
Date: 2017-10-26 15:44:10 +0200
From: Martin van Dinther <<martin.van.dinther>>
New JDBC driver monetdb-jdbc-2.27.jar has been released
| Can't set JSON fields via PreparedStatement | https://api.github.com/repos/MonetDB/MonetDB/issues/6382/comments | 0 | 2020-11-30T15:51:38Z | 2024-06-27T13:00:32Z | https://github.com/MonetDB/MonetDB/issues/6382 | 753,592,462 | 6,382 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-28 23:10:42 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-10-26 14:01:32 +0200
## Comment 25525
Date: 2017-07-28 23:10:42 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:54.0) Gecko/20100101 Firefox/54.0
Build Identifier:
The parsing of 'while (true)' and 'while (select true)' behave differently while it shoudn't
A test file has beeb added
Reproducible: Always
## Comment 25526
Date: 2017-07-28 23:14:51 +0200
From: @mlkersten
test file parsererror.Bug-6381.sql
## Comment 25559
Date: 2017-08-07 20:20:21 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [6aabecbe1486](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=6aabecbe1486) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=6aabecbe1486](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=6aabecbe1486)
Changeset description:
fixes for bug #6381, ie allow select in if/while constructs
approved output
## Comment 25560
Date: 2017-08-07 20:21:28 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [62194816af09](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=62194816af09) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=62194816af09](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=62194816af09)
Changeset description:
fix for bug #6381
## Comment 25561
Date: 2017-08-07 20:21:48 +0200
From: @njnes
fixed, ie handle query in if/while constructs
| Parser misses error messages in conditional | https://api.github.com/repos/MonetDB/MonetDB/issues/6381/comments | 0 | 2020-11-30T15:51:34Z | 2024-06-27T13:00:31Z | https://github.com/MonetDB/MonetDB/issues/6381 | 753,592,429 | 6,381 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-28 00:02:29 +0200
From: Javier Pastorino <<jpastorino.ucdenver>>
To: MonetDB5 devs <<bugs-monetdb5>>
Version: 11.27.1 (Jul2017)
Last updated: 2017-10-26 14:01:42 +0200
## Comment 25509
Date: 2017-07-28 00:02:29 +0200
From: Javier Pastorino <<jpastorino.ucdenver>>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36
Build Identifier:
dba@csci5559:~$ monetdbd create ./myfarm
dba@csci5559:~$ monetdbd start ./myfarm
dba@csci5559:~$ monetdb create voc
created database in maintenance mode: voc
dba@csci5559:~$ monetdb release voc
taken database out of maintenance mode: voc
dba@csci5559:~$ mclient -u monetdb -d voc
password:
TypeException:user.sql.init[5]:'sqlcatalog.create_function' undefined in: sqlcatalog.create_function(X_1:str, X_3:str, X_5:ptr);
MALException:optimizer.MALoptimizer:Start with inconsistent MAL plan
Reproducible: Always
### Steps to Reproduce:
1.Install monetdb and monet client
2.create a farm
3.create a database and connect to initialize
### Actual Results:
Just follow the documentation online for ubuntu and tutorial
### Expected Results:
TypeException:user.sql.init[5]:'sqlcatalog.create_function' undefined in: sqlcatalog.create_function(X_1:str, X_3:str, X_5:ptr);
MALException:optimizer.MALoptimizer:Start with inconsistent MAL plan
initialized and connected.
## Comment 25511
Date: 2017-07-28 11:01:39 +0200
From: @sjoerdmullender
Packaging error on our side for Debian and Ubuntu packages. There is a
file missing in the installer.
To fix: copy this file [1] to /usr/lib/monetdb5/sqlcatalog.mal
This is fixed in the upcoming Jul2017-SP1 release. A release candidate is currently available from [2].
[1]
https://dev.monetdb.org/hg/MonetDB/raw-file/Jul2017_1/sql/backends/monet5/sqlcatalog.mal
[2] https://dev.monetdb.org/downloads/testing/
| unable to create new databases from clean installation | https://api.github.com/repos/MonetDB/MonetDB/issues/6380/comments | 0 | 2020-11-30T15:51:31Z | 2024-06-27T13:00:30Z | https://github.com/MonetDB/MonetDB/issues/6380 | 753,592,399 | 6,380 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-27 19:19:55 +0200
From: Dean De Leo <<dleo>>
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-10-26 14:01:39 +0200
## Comment 25505
Date: 2017-07-27 19:19:55 +0200
From: Dean De Leo <<dleo>>
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.40 Safari/537.36
Build Identifier:
The SQL front-end crashes when a SQL statement contains an invocation to a non existing function
Reproducible: Always
### Steps to Reproduce:
See attached repro.
### Actual Results:
The SQL frontend attempts to access a null pointer and hits a segmentation fault. See attached stack trace.
### Expected Results:
An meaningful error should be raised, possibly not resolving in the crash of the server.
## Comment 25506
Date: 2017-07-27 19:20:15 +0200
From: Dean De Leo <<dleo>>
Created attachment 563
repro.sql
> Attached file: [repro.sql](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6379_repro.sql_563) (application/sql, 129 bytes)
> Description: repro.sql
## Comment 25507
Date: 2017-07-27 19:20:35 +0200
From: Dean De Leo <<dleo>>
Created attachment 564
stacktrace.txt
> Attached file: [stacktrace.txt](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6379_stacktrace.txt_564) (text/plain, 3029 bytes)
> Description: stacktrace.txt
## Comment 25508
Date: 2017-07-27 19:23:33 +0200
From: Dean De Leo <<dleo>>
Sorry I've forgotten to mention I'm on Linux, developer build at commit:
commit 3eeb3cb1a8a9769e76be1dc82a631e8104c3f1d2
Merge: 076f8e6 fe682d4
Author: Sjoerd Mullender <sjoerd@acm.org>
Date: Wed Jul 26 14:11:12 2017 +0200
Merge with Jul2017 branch.
## Comment 25515
Date: 2017-07-28 13:10:39 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [471a91ee613d](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=471a91ee613d) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=471a91ee613d](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=471a91ee613d)
Changeset description:
fixes for bug #6379
## Comment 25516
Date: 2017-07-28 13:11:14 +0200
From: @njnes
fixed, ie protect against missing function (please add the test to the repo)
| Table UDF: SEGV raised when invoking a non existing function | https://api.github.com/repos/MonetDB/MonetDB/issues/6379/comments | 0 | 2020-11-30T15:51:28Z | 2024-06-27T13:00:29Z | https://github.com/MonetDB/MonetDB/issues/6379 | 753,592,348 | 6,379 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-27 19:09:07 +0200
From: Dean De Leo <<dleo>>
To: MonetDB5 devs <<bugs-monetdb5>>
Version: -- development
CC: @PedroTadim
Last updated: 2019-12-05 16:07:59 +0100
## Comment 25504
Date: 2017-07-27 19:09:07 +0200
From: Dean De Leo <<dleo>>
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.40 Safari/537.36
Build Identifier:
The MAL layer throws an error when invoking a table function with actual parameters both range variables and scalars.
Linux, developer build at commit:
commit 3eeb3cb1a8a9769e76be1dc82a631e8104c3f1d2
Merge: 076f8e6 fe682d4
Author: Sjoerd Mullender <sjoerd@acm.org>
Date: Wed Jul 26 14:11:12 2017 +0200
Merge with Jul2017 branch.
Reproducible: Always
### Steps to Reproduce:
1. Create a shared library with an implementation for `test1.impl1' and link to the MAL system using a script such as:
module test1;
pattern impl1(:any... ) :bat[:int]
address TESTimpl1
comment "test table udf";
2. SQL snippet:
> CREATE TABLE table1(A INT, B INT);
> CREATE FUNCTION test1(p1 INT, p2 INT) RETURNS TABLE (ret1 INT) EXTERNAL NAME test1.impl1;
> SELECT * FROM table1, LATERAL test1(table1.A, (SELECT table1.B FROM table1));
TypeException:user.Lest2[19]:'A0' may not be used before being initialized
TypeException:user.s20_1[27]:'user.Lest2' undefined in: user.Lest2(X_135:int, X_137:bte, X_139:oid);
Program contains errors.:(NONE).multiplex
> SELECT * FROM table1, LATERAL test1(table1.A, 2);
TypeException:user.Lest2[19]:'A0' may not be used before being initialized
TypeException:user.s20_1[27]:'user.Lest2' undefined in: user.Lest2(X_135:int, X_137:bte, X_139:oid);
Program contains errors.:(NONE).multiplex
### Actual Results:
An error is raised at the MAL layer.
### Expected Results:
The interpreter should invoke the function TESTimpl1 passing as arguments the column table1.A and the given scalar expression.
## Comment 27445
Date: 2019-12-05 16:07:59 +0100
From: @PedroTadim
I couldn't reproduce this, so it got fixed.
| Table UDF: cannot mix range variables and scalar variables | https://api.github.com/repos/MonetDB/MonetDB/issues/6378/comments | 0 | 2020-11-30T15:51:24Z | 2024-06-27T13:00:28Z | https://github.com/MonetDB/MonetDB/issues/6378 | 753,592,309 | 6,378 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-23 10:05:33 +0200
From: @drstmane
To: Documentation maintainers <<bugs-docs>>
Version: unspecified
CC: @yzchang
Last updated: 2017-07-31 09:42:30 +0200
## Comment 25496
Date: 2017-07-23 10:05:33 +0200
From: @drstmane
It might be my lack of RSS feed knowledge or my (wrong?) use of (buggy?) software (Chrome, Chromium and Firefox on both Fedora 24 and Ubuntu 14.04 LTS),
but when I click on "more" in the "Recent blog posts" panel on the right-hand side of the MonetDB website,
I do not get any more blog posts than the 10 most recent listed in the panel.
Is this a bug in the website or a bug "between keyboard and chair" on my side?
Thanks!
Stefan
## Comment 25497
Date: 2017-07-24 11:07:26 +0200
From: @yzchang
Hai Stefan,
Thanks for reporting this. The URL linked with "more" is wrong. It shouldn't be linked to the RSS page, but to the page that gives you an overview of all blog posts. Probably, the error has slipped in during the upgrade to Drupal 8.
I looked around a bit but gave up. I can't find the place where I can fix the referred URL.
Just for now, if you're looking for all blogs, the correct URL is: https://www.monetdb.org/blog
I don't know how to access the older messages in the RSS page.
Regards,
Jennie
## Comment 25528
Date: 2017-07-31 09:42:30 +0200
From: @sjoerdmullender
This has been fixed.
| MonetDB.org Blogs RSS feed truncated? | https://api.github.com/repos/MonetDB/MonetDB/issues/6377/comments | 0 | 2020-11-30T15:51:21Z | 2024-06-27T13:00:27Z | https://github.com/MonetDB/MonetDB/issues/6377 | 753,592,260 | 6,377 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-23 09:59:08 +0200
From: @drstmane
To: Documentation maintainers <<bugs-docs>>
Version: unspecified
Last updated: 2017-07-23 09:59:08 +0200
## Comment 25495
Date: 2017-07-23 09:59:08 +0200
From: @drstmane
Now since most (if not all) non-C client packages (Java, Perl, PHP, Ruby, Python) have been factored-out into their own repositories and are provided as their own language-specific packages,
it woudl be great if there was also some documentation how to install these packages, preferably also as non system administrator ("root" user) in user-space.
The MonetDB "Software repository" / "Download Area" (https://www.monetdb.org/downloads/) provides such information for the OS-specific packages --- though administor rights are usually required --- but not for the various language clients.
Being able to install these various language clients easily --- even when not familiar with the respective language packaging systems --- would be a great aid, e.g., when developing MonetDB core C code, e.g., in order to run a complete Mtest.
E.g., for Python it could be as simple as
pip install --user https://www.monetdb.org/downloads/Python/pymonetdb-1.0.6.tar.gz
Are there similarly simple commands for the other languages (Perl, PHP, Ruby (Java?))?
Also, it would be great if the respective source repositories came with sufficient but concise documentation --- e.g., a top-level "README" or "HOWTO" --- how to build them and install them for use with a self-compiled MonetDB.
THANKS!
Stefan
| How to install (non-C) MonetDB client packages | https://api.github.com/repos/MonetDB/MonetDB/issues/6376/comments | 0 | 2020-11-30T15:51:18Z | 2024-06-28T13:39:39Z | https://github.com/MonetDB/MonetDB/issues/6376 | 753,592,222 | 6,376 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-17 15:35:14 +0200
From: @kutsurak
To: MonetDB5 devs <<bugs-monetdb5>>
Version: 11.25.23 (Dec2016-SP5)
Last updated: 2017-08-02 10:21:20 +0200
## Comment 25493
Date: 2017-07-17 15:35:14 +0200
From: @kutsurak
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:54.0) Gecko/20100101 Firefox/54.0
Build Identifier:
When the MAL profiler creates the JSON objects, if the textual representation is larger than 8192 bytes, the output gets truncated. This creates incomplete JSON objects that create problems for profiler consumers.
Reproducible: Always
### Steps to Reproduce:
1. mserver5 --dbpath=/tmp/foo
2. stethoscope -d foo -j
3. python -c "print(\"select '\" + 'a'*8192 + \"';\")" | mclient -d foo
### Actual Results:
The JSON object with PC=1 is incomplete.
### Expected Results:
The output should be a stream of complete JSON objects.
## Comment 25494
Date: 2017-07-21 18:25:52 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [7ba8ed6d4865](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=7ba8ed6d4865) made by Panagiotis Koutsourakis <kutsurak@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7ba8ed6d4865](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7ba8ed6d4865)
Changeset description:
Add support for long SQL texts in the profiler
This is related to Bug #6375, but is still incomplete.
The problem is that the profiler truncates JSON objects to 8KB. There are two
solutions to this problem:
1. Since the location of the very big parts of the MAL plan are
known (statement text, short statement text, and value in argument list) we
can truncate only these parts, and thus produce a complete JSON object. The
problem with this approach is that we do not know in advance how much we need
to truncate, because other elements are to follow and everything needs to fit
in 8KB.
2. Transmit more than 8KB chunks. The obvious downside to this is that we will
need more IO and since the query itself can in principle be unbounded, this
can be very expensive.
The solution is the combination of the above. We truncate the big parts and so
effectively we make the amount of IO bounded but also recognizing that they
contain useful information, and try to minimize the information loss.
Still TODO:
The concatenation and transmission of big objects is buggy. The problem can be
reproduced with the command:
python -c "print(\"select * from _tables where name='\" + 'a'*8192 + \"';\")" | mclient -d
## Comment 25498
Date: 2017-07-24 15:00:07 +0200
From: @kutsurak
The last missing piece is fixing the function shortStmtRendering.
For some reason it cuts off the last few chars of really big arguments, and this causes some quote characters to be dropped leading to malformed JSON objects in the end.
## Comment 25499
Date: 2017-07-26 12:47:07 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [3584cfa2625e](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=3584cfa2625e) made by Panagiotis Koutsourakis <kutsurak@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=3584cfa2625e](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=3584cfa2625e)
Changeset description:
Try to reallocate buffers when terms are too long
Part of the fix for bug #6375
## Comment 25500
Date: 2017-07-26 12:47:14 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [55a75e48e3fc](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=55a75e48e3fc) made by Panagiotis Koutsourakis <kutsurak@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=55a75e48e3fc](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=55a75e48e3fc)
Changeset description:
Error handling in case we cannot allocate strings
Part of the fix for bug #6375
## Comment 25501
Date: 2017-07-26 12:47:22 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [9c5b374c3886](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=9c5b374c3886) made by Panagiotis Koutsourakis <kutsurak@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=9c5b374c3886](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=9c5b374c3886)
Changeset description:
Make sure stethoscope outputs each received character only once
This completes the fix for bug #6375
## Comment 25533
Date: 2017-08-02 10:21:20 +0200
From: @sjoerdmullender
Fixed in Jul2017-SP1 release.
| MAL profiler truncates JSON objects larger than 8192 characters | https://api.github.com/repos/MonetDB/MonetDB/issues/6375/comments | 0 | 2020-11-30T15:51:14Z | 2024-06-27T13:00:25Z | https://github.com/MonetDB/MonetDB/issues/6375 | 753,592,166 | 6,375 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-17 14:07:00 +0200
From: Richard Hughes <<richard.monetdb>>
To: SQL devs <<bugs-sql>>
Version: 11.25.23 (Dec2016-SP5)
CC: @njnes, shawpolakcrax12
Last updated: 2019-06-07 09:47:18 +0200
## Comment 25492
Date: 2017-07-17 14:07:00 +0200
From: Richard Hughes <<richard.monetdb>>
Build: default branch [8125f999c846](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=8125f999c846) (2017-7-14)
Reproducible: always
sql>create table sub1 (i int);
operation successful (27.770ms)
sql>create table sub2 (i int);
operation successful (13.348ms)
sql>create merge table mt as select * from sub1 with no data;
operation successful (12.755ms)
sql>alter table mt add table sub1;
operation successful (5.060ms)
sql>alter table mt add table sub2;
operation successful (5.471ms)
sql>insert into sub1 values (1),(2);
2 affected rows (6.190ms)
sql>insert into sub2 values (11),(12);
2 affected rows (6.607ms)
sql>alter table sub1 set read only;
operation successful (6.666ms)
sql>alter table sub2 set read only;
operation successful (6.624ms)
sql>analyze sys.sub1;
sql>analyze sys.sub2;
sql>plan select count(*) from mt where i >= 5 and i < 100;
+------------------------------------------------------------------+
| rel |
+==================================================================+
| project ( |
| | group by ( |
| | | select ( |
| | | | table(sys.sub2) [ "mt"."i", "mt"."%TID%" NOT NULL ] COUNT |
| | | ) [ int "5" <= "mt"."i" < int "100" ] |
| | ) [ ] [ sys.count() NOT NULL as "L3"."L3" ] |
| ) [ "L3" NOT NULL as "L4"."L3" ] |
+------------------------------------------------------------------+
7 tuples (1.274ms)
sql>select count(*) from mt where i >= 5 and i < 100;
+------+
| L3 |
+======+
| 2 |
+------+
1 tuple (3.162ms)
sql>alter table sub1 set read write;
operation successful (13.473ms)
sql>insert into sub1 values (7),(8);
2 affected rows (6.402ms)
sql>alter table sub1 set read only;
operation successful (6.637ms)
sql>analyze sys.sub1;
sql>plan select count(*) from mt where i >= 5 and i < 100;
+------------------------------------------------------------------+
| rel |
+==================================================================+
| project ( |
| | group by ( |
| | | select ( |
| | | | table(sys.sub2) [ "mt"."i", "mt"."%TID%" NOT NULL ] COUNT |
| | | ) [ int "5" <= "mt"."i" < int "100" ] |
| | ) [ ] [ sys.count() NOT NULL as "L3"."L3" ] |
| ) [ "L3" NOT NULL as "L4"."L3" ] |
+------------------------------------------------------------------+
7 tuples (1.264ms)
sql>select count(*) from mt where i >= 5 and i < 100;
+------+
| L3 |
+======+
| 2 |
+------+
1 tuple (4.480ms)
sql>select T.name,S.* from sys.statistics S join sys.columns C on S.column_id=C.id join sys.tables T on T.id=C.table_id where C.name='i' and T.name like 'sub_';
+------+-----------+------+-------+----------------------------+--------+-------+--------+------+--------+--------+--------+-----------+
| name | column_id | type | width | stamp | sample | count | unique | nils | minval | maxval | sorted | revsorted |
+======+===========+======+=======+============================+========+=======+========+======+========+========+========+===========+
| sub1 | 8265 | int | 4 | 2017-07-17 11:15:28.000000 | 4 | 4 | 4 | 0 | 1 | 8 | true | false |
| sub2 | 8268 | int | 4 | 2017-07-17 11:15:28.000000 | 2 | 2 | 2 | 0 | 11 | 12 | true | false |
+------+-----------+------+-------+----------------------------+--------+-------+--------+------+--------+--------+--------+-----------+
2 tuples (66.370ms)
Notice that the final plan is wrong, resulting in the wrong count(*) immediately below. The plan should include 'sub1' and the count should be 4. The contents of sys.statistics is correct.
What's happening is that there's an in-memory cache of the statistics in sql_column::min/max. The following patch, which simply bodges out all usage of those values, confirms this by making the above test work:
diff -r [8125f999c846](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=8125f999c846) sql/storage/store.c
--- a/sql/storage/store.c Fri Jul 14 16:22:22 2017 +0200
+++ b/sql/storage/store.c Mon Jul 17 12:28:45 2017 +0100
@@ -4724,11 +4724,6 @@
sql_schema *sys = find_sql_schema(tr, "sys");
sql_table *stats = find_sql_table(sys, "statistics");
- if (col->min && col->max) {
- *min = col->min;
- *max = col->max;
- return 1;
- }
if (stats) {
sql_column *stats_column_id = find_sql_column(stats, "column_id");
oid rid = table_funcs.column_find_row(tr, stats_column_id, &col->base.id, NULL);
I'm not sure what the right fix is, however I've looked at all the uses of sql_trans_ranges() and concluded that the cache isn't used enough to be important for performance for us. Removing it entirely, therefore, might not necessarily be wrong.
## Comment 25522
Date: 2017-07-28 20:47:43 +0200
From: @njnes
added cleanup code in the sql_analyze call.
## Comment 25524
Date: 2017-07-28 22:45:49 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [6e732bac4683](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=6e732bac4683) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=6e732bac4683](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=6e732bac4683)
Changeset description:
fixed bug #6374, ie clear cached statistics on analyze call
## Comment 25530
Date: 2017-07-31 11:28:10 +0200
From: Richard Hughes <<richard.monetdb>>
This fix doesn't work, I'm afraid.
Running the original test across two connections, one doing all the writes and one read-only, still reproduces the problem:
=====connection 1======
create table sub1 (i int);
create table sub2 (i int);
create merge table mt as select * from sub1 with no data;
alter table mt add table sub1;
alter table mt add table sub2;
insert into sub1 values (1),(2);
insert into sub2 values (11),(12);
alter table sub1 set read only;
alter table sub2 set read only;
analyze sys.sub1;
analyze sys.sub2;
=====connection 2======
sql>plan select count(*) from mt where i >= 5 and i < 100;
+------------------------------------------------------------------+
| rel |
+==================================================================+
| project ( |
| | group by ( |
| | | select ( |
| | | | table(sys.sub2) [ "mt"."i", "mt"."%TID%" NOT NULL ] COUNT |
| | | ) [ int "5" <= "mt"."i" < int "100" ] |
| | ) [ ] [ sys.count() NOT NULL as "L3"."L3" ] |
| ) [ "L3" NOT NULL as "L4"."L3" ] |
+------------------------------------------------------------------+
7 tuples (1.488ms)
sql>select count(*) from mt where i >= 5 and i < 100;
+------+
| L3 |
+======+
| 2 |
+------+
1 tuple (3.514ms)
=====connection 1======
alter table sub1 set read write;
insert into sub1 values (7),(8);
alter table sub1 set read only;
analyze sys.sub1;
=====connection 2======
sql>plan select count(*) from mt where i >= 5 and i < 100;
+------------------------------------------------------------------+
| rel |
+==================================================================+
| project ( |
| | group by ( |
| | | select ( |
| | | | table(sys.sub2) [ "mt"."i", "mt"."%TID%" NOT NULL ] COUNT |
| | | ) [ int "5" <= "mt"."i" < int "100" ] |
| | ) [ ] [ sys.count() NOT NULL as "L3"."L3" ] |
| ) [ "L3" NOT NULL as "L4"."L3" ] |
+------------------------------------------------------------------+
7 tuples (1.322ms)
sql>select count(*) from mt where i >= 5 and i < 100;
+------+
| L3 |
+======+
| 2 |
+------+
1 tuple (3.568ms)
It looks to me like sql_analyze() is wiping out only its own copy of sql_column, not the one(s) belonging to other connections.
## Comment 25650
Date: 2017-09-20 08:55:50 +0200
From: @njnes
the reset_column now also clears the cache on column changes.
| Wrong answer from merge table after content changes | https://api.github.com/repos/MonetDB/MonetDB/issues/6374/comments | 0 | 2020-11-30T15:51:11Z | 2024-06-27T13:00:24Z | https://github.com/MonetDB/MonetDB/issues/6374 | 753,592,137 | 6,374 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6373/comments | 0 | 2020-11-30T15:51:09Z | 2020-11-30T15:51:10Z | https://github.com/MonetDB/MonetDB/issues/6373 | 753,592,099 | 6,373 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6372/comments | 0 | 2020-11-30T15:51:06Z | 2020-11-30T15:51:08Z | https://github.com/MonetDB/MonetDB/issues/6372 | 753,592,063 | 6,372 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6371/comments | 0 | 2020-11-30T15:51:03Z | 2020-11-30T15:51:05Z | https://github.com/MonetDB/MonetDB/issues/6371 | 753,592,034 | 6,371 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6370/comments | 0 | 2020-11-30T15:51:01Z | 2020-11-30T15:51:02Z | https://github.com/MonetDB/MonetDB/issues/6370 | 753,591,994 | 6,370 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6369/comments | 0 | 2020-11-30T15:50:57Z | 2020-11-30T15:51:00Z | https://github.com/MonetDB/MonetDB/issues/6369 | 753,591,957 | 6,369 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6368/comments | 0 | 2020-11-30T15:50:55Z | 2020-11-30T15:50:56Z | https://github.com/MonetDB/MonetDB/issues/6368 | 753,591,922 | 6,368 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6367/comments | 0 | 2020-11-30T15:50:52Z | 2020-11-30T15:50:54Z | https://github.com/MonetDB/MonetDB/issues/6367 | 753,591,887 | 6,367 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6366/comments | 0 | 2020-11-30T15:50:48Z | 2020-11-30T15:50:50Z | https://github.com/MonetDB/MonetDB/issues/6366 | 753,591,838 | 6,366 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6365/comments | 0 | 2020-11-30T15:50:46Z | 2020-11-30T15:50:47Z | https://github.com/MonetDB/MonetDB/issues/6365 | 753,591,811 | 6,365 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6364/comments | 0 | 2020-11-30T15:50:43Z | 2020-11-30T15:50:45Z | https://github.com/MonetDB/MonetDB/issues/6364 | 753,591,777 | 6,364 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6363/comments | 0 | 2020-11-30T15:50:41Z | 2020-11-30T15:50:42Z | https://github.com/MonetDB/MonetDB/issues/6363 | 753,591,722 | 6,363 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6362/comments | 0 | 2020-11-30T15:50:39Z | 2020-11-30T15:50:40Z | https://github.com/MonetDB/MonetDB/issues/6362 | 753,591,687 | 6,362 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6361/comments | 0 | 2020-11-30T15:50:37Z | 2020-11-30T15:50:38Z | https://github.com/MonetDB/MonetDB/issues/6361 | 753,591,656 | 6,361 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6360/comments | 0 | 2020-11-30T15:50:34Z | 2020-11-30T15:50:36Z | https://github.com/MonetDB/MonetDB/issues/6360 | 753,591,634 | 6,360 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6359/comments | 0 | 2020-11-30T15:50:32Z | 2020-11-30T15:50:34Z | https://github.com/MonetDB/MonetDB/issues/6359 | 753,591,596 | 6,359 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6358/comments | 0 | 2020-11-30T15:50:29Z | 2020-11-30T15:50:31Z | https://github.com/MonetDB/MonetDB/issues/6358 | 753,591,536 | 6,358 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6357/comments | 0 | 2020-11-30T15:50:27Z | 2020-11-30T15:50:28Z | https://github.com/MonetDB/MonetDB/issues/6357 | 753,591,501 | 6,357 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6356/comments | 0 | 2020-11-30T15:50:23Z | 2020-11-30T15:50:26Z | https://github.com/MonetDB/MonetDB/issues/6356 | 753,591,451 | 6,356 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6355/comments | 0 | 2020-11-30T15:50:20Z | 2020-11-30T15:50:22Z | https://github.com/MonetDB/MonetDB/issues/6355 | 753,591,409 | 6,355 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T15:44:36Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/6354/comments | 0 | 2020-11-30T15:50:18Z | 2020-11-30T15:50:19Z | https://github.com/MonetDB/MonetDB/issues/6354 | 753,591,375 | 6,354 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-14 14:23:25 +0200
From: @bartscheers
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-10-26 14:01:35 +0200
## Comment 25466
Date: 2017-07-14 14:23:25 +0200
From: @bartscheers
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:54.0) Gecko/20100101 Firefox/54.0
Build Identifier:
Statement is "insert into tab1 (col1, col2) select col1, col2 from tab2". Here tab2 is defined as a remote table and has a third column col3 which is nullable. If not explicitly written in the statement the query cannot be registered.
Reproducible: Always
### Steps to Reproduce:
1. run monetdb on master and worker node
2. execute the run.sh (first part on worker, then second part on master)
3.
### Actual Results:
(mapi:monetdb://monetdb@rocks078/worker) Cannot register
project (
table(sys.tmpextrsrc) [ "tmpextrsrc"."id" NOT NULL, "tmpextrsrc"."total_flux" NOT NULL, "tmpextrsrc"."x" NOT NULL, "tmpextrsrc"."y" NOT NULL, "tmpextrsrc"."z" NOT NULL, "tmpextrsrc"."%TID%" NOT NULL ] COUNT
) [ "tmpextrsrc"."id" NOT NULL, "tmpextrsrc"."total_flux" NOT NULL, double "NULL" as "L7"."L7", "tmpextrsrc"."x" NOT NULL, "tmpextrsrc"."y" NOT NULL, "tmpextrsrc"."z" NOT NULL ] REMOTE mapi:monetdb://rocks078:50078/worker
### Expected Results:
operation successful
2 affected rows
This was tested on the Jul2017 branch (version d0720566e7f8)
## Comment 25467
Date: 2017-07-14 14:24:27 +0200
From: @bartscheers
Created attachment 558
init db
> Attached file: [init_db.sh](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6353_init_db.sh_558) (application/x-shellscript, 243 bytes)
> Description: init db
## Comment 25468
Date: 2017-07-14 14:27:03 +0200
From: @bartscheers
Created attachment 559
worker stuff
> Attached file: [worker.sql](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6353_worker.sql_559) (application/sql, 327 bytes)
> Description: worker stuff
## Comment 25469
Date: 2017-07-14 14:27:30 +0200
From: @bartscheers
Created attachment 560
master stuff
> Attached file: [master.sql](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6353_master.sql_560) (application/sql, 702 bytes)
> Description: master stuff
## Comment 25470
Date: 2017-07-14 14:27:47 +0200
From: @bartscheers
Created attachment 561
master buggy
> Attached file: [master_bug.sql](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6353_master_bug.sql_561) (application/sql, 678 bytes)
> Description: master buggy
## Comment 25471
Date: 2017-07-14 14:28:53 +0200
From: @bartscheers
Created attachment 562
run script
> Attached file: [run.sh](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6353_run.sh_562) (application/x-shellscript, 186 bytes)
> Description: run script
## Comment 25512
Date: 2017-07-28 13:01:25 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [d876a2cc5823](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=d876a2cc5823) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=d876a2cc5823](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=d876a2cc5823)
Changeset description:
fixes for bug #6353, ie handle NULL in rel_read better
## Comment 25513
Date: 2017-07-28 13:01:50 +0200
From: @njnes
fixed, ie handle the NULL when parsing atoms in the rel_read code
| implicit NULL value not propagated in distributed/remote query | https://api.github.com/repos/MonetDB/MonetDB/issues/6353/comments | 0 | 2020-11-30T15:50:14Z | 2024-06-27T13:00:17Z | https://github.com/MonetDB/MonetDB/issues/6353 | 753,591,317 | 6,353 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-14 09:47:14 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-10-26 14:01:25 +0200
## Comment 25465
Date: 2017-07-14 09:47:14 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:54.0) Gecko/20100101 Firefox/54.0
Build Identifier:
On TPCH SF1 Sqlsmith discovered a name resolution issue causing an assertion to be raised in the optimizers
could not find sample_3.majflt
sample_0.rowid
sample_0.fldid
sample_0.message
sample_0.input
Reproducible: Always
### Steps to Reproduce:
insert into sys.netcdf_files values (
50,
case when 84 is NULL then case when EXISTS (
select
sample_3.majflt as c0
from
sys.rejects as sample_0
inner join sys.tracelog as sample_3
left join sys.querylog_calls as sample_4
on (sample_3.rrsmb = sample_4.tuples )
on (sample_0.rowid = sample_3.ticks )
where false)
then cast(null as char) else cast(null as char) end
else case when EXISTS (
select
sample_3.majflt as c0
from
sys.rejects as sample_0
inner join sys.tracelog as sample_3
left join sys.querylog_calls as sample_4
on (sample_3.rrsmb = sample_4.tuples )
on (sample_0.rowid = sample_3.ticks )
where false) then cast(null as char) else cast(null as char) end
end
);
## Comment 25519
Date: 2017-07-28 20:25:57 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [e9c19dd387e7](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=e9c19dd387e7) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=e9c19dd387e7](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=e9c19dd387e7)
Changeset description:
fixed bug #6348, only rename non constant expressions
fixed bug #6352, when pushing projects up, we should include subqueries.
## Comment 25521
Date: 2017-07-28 20:26:18 +0200
From: @njnes
fixed, include subqueries when pushing up projections
| Scope resolution problem (sqlsmith) | https://api.github.com/repos/MonetDB/MonetDB/issues/6352/comments | 0 | 2020-11-30T15:50:11Z | 2024-06-27T13:00:16Z | https://github.com/MonetDB/MonetDB/issues/6352 | 753,591,281 | 6,352 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-12 11:01:02 +0200
From: @swingbit
To: SQL devs <<bugs-sql>>
Version: 11.25.23 (Dec2016-SP5)
CC: martin.van.dinther, @mlkersten, @njnes, @PedroTadim
Last updated: 2020-06-16 09:55:07 +0200
## Comment 25453
Date: 2017-07-12 11:01:02 +0200
From: @swingbit
User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36
Build Identifier:
The following behaviour seems not right to me:
sql>prepare select 1;
execute prepared statement using: EXEC 7(...)
+---------+--------+-------+--------+-------+--------+
| type | digits | scale | schema | table | column |
+=========+========+=======+========+=======+========+
| tinyint | 1 | 0 | | L2 | L2 |
+---------+--------+-------+--------+-------+--------+
1 tuple (0.560ms)
sql>exec 7();
+------+
| L2 |
+======+
| 1 |
+------+
1 tuple (0.301ms)
sql>exec 7();
+------+
| L2 |
+======+
| 1 |
+------+
1 tuple (0.332ms)
sql>some error;
syntax error, unexpected SOME in: "some"
sql>exec 7();
EXEC: no prepared statement with id: 7
sql>
The error discarded the prepared statement.
I see no reason for that, especially in auto-commit mode.
Reproducible: Always
$ mserver5 --version
MonetDB 5 server v11.25.24 (64-bit, 128-bit integers)
This is an unreleased version
Copyright (c) 1993-July 2008 CWI
Copyright (c) August 2008-2017 MonetDB B.V., all rights reserved
Visit http://www.monetdb.org/ for further information
Found 15.6GiB available memory, 8 available cpu cores
Libraries:
libpcre: 8.40 2017-01-11 (compiled with 8.40)
openssl: OpenSSL 1.0.2j 26 Sep 2016 (compiled with OpenSSL 1.0.2j-fips 26 Sep 2016)
libxml2: 2.9.3 (compiled with 2.9.3)
Compiled by: roberto@photon.hq.spinque.com (x86_64-unknown-linux-gnu)
Compilation: gcc -O3 -fomit-frame-pointer -pipe -Werror -Wall -Wextra -W -Werror-implicit-function-declaration -Wpointer-arith -Wdeclaration-after-statement -Wundef -Wformat=2 -Wno-format-nonliteral -Winit-self -Winvalid-pch -Wmissing-declarations -Wmissing-format-attribute -Wmissing-prototypes -Wold-style-definition -Wpacked -Wunknown-pragmas -Wvariadic-macros -fstack-protector-all -Wstack-protector -Wpacked-bitfield-compat -Wsync-nand -Wjump-misses-init -Wmissing-include-dirs -Wlogical-op -Wunreachable-code -D_FORTIFY_SOURCE=2
Linking : /usr/bin/ld -m elf_x86_64
## Comment 25514
Date: 2017-07-28 13:05:06 +0200
From: @njnes
on errors we cleanup, ie we also cleanup any prepared statements. We basically reset the connection to a known previous state.
## Comment 25517
Date: 2017-07-28 13:17:23 +0200
From: @swingbit
Thanks Niels.
Just to be sure, you don't think this behaviour is odd from a user point of view?
I do understand the reason behind it, but it still surprises me much.
## Comment 25605
Date: 2017-08-31 19:00:16 +0200
From: Martin van Dinther <<martin.van.dinther>>
MonetDB should not destroy the prepared statement when an error occurs, for exmaple in the calling of exec (<args>) where <args> contains some invalid value. It produces unexpected and annoying problems in mclient but also in JDBC and ODBC drivers.
Just return the error msg and keep the prepared statement(s) active.
## Comment 25636
Date: 2017-09-15 09:01:09 +0200
From: @njnes
again each error is the moment to cleanup, which is similar to a session reset.
I do not think this is very strange. Users should handle errors properly.
## Comment 27106
Date: 2019-07-04 11:54:50 +0200
From: @swingbit
I must reopen this bug, because I just don't agree with how it was closed.
1) cleaning up (resetting the connection) after any error sounds like an arbitrary and questionable choice, not like a proper reason
2) it is not even true/consistently done, because variables are still there after an error:
auto commit mode: on
sql>declare a int;
operation successful
sql>set a = 2;
operation successful
sql>select a;
+------+
| L2 |
+======+
| 2 |
+------+
1 tuple
sql>some error;
syntax error, unexpected IDENT in: "some"
sql>select a;
+------+
| L2 |
+======+
| 2 |
+------+
1 tuple
## Comment 27492
Date: 2019-12-18 16:32:43 +0100
From: MonetDB Mercurial Repository <<hg>>
Changeset [cfcedd05281b](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=cfcedd05281b) made by Pedro Ferreira <pedro.ferreira@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=cfcedd05281b](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=cfcedd05281b)
Changeset description:
Delete session declared variables at an error, rollback or commit.
I'm against this too, but as requested on bug #6351, we have to keep it consistent with the deletion of prepared statements.
## Comment 27596
Date: 2020-03-15 11:10:55 +0100
From: @mlkersten
Follow up in BUG #6832
## Comment 27856
Date: 2020-06-16 09:55:07 +0200
From: @PedroTadim
Discussing continues on bug #6832.
| Prepared statements lost after any error, also in auto commit mode | https://api.github.com/repos/MonetDB/MonetDB/issues/6351/comments | 1 | 2020-11-30T15:50:08Z | 2024-06-28T13:10:52Z | https://github.com/MonetDB/MonetDB/issues/6351 | 753,591,229 | 6,351 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-11 16:45:10 +0200
From: Leon Helwerda <<l.s.helwerda>>
To: Martin van Dinther <<martin.van.dinther>>
Version: 11.25.15 (Dec2016-SP3)
CC: martin.van.dinther
Last updated: 2017-10-26 14:01:29 +0200
## Comment 25451
Date: 2017-07-11 16:45:10 +0200
From: Leon Helwerda <<l.s.helwerda>>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:54.0) Gecko/20100101 Firefox/54.0
Build Identifier:
Certain control characters, such as the carriage return ('\r') and form feed ('\f') are returned as the character that they are represented as, but even without the backslash ('r' and 'f', respectively) rather than as an actual control character. This makes such special characters indistinguishable from unrelated normal characters. Other characters like '\n' do work as expected.
Reproducible: Always
### Steps to Reproduce:
$ java -jar ~/Downloads/jdbcclient.jar -d db
password:
Welcome to the MonetDB interactive JDBC terminal!
Database Server: MonetDB v11.25.15
JDBC Driver: MonetDB Native Driver v2.24 (Liberica 20161215 based on MCL v1.14)
Current Schema: sys
Type \q to quit, \h for a list of available commands
auto commit mode: on
sql> CREATE TABLE test (message TEXT);
Operation successful. Elapsed Time: 39 ms
sql> INSERT INTO test VALUES ('hello\r\nworld');
1 affected row. Elapsed Time: 26 ms
sql> INSERT INTO test VALUES ('hello\fworld');
1 affected row. Elapsed Time: 23 ms
### Actual Results:
$ java -jar ~/Downloads/jdbcclient.jar
sql> SELECT * FROM test;
+--------------+
| message |
+==============+
| hellor
world |
| hellofworld |
+--------------+
2 rows
Elapsed Time: 24 ms
The 'r' and 'f' in these results are normal characters also visible in ResultSets, while the newline is (correctly) presented in the result.
### Expected Results:
The result from mclient on the same table as create above (regardless of whether the rows were created with jdbcclient or mclient) shows a more expected result:
$ mclient -d db
password:
Welcome to mclient, the MonetDB/SQL interactive terminal (Dec2016-SP4)
Database: MonetDB v11.25.15 (Dec2016-SP3), 'mapi:monetdb://localhost:50000/db'
Type \q to quit, \? for a list of available commands
auto commit mode: on
sql>SELECT * FROM test;
+--------------+
| message |
+==============+
| hello |
: world :
| hello
world |
+--------------+
Obviously, in Java applications, the strings are supposed contain the appropriate control characters instead of the fancy output as shown here.
## Comment 25673
Date: 2017-09-28 15:05:32 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset e5c67a23d7d6, made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the monetdb-java repo, refers to this bug.
For complete details, see https://dev.monetdb.org/hg/monetdb-java?cmd=changeset;node=e5c67a23d7d6
Changeset description:
Fix for bug #6350
## Comment 25674
Date: 2017-09-28 16:56:26 +0200
From: Martin van Dinther <<martin.van.dinther>>
Fix will be included in the next release of jdbcclient.jar and JDBC driver (v2.27)
| Carriage return and form feed in TEXT fields are returned as 'r' and 'f' in jdbcclient and ResultSets | https://api.github.com/repos/MonetDB/MonetDB/issues/6350/comments | 0 | 2020-11-30T15:50:05Z | 2024-06-27T13:00:14Z | https://github.com/MonetDB/MonetDB/issues/6350 | 753,591,181 | 6,350 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-11 13:31:57 +0200
From: Manuel <<manuel>>
To: Martin van Dinther <<martin.van.dinther>>
Version: 11.25.23 (Dec2016-SP5)
CC: martin.van.dinther
Last updated: 2017-07-28 09:53:35 +0200
## Comment 25450
Date: 2017-07-11 13:31:57 +0200
From: Manuel <<manuel>>
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36
Build Identifier:
I am using MonetPreparedStatement to build an sql. I am unable to set the string value in a where condition for a column of type CLOB.
An SQLException with message Conversion not allowed, and sqlState = "M1M05".
My guess is that the method
public void setObject(int parameterIndex, Object x, int targetSqlType, int scale) throws SQLException;
in MonetPreparedStatement, is missing the case 2005 (CLOB).
Reproducible: Always
### Steps to Reproduce:
1. have a table with a column if type CLOB
2. use the jdbc driver to open a connection to a database, and build a precompiled SQL statement using PreparedStatement
3.set the value strign value to a CLOB argument.
### Actual Results:
An sql exception is thrown
### Expected Results:
the query gets built
## Comment 25461
Date: 2017-07-13 15:35:15 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset 0112d8496636, made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the monetdb-java repo, refers to this bug.
For complete details, see https://dev.monetdb.org/hg/monetdb-java?cmd=changeset;node=0112d8496636
Changeset description:
Added test program for bug #6349.
Added case Types.CLOB: at several places in setObject(). This fixes bug #6349
Also replaced "" + x into direct <Class>.toString(x)
## Comment 25462
Date: 2017-07-13 15:40:36 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [7868f3b4b78d](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=7868f3b4b78d) made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7868f3b4b78d](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7868f3b4b78d)
Changeset description:
Activate new java JDBC test program for bug #6349
## Comment 25510
Date: 2017-07-28 09:53:35 +0200
From: @sjoerdmullender
A new JDBC driver (2.26) has been released.
| Cannot set String value to a CLOB column in MonetPreparedStatement | https://api.github.com/repos/MonetDB/MonetDB/issues/6349/comments | 0 | 2020-11-30T15:50:01Z | 2024-06-27T13:00:13Z | https://github.com/MonetDB/MonetDB/issues/6349 | 753,591,138 | 6,349 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-08 22:56:14 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-10-26 14:01:43 +0200
## Comment 25440
Date: 2017-07-08 22:56:14 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:54.0) Gecko/20100101 Firefox/54.0
Build Identifier:
The sample test causes monetdb default branch to crash in the following
A test file has been added
hread 6 "mserver5" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7f302cd56700 (LWP 10293)]
0x00007f304e456433 in count_col (tr=0x7f301c001b00, c=0x7f301c1369c0, all=1) at /export/scratch1/mk/default//package/sql/storage/bat/bat_storage.c:961
961 c->data = timestamp_delta(oc->data, tr->stime);
(gdb) where
0 0x00007f304e456433 in count_col (tr=0x7f301c001b00, c=0x7f301c1369c0, all=1) at /export/scratch1/mk/default//package/sql/storage/bat/bat_storage.c:961
1 0x00007f304e3320a5 in stmt_tid (be=0x7f301c0db0f0, t=0x7f301c136860, partition=1) at /export/scratch1/mk/default//package/sql/backends/monet5/sql_statement.c:476
2 0x00007f304e316a60 in rel2bin_basetable (be=0x7f301c0db0f0, rel=0x7f301c134550) at /export/scratch1/mk/default//package/sql/backends/monet5/rel_bin.c:1171
3 0x00007f304e326f59 in subrel_bin (be=0x7f301c0db0f0, rel=0x7f301c134550, refs=0x7f301c134fc0) at /export/scratch1/mk/default//package/sql/backends/monet5/rel_bin.c:4755
4 0x00007f304e31c975 in rel2bin_project (be=0x7f301c0db0f0, rel=0x7f301c134930, refs=0x7f301c134fc0, topn=0x0) at /export/scratch1/mk/default//package/sql/backends/monet5/rel_bin.c:2418
5 0x00007f304e3270bc in subrel_bin (be=0x7f301c0db0f0, rel=0x7f301c134930, refs=0x7f301c134fc0) at /export/scratch1/mk/default//package/sql/backends/monet5/rel_bin.c:4789
6 0x00007f304e3273b3 in output_rel_bin (be=0x7f301c0db0f0, rel=0x7f301c134930) at /export/scratch1/mk/default//package/sql/backends/monet5/rel_bin.c:4854
7 0x00007f304e341031 in sql_relation2stmt (be=0x7f301c0db0f0, r=0x7f301c134930) at /export/scratch1/mk/default//package/sql/backends/monet5/sql_gencode.c:474
8 0x00007f304e341390 in backend_dumpstmt (be=0x7f301c0db0f0, mb=0x7f301c107fe0, r=0x7f301c134930, top=1, add_end=1, query=0x7f301c001a30 "select * from cls_0_statistics;")
at /export/scratch1/mk/default//package/sql/backends/monet5/sql_gencode.c:515
9 0x00007f304e341bf5 in backend_dumpproc (be=0x7f301c0db0f0, c=0x7f304fbf6348, cq=0x7f301c118010, r=0x7f301c134930) at /export/scratch1/mk/default//package/sql/backends/monet5/sql_gencode.c:642
10 0x00007f304e309d30 in SQLparser (c=0x7f304fbf6348) at /export/scratch1/mk/default//package/sql/backends/monet5/sql_scenario.c:1145
Reproducible: Always
### Steps to Reproduce:
create table cls_0(i integer, j integer, k string);
insert into cls_0 values
( 1, 99, 'b'),
( 2, 99, 'b'),
( 3, 99, 'b'),
( 4, 99, 'a'),
( 5, 99, 'a'),
( 6, 99, 'a'),
( 7, 33, 'b'),
( 8, 33, 'b'),
( 9, 33, 'b'),
( 10, 33, 'a'),
( 11, 33, 'a'),
( 12, 33, 'a');
select * from cls_0;
-- The metric functions over all columns are derived in one blow.
-- The result can be cached for easy recall
create procedure cls_0_statistics()
begin
create table cls_0_statistics ( col string, fcn string, tpe string, val string);
insert into cls_0_statistics values
('i', 'min', 'int', (select cast( min( i ) as string) from cls_0)),
('i', 'max', 'int', (select cast( min( i ) as string) from cls_0)),
('j', 'min', 'int', (select cast( min( i ) as string) from cls_0));
end;
select * from cls_0_statistics;
## Comment 25454
Date: 2017-07-12 11:22:55 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [11c1724677ad](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=11c1724677ad) made by Martin Kersten <mk@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=11c1724677ad](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=11c1724677ad)
Changeset description:
Added file for bug #6348
## Comment 25518
Date: 2017-07-28 20:24:31 +0200
From: @njnes
fixed, problem was in the multiple rows of insert values
## Comment 25520
Date: 2017-07-28 20:25:59 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [e9c19dd387e7](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=e9c19dd387e7) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=e9c19dd387e7](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=e9c19dd387e7)
Changeset description:
fixed bug #6348, only rename non constant expressions
fixed bug #6352, when pushing projects up, we should include subqueries.
| Interference of procedure/table name | https://api.github.com/repos/MonetDB/MonetDB/issues/6348/comments | 0 | 2020-11-30T15:49:58Z | 2024-06-27T13:00:12Z | https://github.com/MonetDB/MonetDB/issues/6348 | 753,591,096 | 6,348 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-08 16:32:27 +0200
From: Sreejith Sharma <<sreejith.sharma>>
To: MonetDB5 devs <<bugs-monetdb5>>
Version: 11.25.3 (Dec2016)
CC: @mlkersten, sreejith.sharma
Last updated: 2019-04-30 12:36:04 +0200
## Comment 25437
Date: 2017-07-08 16:32:27 +0200
From: Sreejith Sharma <<sreejith.sharma>>
User-Agent: Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.103 Safari/537.36
Build Identifier:
If any SQL size if more than 1.2MB it fails with "memory exhausted in" error in the server.
Reproducible: Always
### Steps to Reproduce:
1.Execute any SQL with more than 1.2MB in size will fail in any scenario.
2.
3.
### Actual Results:
SQL fails with "memory exhausted in" error
### Expected Results:
SQL should have gone through and given the expected results. Any query less than 1.2MB only will go through at this moment.
## Comment 25438
Date: 2017-07-08 16:43:25 +0200
From: Sreejith Sharma <<sreejith.sharma>>
Please let me know if there is any alternative or any limitations on the SQL query size.
## Comment 25439
Date: 2017-07-08 20:41:08 +0200
From: @mlkersten
Assuming you mean a single SQL statement of more then 1.2M in size, this sounds like a run away front-end query generator. Could also be containing a very large cblob value. More details on the actual query you sent is in order to make any suggestion.
Furthermore, could also be a limitation in your windows environment.
## Comment 25441
Date: 2017-07-09 04:29:24 +0200
From: Sreejith Sharma <<sreejith.sharma>>
(In reply to Martin Kersten from comment 2)
> Assuming you mean a single SQL statement of more then 1.2M in size, this
> sounds like a run away front-end query generator. Could also be containing a
> very large cblob value. More details on the actual query you sent is in
> order to make any suggestion.
>
> Furthermore, could also be a limitation in your windows environment.
Thanks Martin.
The SQL is generated from front end and it's not a clob value. It's just a combination of OR/AND condition. Also, it's running on Linux platform with enough memory. Is there any limitation as such for a single SQL size?
## Comment 25442
Date: 2017-07-09 08:19:37 +0200
From: Sreejith Sharma <<sreejith.sharma>>
Created attachment 557
Attached is the zipped version of SQL which exceed 2.5mb and fails in server.
As SQL generated by the application is more of user-driven selections, we don't really have much control on the size. So, I would expect the SQL might group even more than 3MB.
> Attached file: [sql_2500mb_plus.zip](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6347_sql_2500mb_plus.zip_557) (application/x-zip-compressed, 160380 bytes)
> Description: Attached is the zipped version of SQL which exceed 2.5mb and fails in server.
## Comment 25443
Date: 2017-07-09 09:53:45 +0200
From: @mlkersten
It seems your zip file contains an error, because the predicate uses c_5 twice. This leads to an empty result. Perhaps the file should be regenerated with c_6
AND ("TEST_MONETDB".c_5 = 26450)
AND ("TEST_MONETDB".c_5 = 1109937)
## Comment 25444
Date: 2017-07-09 16:50:23 +0200
From: Sreejith Sharma <<sreejith.sharma>>
(In reply to Martin Kersten from comment 5)
> It seems your zip file contains an error, because the predicate uses c_5
> twice. This leads to an empty result. Perhaps the file should be regenerated
> with c_6
>
> AND ("TEST_MONETDB".c_5 = 26450)
> AND ("TEST_MONETDB".c_5 = 1109937)
Apologisze, my bad... however even having it fixed "AND ("TEST_MONETDB".c_6=1109937) across SQL it will fail with below error if the size exceeds 1.2MB.
-------------------------------
memory exhausted in: "select sum(c1)
from "TEST_MONETDB"
where (
("TES"
syntax error, unexpected IDENT in: "c_6"
--------------------------------
if it's within 1.2MB limit, I will get below results.
+--------+
| L3 |
+========+
| 488835 |
+--------+
1 tuple
Is there any limitation as such for SQL size?
## Comment 25445
Date: 2017-07-09 17:17:09 +0200
From: Sreejith Sharma <<sreejith.sharma>>
(In reply to Sreejith Sharma from comment 6)
> (In reply to Martin Kersten from comment 5)
> > It seems your zip file contains an error, because the predicate uses c_5
> > twice. This leads to an empty result. Perhaps the file should be regenerated
> > with c_6
> >
> > AND ("TEST_MONETDB".c_5 = 26450)
> > AND ("TEST_MONETDB".c_5 = 1109937)
>
> Apologize, my bad... however even having it fixed "AND
> ("TEST_MONETDB".c_6=1109937) across SQL it will fail with below error if the
> size exceeds 2.5MB.
>
> -------------------------------
> memory exhausted in: "select sum(c1)
> from "TEST_MONETDB"
> where (
> ("TES"
> syntax error, unexpected IDENT in: "c_6"
> --------------------------------
>
> if it's within 2.5MB limit, I will get below results.
>
> +--------+
> | L3 |
> +========+
> | 488835 |
> +--------+
> 1 tuple
>
> Also, one more correction as made the changes and raised the size it's > > >going through till 2.5.
>
> Is there any limitation as such for SQL size?
> Correcting the subject line ---
## Comment 25446
Date: 2017-07-09 17:59:53 +0200
From: Sreejith Sharma <<sreejith.sharma>>
(In reply to Sreejith Sharma from comment 7)
> (In reply to Sreejith Sharma from comment 6)
> > (In reply to Martin Kersten from comment 5)
> > > It seems your zip file contains an error, because the predicate uses c_5
> > > twice. This leads to an empty result. Perhaps the file should be regenerated
> > > with c_6
> > >
> > > AND ("TEST_MONETDB".c_5 = 26450)
> > > AND ("TEST_MONETDB".c_5 = 1109937)
> >
> > Apologize, my bad... however even having it fixed "AND
> > ("TEST_MONETDB".c_6=1109937) across SQL it will fail with below error if the
> > size exceeds 2.5MB.
> >
> > -------------------------------
> > memory exhausted in: "select sum(c1)
> > from "TEST_MONETDB"
> > where (
> > ("TES"
> > syntax error, unexpected IDENT in: "c_6"
> > --------------------------------
> >
> > if it's within 2.5MB limit, I will get below results.
> >
> > +--------+
> > | L3 |
> > +========+
> > | 488835 |
> > +--------+
> > 1 tuple
> >
> > Also, one more correction as made the changes and raised the size it's > > >going through till 2.5.
> >
> > Is there any limitation as such for SQL size?
> > Correcting the subject line ---
Also, is there any way where we can increase the memory limit? As I noticed the limit varies from 2.5 to 2.7.
## Comment 25447
Date: 2017-07-10 12:53:42 +0200
From: @sjoerdmullender
The problem is that the BISON parser runs out of stack size. This stack is an array internal to the parser that is 10000 entries big. This means the problem has nothing to do with the size of the query per se. Instead it has to do with the complexity of the query.
Given that it the bison parser that complains, the only solution for being able to parse the given query would be to increase the stack size (YYMAXDEPTH) and recompile.
An alternative would be to radically change the query. I'm thinking of instead of having this enormous list of OR-separated clauses, do a JOIN with a table that you specify in-line.
Something like
FROM "TEST_MONETDB", (VALUES (271459, 1150308, ...), (...)) AS t(c_1, c_2, ...);
WHERE "TEST_MONETDB".c_1 = t.c_1 and "TEST_MONETDB".c_2 = t.c_t ...
## Comment 25448
Date: 2017-07-11 10:35:19 +0200
From: Sreejith Sharma <<sreejith.sharma>>
(In reply to Sjoerd Mullender from comment 9)
> The problem is that the BISON parser runs out of stack size. This stack is
> an array internal to the parser that is 10000 entries big. This means the
> problem has nothing to do with the size of the query per se. Instead it has
> to do with the complexity of the query.
>
> Given that it the bison parser that complains, the only solution for being
> able to parse the given query would be to increase the stack size
> (YYMAXDEPTH) and recompile.
>
> An alternative would be to radically change the query. I'm thinking of
> instead of having this enormous list of OR-separated clauses, do a JOIN with
> a table that you specify in-line.
>
> Something like
>
> FROM "TEST_MONETDB", (VALUES (271459, 1150308, ...), (...)) AS t(c_1, c_2,
> ...);
> WHERE "TEST_MONETDB".c_1 = t.c_1 and "TEST_MONETDB".c_2 = t.c_t ...
Thanks Sjoerd. Resizing of default parameter for YYMAXDEPTH from 10000 has worked! However, there's limitation of using conjunction/disjunction OR/AND clause while usage/parsing the SQL.
## Comment 25449
Date: 2017-07-11 13:22:07 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [fff18316fff8](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=fff18316fff8) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=fff18316fff8](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=fff18316fff8)
Changeset description:
Use left recursion instead of right recursion in parser.
This fixes bug #6347, at least as far as parsing the query is
concerned. It may still happen that there is a too deep recursion
when creating the query plan.
| Mserver returns with "memory exhausted in" error if query size exceeds 2.5MB | https://api.github.com/repos/MonetDB/MonetDB/issues/6347/comments | 0 | 2020-11-30T15:49:55Z | 2024-06-27T13:00:11Z | https://github.com/MonetDB/MonetDB/issues/6347 | 753,591,052 | 6,347 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-07 11:38:23 +0200
From: @lsidir
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2018-03-29 15:39:18 +0200
## Comment 25434
Date: 2017-07-07 11:38:23 +0200
From: @lsidir
User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36
Build Identifier:
In gdk_batop.c line 1281 the following check is done:
if (sorted == NULL && order == NULL && groups == NULL) {
/* no place to put result, so we're done quickly */
return GDK_SUCCEED;
}
I can understand that if the caller has not allocate a bat for the result to be stored, then BATsort can not sort, but is this a success? or should BATsort return GDK_FAIL in this case?
Reproducible: Always
## Comment 25611
Date: 2017-09-05 11:14:39 +0200
From: @sjoerdmullender
It does exactly what you ask it to do, so why would that be wrong? ;-)
In any case, sorted==NULL is perfectly legitimate, but indeed if all outputs are NULL, and in particular both sorted and order are NULL, there is no point calling the function. I'll turn that into an error. Just having groups as a non-NULL pointer doesn't really make much sense either, so I'll disallow that as well.
## Comment 26317
Date: 2018-03-29 15:39:18 +0200
From: @sjoerdmullender
The Mar2018 version has been released.
| BATsort returns GDK_SUCCEED when **sorted bat is null | https://api.github.com/repos/MonetDB/MonetDB/issues/6346/comments | 0 | 2020-11-30T15:49:51Z | 2024-06-27T13:00:10Z | https://github.com/MonetDB/MonetDB/issues/6346 | 753,591,013 | 6,346 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-06 17:36:49 +0200
From: @yzchang
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @yzchang
Last updated: 2017-07-07 15:50:26 +0200
## Comment 25431
Date: 2017-07-06 17:36:49 +0200
From: @yzchang
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_5) AppleWebKit/603.2.4 (KHTML, like Gecko) Version/10.1.1 Safari/603.2.4
Build Identifier:
[Actually, I tried this in Jul2017]
According to its documentation [1], the SAMPLE operator takes an optional prefix UNIFORM. However, it accepts loads more. It seems that I can pass an arbitrary word before SAMPLE and the extra word will just be ignored:
sql>select name from _tables doesnotexist sample 0.1;
+------------------+
| name |
+==================+
| schemas |
| args |
| _tables |
| idxs |
| querylog_catalog |
+------------------+
5 tuples (1.538ms)
sql>select name from _tables blahblahblah sample 0.1;
+------------------+
| name |
+==================+
| functions |
| _tables |
| keys |
| querylog_catalog |
| statistics |
+------------------+
5 tuples (1.681ms)
sql>select name from _tables blahbl132ahblah sample 0.1;
+-----------------+
| name |
+=================+
| sequences |
| privileges |
| optimizers |
| keywords |
| privilege_codes |
+-----------------+
5 tuples (2.777ms)
Regards,
Jennie
[1] https://www.monetdb.org/Documentation/Cookbooks/SQLrecipes/Sampling
Reproducible: Always
### Actual Results:
sampled query results
### Expected Results:
syntax error
## Comment 25432
Date: 2017-07-06 19:06:00 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [13a7f8456665](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=13a7f8456665) made by Jennie Zhang <y.zhang@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=13a7f8456665](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=13a7f8456665)
Changeset description:
Added test for Bug #6345
## Comment 25433
Date: 2017-07-07 10:16:17 +0200
From: @sjoerdmullender
Has nothing to do with SAMPLE:
select * from _tables blah;
## Comment 25435
Date: 2017-07-07 15:48:02 +0200
From: @yzchang
The arbitrary word is not garbage, it is correctly regarded as the alias of the table "_tables". It is not a prefix to SAMPLE. So, this is not a bug.
The confusing was caused by mismatching implementation of SAMPLE and its documentation. According to the documentation, UNIFORM is an optional keyword that can be past to SAMPLE, but UNIFORM is not implemented as a keyword. So, in a query like "SELECT name FROM _tables UNIFORM SAMPLE 0.1;", UNIFORM becomes the alias of "_tables":
sql>SELECT UNIFORM.name FROM _tables UNIFORM SAMPLE 0.1;
+------------+
| name |
+============+
| types |
| _columns |
| privileges |
| tracelog |
| keywords |
+------------+
5 tuples (2.023ms)
Sjoerd and I agreed that the best solution at the moment is to fix the documentation.
## Comment 25436
Date: 2017-07-07 15:50:26 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [2d5fff479db3](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=2d5fff479db3) made by Jennie Zhang <y.zhang@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=2d5fff479db3](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=2d5fff479db3)
Changeset description:
Remove test of invalid Bug #6345
| SAMPLE accepts arbitrary word as its prefix and ignores it | https://api.github.com/repos/MonetDB/MonetDB/issues/6345/comments | 0 | 2020-11-30T15:49:48Z | 2024-06-28T13:10:51Z | https://github.com/MonetDB/MonetDB/issues/6345 | 753,590,976 | 6,345 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-06 14:59:03 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-08-02 10:21:30 +0200
## Comment 25426
Date: 2017-07-06 14:59:03 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:54.0) Gecko/20100101 Firefox/54.0
Build Identifier:
A short query causes assertion errors in gdk_bbp sometimes. Furthermore, dropping some of the lines leads to 'object not found' and 'illegal argument'.
Test file has neen added.
Reproducible: Always
### Steps to Reproduce:
select
sample_2.name as c0,
sample_2.id as c1,
sample_2.name as c2,
cast(coalesce(sample_2.id,
sample_2.id) as int) as c3,
sample_2.id as c4,
11 as c5,
sample_2.id as c6
from
sys.objects as sample_2
where sample_2.name is not NULL
limit 61;
## Comment 25428
Date: 2017-07-06 15:26:39 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [f09003c62bd1](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=f09003c62bd1) made by Martin Kersten <mk@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=f09003c62bd1](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=f09003c62bd1)
Changeset description:
Add test for bug #6344
## Comment 25452
Date: 2017-07-12 10:36:06 +0200
From: @njnes
fixed by proper cleanup and variable reassignment in push select optimizer
## Comment 25538
Date: 2017-08-02 10:21:30 +0200
From: @sjoerdmullender
Fixed in Jul2017-SP1 release.
| Spurious errors and assertions (SQLsmith) | https://api.github.com/repos/MonetDB/MonetDB/issues/6344/comments | 0 | 2020-11-30T15:49:44Z | 2024-06-27T13:00:08Z | https://github.com/MonetDB/MonetDB/issues/6344 | 753,590,919 | 6,344 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-06 00:26:21 +0200
From: @joerivanruth
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-10-26 14:01:31 +0200
## Comment 25423
Date: 2017-07-06 00:26:21 +0200
From: @joerivanruth
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36
Build Identifier:
This is the third report in a series of three issues found while
testing the Air Traffic queries. I'm filing separate reports because
the underlying cause may or may not be the same.
In this case, the attached queries cause the mserver5 to exit with
something like the following:
unable to find atraf.ontime_slave1(Year)
I would expect an empty query result instead.
To reproduce, first start three mserver5's as in the attached
restart.sh. Then run demo.sh. This initializes the databases and
runs the failing queries.
Be careful: restart.sh first uses 'killall' to kill ALL existing
processes named 'mserver5'.
I tested this with Mercurial revision [6a7ff140ae05](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=6a7ff140ae05) on the Jul2017
branch.
Reproducible: Always
## Comment 25424
Date: 2017-07-06 00:27:13 +0200
From: @joerivanruth
Created attachment 556
Script to reproduce the problem
> Attached file: [6343.tar.gz](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6343_6343.tar.gz_556) (application/x-gzip, 2179 bytes)
> Description: Script to reproduce the problem
## Comment 25458
Date: 2017-07-13 08:57:58 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [7bc1969a7552](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=7bc1969a7552) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7bc1969a7552](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7bc1969a7552)
Changeset description:
fixes for bugs 6341, 6342 and 6343
For 6343 more is needed.
added missing sql_div definitions (for decimals)
added parsing of anti ( ! ) in rel dump
handle aliases with in the group by operator
pushed merge table rewrites down functions
## Comment 25543
Date: 2017-08-02 17:07:42 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [c3120c666919](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=c3120c666919) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=c3120c666919](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=c3120c666919)
Changeset description:
fixed for bug #6343, ie handle passing stack variables
## Comment 25545
Date: 2017-08-02 17:22:39 +0200
From: @njnes
added support for passing function (stack based) arguments to remote
function calls.
| MERGE TABLE issue: unable to find | https://api.github.com/repos/MonetDB/MonetDB/issues/6343/comments | 0 | 2020-11-30T15:49:41Z | 2024-06-27T13:00:06Z | https://github.com/MonetDB/MonetDB/issues/6343 | 753,590,869 | 6,343 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-06 00:22:31 +0200
From: @joerivanruth
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-08-02 10:21:35 +0200
## Comment 25421
Date: 2017-07-06 00:22:31 +0200
From: @joerivanruth
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36
Build Identifier:
This is the second report in a series of three issues found while
testing the Air Traffic queries. I'm filing separate reports because
the underlying cause may or may not be the same.
In this case, the attached queries cause the slave mserver5's to crash.
To reproduce, first start three mserver5's as in the attached
restart.sh. Then run demo.sh. This initializes the databases and
runs the failing queries.
Be careful: restart.sh first uses 'killall' to kill ALL existing
processes named 'mserver5'.
I tested this with Mercurial revision [6a7ff140ae05](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=6a7ff140ae05) on the Jul2017
branch.
Reproducible: Always
## Comment 25422
Date: 2017-07-06 00:23:48 +0200
From: @joerivanruth
Created attachment 555
Script to reproduce
> Attached file: [6342.tar.gz](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6342_6342.tar.gz_555) (application/x-gzip, 2100 bytes)
> Description: Script to reproduce
## Comment 25457
Date: 2017-07-13 08:57:55 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [7bc1969a7552](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=7bc1969a7552) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7bc1969a7552](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7bc1969a7552)
Changeset description:
fixes for bugs 6341, 6342 and 6343
For 6343 more is needed.
added missing sql_div definitions (for decimals)
added parsing of anti ( ! ) in rel dump
handle aliases with in the group by operator
pushed merge table rewrites down functions
## Comment 25460
Date: 2017-07-13 09:01:51 +0200
From: @njnes
fixed by adding more sql_div definitions
## Comment 25541
Date: 2017-08-02 10:21:35 +0200
From: @sjoerdmullender
Fixed in Jul2017-SP1 release.
| MERGE TABLE issue: hang | https://api.github.com/repos/MonetDB/MonetDB/issues/6342/comments | 0 | 2020-11-30T15:49:38Z | 2024-06-27T13:00:05Z | https://github.com/MonetDB/MonetDB/issues/6342 | 753,590,819 | 6,342 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-06 00:14:52 +0200
From: @joerivanruth
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-08-02 10:21:26 +0200
## Comment 25419
Date: 2017-07-06 00:14:52 +0200
From: @joerivanruth
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36
Build Identifier:
This is the first report in a series of three issues found while
testing the Air Traffic queries. I'm filing separate reports because
the underlying cause may or may not be the same.
In this case, the attached queries cause the mserver5 to exit with
something like the following:
(mapi:monetdb://monetdb@localhost/slave1) Cannot register A0 char(5,0),A1 date(0,0),A2 date(0,0)
group by (
project (
select (
table(atraf.ontime_slave1) [ "ontime_slave1"."Year" as "ontime"."Year", ...
I would expect an empty query result instead.
To reproduce, first start three mserver5's as in the attached
restart.sh. Then run demo.sh. This initializes the databases and
runs the failing queries.
Be careful: restart.sh first uses 'killall' to kill ALL existing
processes named 'mserver5'.
I tested this with Mercurial revision [6a7ff140ae05](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=6a7ff140ae05) on the Jul2017
branch.
Reproducible: Always
## Comment 25420
Date: 2017-07-06 00:16:30 +0200
From: @joerivanruth
Created attachment 554
script to reproduce
> Attached file: [6341.tar.gz](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6341_6341.tar.gz_554) (application/x-gzip, 2222 bytes)
> Description: script to reproduce
## Comment 25456
Date: 2017-07-13 08:57:51 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [7bc1969a7552](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=7bc1969a7552) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7bc1969a7552](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=7bc1969a7552)
Changeset description:
fixes for bugs 6341, 6342 and 6343
For 6343 more is needed.
added missing sql_div definitions (for decimals)
added parsing of anti ( ! ) in rel dump
handle aliases with in the group by operator
pushed merge table rewrites down functions
## Comment 25459
Date: 2017-07-13 09:01:04 +0200
From: @njnes
fixed by handling aliases in the group by
## Comment 25536
Date: 2017-08-02 10:21:26 +0200
From: @sjoerdmullender
Fixed in Jul2017-SP1 release.
| MERGE TABLE issue: Cannot register | https://api.github.com/repos/MonetDB/MonetDB/issues/6341/comments | 0 | 2020-11-30T15:49:35Z | 2024-06-27T13:00:04Z | https://github.com/MonetDB/MonetDB/issues/6341 | 753,590,780 | 6,341 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-05 11:56:48 +0200
From: pavlos <<pavlos.katsogridakis>>
To: SQL devs <<bugs-sql>>
Version: 11.23.3 (Jun2016)
CC: @njnes, @yzchang
Last updated: 2017-08-02 10:21:31 +0200
## Comment 25409
Date: 2017-07-05 11:56:48 +0200
From: pavlos <<pavlos.katsogridakis>>
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:54.0) Gecko/20100101 Firefox/54.0
Build Identifier:
Tried to run the tpch 03 query, taking a 10% sample on each table,
so I appended sample 0.1; to the query.
I expected the query to take a 10% sample of the tables and then run the
query. Instead, it executed the query and then it took a 10% sample.
Reproducible: Always
### Steps to Reproduce:
1.Setup tpch with sf10
2.Append sample 0.1; to the 03.sql query
## Comment 25413
Date: 2017-07-05 18:33:37 +0200
From: @njnes
indeed the sample operator (like limit) does it job on top of the result. Sampling per input should be done by hand.
## Comment 25425
Date: 2017-07-06 11:14:36 +0200
From: @yzchang
Yeah, we got that sampling the base tables requires some manual work.
But how about the combined use of SAMPLE and LIMIT? Shouldn't SAMPLE be executed before LIMIT? Or should we actually disallow their combined use, since they both reduce the result set, only using different algorithms?
## Comment 25539
Date: 2017-08-02 10:21:31 +0200
From: @sjoerdmullender
Fixed in Jul2017-SP1 release.
| sample operator takes effect after the execution of the query, expected before | https://api.github.com/repos/MonetDB/MonetDB/issues/6340/comments | 0 | 2020-11-30T15:49:31Z | 2024-06-27T13:00:03Z | https://github.com/MonetDB/MonetDB/issues/6340 | 753,590,731 | 6,340 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-04 14:22:06 +0200
From: @joerivanruth
To: SQL devs <<bugs-sql>>
Version: 11.25.23 (Dec2016-SP5)
CC: @njnes
Last updated: 2017-08-02 10:21:23 +0200
## Comment 25408
Date: 2017-07-04 14:22:06 +0200
From: @joerivanruth
The SQL script
CREATE TABLE t(a INT);
CREATE TABLE x(b INT, c1 INT);
SELECT * FROM
(SELECT c1, (SELECT COUNT(a) FROM t WHERE a = b) AS c2 FROM x) AS dummy
WHERE c1 = c2;
makes mserver5 crash with
sql/server/rel_optimizer.c:5494: rel_push_project_up: Assertion `e' failed.
This happens both on Dec2016_SP5 and on the Jul2017 branch.
## Comment 25412
Date: 2017-07-05 18:32:03 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [6a7ff140ae05](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=6a7ff140ae05) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=6a7ff140ae05](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=6a7ff140ae05)
Changeset description:
fixed bug #6339, ie fixed problem in column lookup in combination with
the apply operator
## Comment 25535
Date: 2017-08-02 10:21:23 +0200
From: @sjoerdmullender
Fixed in Jul2017-SP1 release.
| Mserver5 crashes on nested SELECT | https://api.github.com/repos/MonetDB/MonetDB/issues/6339/comments | 0 | 2020-11-30T15:49:29Z | 2024-06-27T13:00:01Z | https://github.com/MonetDB/MonetDB/issues/6339 | 753,590,703 | 6,339 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-03 05:53:16 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-07-06 15:26:30 +0200
## Comment 25405
Date: 2017-07-03 05:53:16 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:54.0) Gecko/20100101 Firefox/54.0
Build Identifier:
A compound SQL query causes an assertion to be raised in a query.
The script is added to the testsuite.
Reproducible: Always
## Comment 25411
Date: 2017-07-05 15:49:26 +0200
From: @njnes
what is the name of the test script
## Comment 25427
Date: 2017-07-06 15:26:30 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [d2d82627e285](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=d2d82627e285) made by Martin Kersten <mk@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=d2d82627e285](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=d2d82627e285)
Changeset description:
Added test file for bug #6338
| SQL raises assert in relbin_groupby | https://api.github.com/repos/MonetDB/MonetDB/issues/6338/comments | 0 | 2020-11-30T15:49:25Z | 2024-06-27T13:00:00Z | https://github.com/MonetDB/MonetDB/issues/6338 | 753,590,657 | 6,338 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-07-01 19:52:03 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @PedroTadim
Last updated: 2020-02-04 16:41:28 +0100
## Comment 25404
Date: 2017-07-01 19:52:03 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:53.0) Gecko/20100101 Firefox/53.0
Build Identifier:
The following statement does not work
SET col = (SELECT color FROM colors where hdl is NULL SAMPLE 1);
and should be done along this verbose construact:
CREATE TABLE cs(color string);
INSERT INTO cs SELECT color FROM colors where hdl is NULL SAMPLE 1;
SET col = (SELECT * FROM cs );
Reproducible: Always
### Steps to Reproduce:
CREATE TABLE colors(hdl STRING, color STRING);
INSERT INTO colors VALUES
(null,'yellow'),
(null,'purple'),
(null,'green');
DECLARE col STRING;
CREATE TABLE cs(color string);
INSERT INTO cs SELECT color FROM colors where hdl is NULL SAMPLE 1;
SET col = (SELECT * FROM cs );
SET col = (SELECT color FROM colors where hdl is NULL SAMPLE 1);
## Comment 27190
Date: 2019-07-29 14:41:33 +0200
From: @PedroTadim
SAMPLE 1 is being called inside a sub-query, so this applies to the LIMIT clause inside sub-queries as well. Because SQL is a set language calling SAMPLE n inside a sub-query can lead to non-deterministic results. So we shouldn't implement this.
| Allow SAMPLE 1 in scalar returning query | https://api.github.com/repos/MonetDB/MonetDB/issues/6337/comments | 0 | 2020-11-30T15:49:22Z | 2024-06-28T13:39:38Z | https://github.com/MonetDB/MonetDB/issues/6337 | 753,590,609 | 6,337 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-06-29 07:35:26 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-08-02 10:21:22 +0200
## Comment 25401
Date: 2017-06-29 07:35:26 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:53.0) Gecko/20100101 Firefox/53.0
Build Identifier:
Insertion of VALUES based on multiple scalar subqueries are silently ignored.
Test file added
Reproducible: Always
### Steps to Reproduce:
START TRANSACTION;
create table data(i integer);
insert into data values(0),(1),(2);
create table multiples(i integer);
insert into multiples VALUES((select count(*) from data)), ((select count(distinct i) from data));
select * from multiples;
insert into multiples VALUES((select count(*) from data));
insert into multiples VALUES((select count(distinct i) from data));
select * from multiples;
ROLLBACK;
### Actual Results:
sql>insert into multiples VALUES((select count(*) from data)), ((select count(distinct i) from data));
...
sql>select * from multiples;
+---+
| i |
+===+
sql>insert into multiples VALUES((select count(*) from data));
1 affected row (0.917ms)
sql>insert into multiples VALUES((select count(distinct i) from data));
1 affected row (1.069ms)
sql>select * from multiples;
+------+
| i |
+======+
| 3 |
| 3 |
+------+
+---+
## Comment 25402
Date: 2017-06-29 09:30:08 +0200
From: @sjoerdmullender
This syntax is not supported. Is it at all legal according to the standard?
The syntax is
INSERT INTO table SELECT ...;
You can use a UNION to insert from multiple SELECT queries in one go.
## Comment 25414
Date: 2017-07-05 21:19:47 +0200
From: @njnes
the standard seems to allow this as long as the list holds single values, ie
a subquery which results in a single value is allowed.
Added a fix for this, ie handle value lists with queries.
## Comment 25415
Date: 2017-07-05 22:12:35 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [2cbddfe7dcb8](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=2cbddfe7dcb8) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=2cbddfe7dcb8](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=2cbddfe7dcb8)
Changeset description:
handle subqueries in value lists, solves bug #6336
## Comment 25429
Date: 2017-07-06 15:26:44 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [c162d873d44a](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=c162d873d44a) made by Martin Kersten <mk@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=c162d873d44a](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=c162d873d44a)
Changeset description:
Addition of bug #6336
## Comment 25430
Date: 2017-07-06 15:26:50 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [37aeb1b76bbf](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=37aeb1b76bbf) made by Martin Kersten <mk@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=37aeb1b76bbf](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=37aeb1b76bbf)
Changeset description:
Addition of bug #6336
## Comment 25534
Date: 2017-08-02 10:21:22 +0200
From: @sjoerdmullender
Fixed in Jul2017-SP1 release.
| VALUES multiple inserts error | https://api.github.com/repos/MonetDB/MonetDB/issues/6336/comments | 0 | 2020-11-30T15:49:20Z | 2024-06-27T12:59:58Z | https://github.com/MonetDB/MonetDB/issues/6336 | 753,590,581 | 6,336 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-06-26 11:11:10 +0200
From: @PedroTadim
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-10-26 14:01:28 +0200
## Comment 25394
Date: 2017-06-26 11:11:10 +0200
From: @PedroTadim
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.109 Safari/537.36
Build Identifier:
Mserver takes too long processing a single but complex SQL query. After a certain time, it receives a SIGKILL signal.
Reproducible: Always
### Steps to Reproduce:
1. SELECT DISTINCT 8 * + + CASE + NULLIF ( - - CASE - ( 52 ) WHEN + 18 - - COUNT ( * ) THEN + 4 WHEN + + 93 - - ( - 40 ) THEN - - 1 * 27 * MIN ( DISTINCT - 85 ) + - - 29 * - + CASE - CASE - 66 WHEN - + 87 / 51 THEN 25 * 78 END WHEN - 12 * - 38 / + - 64 * - - CASE - - 61 WHEN + - 84 + 83 + + 36 * NULLIF ( - 46, - 21 ) * 21 + - 18 + - 21 THEN + 8 * 45 WHEN + 40 THEN NULL END * 7 THEN NULL WHEN - 8 THEN SUM ( + 62 ) * 44 WHEN - ( - ( + ( 6 ) ) ) THEN + 91 + + 38 ELSE NULL END END, + 57 ) WHEN - CAST ( NULL AS INTEGER ) THEN ( 46 ) ELSE NULL END * - 77;
### Actual Results:
The server receives a SIGKILL signal after a while.
### Expected Results:
A single NULL value.
The merovingian.log only shows the SIGKILL line:
2017-06-26 10:49:42 MSG merovingian[2946]: database 'db' (26839) was killed by signal SIGKILL
## Comment 25529
Date: 2017-07-31 09:58:36 +0200
From: @sjoerdmullender
Something goes seiously wrong in the SQL optimizer. I stopped the execution at a point where the MAL variable stack was 3853412 entries big and still growing. That seems excessive for this query.
## Comment 25589
Date: 2017-08-18 13:37:33 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [484ca02b0773](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=484ca02b0773) made by Pedro Ferreira <pedro.ferreira@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=484ca02b0773](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=484ca02b0773)
Changeset description:
Added test for Bug #6335
## Comment 25637
Date: 2017-09-17 09:26:27 +0200
From: @njnes
fixed crash, we now rewrite large expressions into smaller onces (which solves
the explosion when rewriting for division by zero and similar problems)
## Comment 25641
Date: 2017-09-17 18:01:40 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [9ccae5c8d830](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=9ccae5c8d830) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=9ccae5c8d830](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=9ccae5c8d830)
Changeset description:
re-implemented declared tables, to solve the side-effect bug #6397.
also lots of fixes for the sqllitelogistest_case bug #6335 (doesn't
crash (or run out of memory) anymore).
## Comment 25662
Date: 2017-09-20 12:20:31 +0200
From: @PedroTadim
The crash does not happen anymore, but a MAL error occurs:
!TypeException:user.main[236]:'calc.*' undefined in: X_26992:sht := calc.*(X_26990:int, X_26991:bte);
So I guess the calc operations are not implemented across numeric data types.
## Comment 25724
Date: 2017-10-15 18:57:27 +0200
From: @njnes
The function type problem was caused in the inplace type casting code.
## Comment 25725
Date: 2017-10-15 20:25:51 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [b1efe52241ae](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=b1efe52241ae) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=b1efe52241ae](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=b1efe52241ae)
Changeset description:
fixed bug #6335, ie don't do inplace conversions on shared atoms
| Sqlitelogictest crash in complex SQL query | https://api.github.com/repos/MonetDB/MonetDB/issues/6335/comments | 0 | 2020-11-30T15:49:16Z | 2024-06-27T12:59:57Z | https://github.com/MonetDB/MonetDB/issues/6335 | 753,590,531 | 6,335 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-06-20 15:44:15 +0200
From: @eyalroz
To: clients devs <<bugs-clients>>
Version: 11.25.5 (Dec2016-SP1)
Last updated: 2017-06-20 15:44:15 +0200
## Comment 25385
Date: 2017-06-20 15:44:15 +0200
From: @eyalroz
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:53.0) Gecko/20100101 Firefox/53.0
Build Identifier:
The reason for rejecting COPY INTO using relative pathnames is that these cannot be relative to the CWD of the client, but rather of the server process - which the user is unlikely to know; so we demand an absolute pathname.
Now, the current error message just says:
ERROR = !COPY INTO: filename must have absolute path: relative/path/goes/here/some_file.txt
but that might just lead the user to think they can supply a relative path on their own machine. It would be better to say:
ERROR = !COPY INTO: filename must have absolute path on the server machine: relative/path/goes/here/some_file.txt
Reproducible: Always
| Need better error message when rejecting a COPY INTO from a relative path | https://api.github.com/repos/MonetDB/MonetDB/issues/6334/comments | 0 | 2020-11-30T15:49:13Z | 2024-06-28T13:10:50Z | https://github.com/MonetDB/MonetDB/issues/6334 | 753,590,489 | 6,334 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-06-06 18:37:05 +0200
From: @lsidir
To: SQL devs <<bugs-sql>>
Version: -- development
CC: martin.van.dinther, @drstmane
Last updated: 2019-12-11 17:29:59 +0100
## Comment 25373
Date: 2017-06-06 18:37:05 +0200
From: @lsidir
User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36
Build Identifier:
they are especially usefull if you want to select columns that are not part of a group by. For example if the table is
|a|b|
-----
|1|x|
+-+-+
|1|y|
+-+-+
|2|z|
+-+-+
select table.a, first(table.b) from table group by table.a;
will return
+-+-+
|1|x|
+-+-+
|2|z|
+-+-+
Reproducible: Always
## Comment 25377
Date: 2017-06-08 13:35:22 +0200
From: Martin van Dinther <<martin.van.dinther>>
SQL is a set-oriented language in which the (storage) order is not relevant, meaning it can change between queries (different data processing strategies) and during its life time (e.g. when an internal reorganisation or backup/restore is done).
So aggregates FIRST() and LAST() could return different values when running the same query multiple times.
Instead you could/should use MIN() and MAX() aggregates.
create table tabel (a int, b char(1));
insert into tabel values (1, 'x'), (1, 'y'), (2, 'z');
select a, min(b) as min_b, max(b) as max_b from tabel group by a order by a;
## Comment 25378
Date: 2017-06-08 13:42:55 +0200
From: @lsidir
MIN and MAX are extra work. Returning differernt values if the storage has been updated is ok. See oracle or postgress how they support first/last:
https://wiki.postgresql.org/wiki/First/last_(aggregate)
## Comment 25455
Date: 2017-07-12 14:56:56 +0200
From: @drstmane
As Martin vD suggested in his comment, FIRST() and LAST() have no clearly defined (intuitive) semantics in set-oriented SQL --- even LIMIT n without order by does not, other some returning "any" n tuples, possibly different ones with each query invocation (i.e., not even "guaranteed" to be deterministic).
Thus, maybe we could consider "ANY()" instead of "FIRST()" & "LAST()"?
Stefan
## Comment 27477
Date: 2019-12-11 17:29:59 +0100
From: Martin van Dinther <<martin.van.dinther>>
As of Apr2019 release (11.33.3) we support SQL window functions:
FIRST_VALUE '(' query_expression ')'
LAST_VALUE '(' query_expression ')'
and
NTH_VALUE '(' query_expression ',' query_expression ')'
See https://www.monetdb.org/blog/extended_sql_window_functions
With these new functions you could implement your desired query outputs by using for instance:
sql>select distinct a, first_value(b) over(partition by a) from tabel;
+------+------+
| a | L2 |
+======+======+
| 1 | x |
| 2 | z |
+------+------+
2 tuples
sql>select distinct a, last_value(b) over(partition by a) from tabel;
+------+------+
| a | L2 |
+======+======+
| 1 | y |
| 2 | z |
+------+------+
2 tuples
Alternatively you can create your own aggregate function using CREATE AGGREGATE command.
See https://www.monetdb.org/Documentation/Manuals/SQLreference/SQLSyntaxOverviewCREATE_AGGREGATE_LANGUAGE
for an example defined in language Python. Other supported languages are C, C++ and R if you prefer those.
| Feature Request: Support aggregation functions FIRST() and LAST() | https://api.github.com/repos/MonetDB/MonetDB/issues/6333/comments | 0 | 2020-11-30T15:49:10Z | 2024-06-28T13:39:38Z | https://github.com/MonetDB/MonetDB/issues/6333 | 753,590,451 | 6,333 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-06-06 10:29:59 +0200
From: @PedroTadim
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-07-17 16:07:18 +0200
## Comment 25372
Date: 2017-06-06 10:29:59 +0200
From: @PedroTadim
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.86 Safari/537.36
Build Identifier:
The mserver crashes after executing two SELECT queries consecutively. The first one throws a MAL exception. Then after executing the second one, the server crashes.
Reproducible: Always
### Steps to Reproduce:
1. CREATE TABLE tab0(col0 INTEGER, col1 INTEGER, col2 INTEGER);
2. SELECT DISTINCT col1 * + + col1 + + col2 * 0 / + - CAST ( + - col0 AS INTEGER ) AS col1 FROM tab0 cor0;
3. SELECT ALL + 93 + col0 FROM tab0 AS cor0;
### Actual Results:
During the first query, the following error exception is thrown:
TypeException:user.main[19]:'bat.append' undefined in: bat.append(X_24937:bat[:lng],X_24943:bte);
During the second one, the server gets a SIGSEV.
### Expected Results:
A empty result set for both SELECT queries.
While running the first SELECT query the following warning is thrown
WARNING To speedup calc./ a bulk operator implementation is needed
X_455:bat[:lng] := mal.multiplex("calc":str,"/":str,0:bte,X_454:bat[:int]);
The bt of the SIGSEV:
0 0x00007ffff4c0e92a in __memset_avx2_erms () from /lib64/libc.so.6
1 0x00007ffff75e8b87 in GDKfree (blk=0x7fffb010fa60) at gdk_utils.c:1745
2 0x00007ffff797c070 in freeInstruction (p=0x7fffb010fa60) at mal_instruction.c:482
3 0x00007ffff79a9946 in MSresetInstructions (mb=0x7fffb00cf2a0, start=1) at mal_session.c:372
4 0x00007fffe8c958c1 in SQLengineIntern (c=0x7fffea35b328, be=0x7fffb00c1b00) at sql_execute.c:717
5 0x00007fffe8c929ea in SQLengine (c=0x7fffea35b328) at sql_scenario.c:1192
6 0x00007ffff79a80b8 in runPhase (c=0x7fffea35b328, phase=4) at mal_scenario.c:517
7 0x00007ffff79a8291 in runScenarioBody (c=0x7fffea35b328) at mal_scenario.c:559
8 0x00007ffff79a839f in runScenario (c=0x7fffea35b328) at mal_scenario.c:579
9 0x00007ffff79a9dd6 in MSserveClient (dummy=0x7fffea35b328) at mal_session.c:448
10 0x00007ffff79a98db in MSscheduleClient (command=0x7fffb00008d0 "0", challenge=0x7fffc46cedf0 "VKaSG5wf", fin=0x7fffb0002980, fout=0x7fffbc002b60) at mal_session.c:339
11 0x00007ffff7a49d19 in doChallenge (data=0x7fffbc0008d0) at mal_mapi.c:188
12 0x00007ffff762d0f1 in thread_starter (arg=0x7fffbc004c50) at gdk_system.c:485
13 0x00007ffff4e8c6ca in start_thread () from /lib64/libpthread.so.0
14 0x00007ffff4bc6f7f in clone () from /lib64/libc.so.6
## Comment 25391
Date: 2017-06-22 14:29:21 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [3e038283a86f](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=3e038283a86f) made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=3e038283a86f](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=3e038283a86f)
Changeset description:
Add test for bug #6332
## Comment 25395
Date: 2017-06-28 10:12:32 +0200
From: @njnes
fixed bug in optimizer
| Sqlitelogictest crash related to an undefined MAL function | https://api.github.com/repos/MonetDB/MonetDB/issues/6332/comments | 0 | 2020-11-30T15:49:06Z | 2024-06-27T12:59:54Z | https://github.com/MonetDB/MonetDB/issues/6332 | 753,590,403 | 6,332 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-06-01 19:02:33 +0200
From: Martin van Dinther <<martin.van.dinther>>
To: SQL devs <<bugs-sql>>
Version: 11.25.23 (Dec2016-SP5)
Last updated: 2018-03-29 16:44:25 +0200
## Comment 25370
Date: 2017-06-01 19:02:33 +0200
From: Martin van Dinther <<martin.van.dinther>>
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:53.0) Gecko/20100101 Firefox/53.0
Build Identifier:
It appears that column "nils" of sys.statistics always contains 0 after running ANALYSE on all columns of all tables of schemas: sys, tmp and profiler.
I expected that columns "nils" would contain the number of nils/NULLs found in the specific column.
Reproducible: Always
### Steps to Reproduce:
1. Run SQL commands:
ANALYSE sys;
ANALYSE tmp;
ANALYSE profiler;
SELECT * FROM sys.statistics WHERE nils <> 0.
SELECT * FROM sys.statistics WHERE minval = 'nil' OR maxval = 'nil';
SELECT * FROM sys.statistics WHERE (minval = 'nil' OR maxval = 'nil') AND nils <> 0;
### Actual Results:
nils always contain 0 for all columns analyzed, although there are many columns which have 'nil' as minimum or maximum value.
### Expected Results:
For columns that have NULLs/nils value, the number of occurences in that column.
## Comment 25383
Date: 2017-06-20 11:28:37 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [b4f5f29a002f](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=b4f5f29a002f) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=b4f5f29a002f](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=b4f5f29a002f)
Changeset description:
Fix BATselect comment so that it tells the truth; fix call to select for NIL.
This fixes bug #6331.
## Comment 25390
Date: 2017-06-22 14:14:15 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [567718e3de97](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=567718e3de97) made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=567718e3de97](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=567718e3de97)
Changeset description:
Add test and output files for bug #6331
## Comment 26330
Date: 2018-03-29 16:44:25 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [8920ec11dedc](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=8920ec11dedc) made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=8920ec11dedc](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=8920ec11dedc)
Changeset description:
Fix memory leak reported on
http://monetdb.cwi.nl/testweb/web/showtestoutput.php?serial=66173:f34a57f73307&target=GNU-Fedora-x86_64-sanitize&module=sql&test=sql%2Ftest%2FBugTracker-2017%2Fstatistics_nils_not_eq_zero.Bug-6331&which=err
The leak is introduced recently via [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=78237ac1720c](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=78237ac1720c)
I also found and removed two unneccesary GDKfree() calls to pointers which are already NULL.
| sys.statistics column "nils" always contains 0. Expected a positive value for columns that have one or more nils/NULLs | https://api.github.com/repos/MonetDB/MonetDB/issues/6331/comments | 0 | 2020-11-30T15:49:03Z | 2024-06-27T12:59:53Z | https://github.com/MonetDB/MonetDB/issues/6331 | 753,590,349 | 6,331 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-06-01 12:14:01 +0200
From: @PedroTadim
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-07-17 16:07:31 +0200
## Comment 25369
Date: 2017-06-01 12:14:01 +0200
From: @PedroTadim
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.36
Build Identifier:
Mserver crashes with an assertion query in a SELECT query without the FROM clause.
Reproducible: Always
### Steps to Reproduce:
1. SELECT 45 - - CASE + + 70 WHEN COUNT ( * ) + SUM ( DISTINCT 91 ) + MIN ( DISTINCT 41 ) / 65 THEN + 98 END * + - 0 - - COUNT ( * ) AS col0;
### Actual Results:
The server gets an assertion error and crashes.
### Expected Results:
A single NULL value in the result set.
Error log:
2017-06-01 12:05:31 ERR db[26005]: could not find L3.L3
2017-06-01 12:05:31 ERR db[26005]: mserver5: rel_bin.c:2708: rel2bin_groupby: Assertion `0' failed.
2017-06-01 12:05:35 MSG merovingian[31068]: database 'db' (26005) has crashed (dumped core)
Backtrace:
0 0x00007ffff4b6591f in raise () from /lib64/libc.so.6
1 0x00007ffff4b6751a in abort () from /lib64/libc.so.6
2 0x00007ffff4b5dda7 in __assert_fail_base () from /lib64/libc.so.6
3 0x00007ffff4b5de52 in __assert_fail () from /lib64/libc.so.6
4 0x00007fffe5a8cfbe in rel2bin_groupby (be=0x7fffac0cb0c0, rel=0x7fffac0e1ee0, refs=0x7fffac0ea360) at rel_bin.c:2708
5 0x00007fffe5a96264 in subrel_bin (be=0x7fffac0cb0c0, rel=0x7fffac0e1ee0, refs=0x7fffac0ea360) at rel_bin.c:4797
6 0x00007fffe5a8bac1 in rel2bin_project (be=0x7fffac0cb0c0, rel=0x7fffac0e2200, refs=0x7fffac0ea360, topn=0x0) at rel_bin.c:2418
7 0x00007fffe5a96208 in subrel_bin (be=0x7fffac0cb0c0, rel=0x7fffac0e2200, refs=0x7fffac0ea360) at rel_bin.c:4789
8 0x00007fffe5a964ff in output_rel_bin (be=0x7fffac0cb0c0, rel=0x7fffac0e2200) at rel_bin.c:4854
9 0x00007fffe5ab00b7 in sql_relation2stmt (be=0x7fffac0cb0c0, r=0x7fffac0e2200) at sql_gencode.c:472
10 0x00007fffe5ab0416 in backend_dumpstmt (be=0x7fffac0cb0c0, mb=0x7fffac0cb680, r=0x7fffac0e2200, top=1, add_end=0,
query=0x7fffac109d20 "select 45 - - case + + 70 when count ( * ) + sum ( distinct 91 ) + min ( distinct 41 ) / 65 then + 98 end * + - 0 - - count ( * ) as col0;") at sql_gencode.c:513
11 0x00007fffe5a76f83 in SQLparser (c=0x7fffe7140338) at sql_scenario.c:1119
12 0x00007ffff7a1aab7 in runPhase (c=0x7fffe7140338, phase=1) at mal_scenario.c:517
13 0x00007ffff7a1abfb in runScenarioBody (c=0x7fffe7140338) at mal_scenario.c:550
14 0x00007ffff7a1ad9e in runScenario (c=0x7fffe7140338) at mal_scenario.c:579
15 0x00007ffff7a1c844 in MSserveClient (dummy=0x7fffe7140338) at mal_session.c:462
16 0x00007ffff7a1c334 in MSscheduleClient (command=0x7fffac0008d0 "0", challenge=0x7fffc2269df0 "J3gbRCwmGp", fin=0x7fffac002990, fout=0x7fffb8002b70, protocol=PROTOCOL_9, blocksize=8190,
compute_column_widths=0) at mal_session.c:353
17 0x00007ffff7abd4cc in doChallenge (data=0x7fffb80008d0) at mal_mapi.c:269
18 0x00007ffff76a24f3 in thread_starter (arg=0x7fffb8004c60) at gdk_system.c:485
19 0x00007ffff4efd6ca in start_thread () from /lib64/libpthread.so.0
20 0x00007ffff4c37f7f in clone () from /lib64/libc.so.6
## Comment 25371
Date: 2017-06-03 13:01:37 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [9a36e7f9cd0b](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=9a36e7f9cd0b) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=9a36e7f9cd0b](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=9a36e7f9cd0b)
Changeset description:
fixed bug #6330, ie do not clean expression which are later used in same
the projection.
## Comment 25389
Date: 2017-06-22 14:10:34 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [5a7ed5bff986](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=5a7ed5bff986) made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=5a7ed5bff986](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=5a7ed5bff986)
Changeset description:
Add and approve output files for bug #6330
## Comment 25393
Date: 2017-06-25 12:03:17 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [94b5d2589cba](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=94b5d2589cba) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=94b5d2589cba](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=94b5d2589cba)
Changeset description:
fix bug #6330, ie handle null*0 in times Zero optimization.
| Sqlitelogictest crash on a complex SELECT query | https://api.github.com/repos/MonetDB/MonetDB/issues/6330/comments | 0 | 2020-11-30T15:49:00Z | 2024-06-27T12:59:52Z | https://github.com/MonetDB/MonetDB/issues/6330 | 753,590,301 | 6,330 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-30 18:35:53 +0200
From: Richard Hughes <<richard.monetdb>>
To: SQL devs <<bugs-sql>>
Version: 11.25.21 (Dec2016-SP4)
CC: @njnes
Last updated: 2020-11-13 10:59:16 +0100
## Comment 25358
Date: 2017-05-30 18:35:53 +0200
From: Richard Hughes <<richard.monetdb>>
Build is Jul2017 [24dd2ff6c348](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=24dd2ff6c348) Also reproduced on Jun2016.
sql>create table a (x int, y int not null);
operation successful (27.710ms)
sql>\d a
CREATE TABLE "sys"."a" (
"x" INTEGER,
"y" INTEGER NOT NULL
);
sql>create table b as select * from a with no data;
operation successful (29.244ms)
sql>\d b
CREATE TABLE "sys"."b" (
"x" INTEGER,
"y" INTEGER
);
Expected result: column b.y should be "NOT NULL".
I haven't made any effort to debug this myself - the issue isn't important to me.
## Comment 25392
Date: 2017-06-22 14:58:32 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [605a704cbd8a](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=605a704cbd8a) made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=605a704cbd8a](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=605a704cbd8a)
Changeset description:
Add test and output files for bug #6329
## Comment 25585
Date: 2017-08-16 21:55:36 +0200
From: @njnes
not sure this realy is a bug. A constraint isn't part of a select * or
more complex query, its part of the base table. I think a better solution
is to use create table a like ..
## Comment 25587
Date: 2017-08-17 14:00:03 +0200
From: Richard Hughes <<richard.monetdb>>
Survey of databases I have easy access to (key: Y=nullable constraints are copied, N=not copied i.e. like MonetDB):
N PostgreSQL
N SQLite
Y MS SQL Server
Y MariaDB
Conclusion: there's zero consistency and MonetDB generally behaves like PostgreSQL in other areas, therefore this is not a bug.
## Comment 28266
Date: 2020-11-13 10:59:16 +0100
From: @sjoerdmullender
Time to properly close this, since it's not a bug.
| "create table as" doesn't copy "not null" property | https://api.github.com/repos/MonetDB/MonetDB/issues/6329/comments | 0 | 2020-11-30T15:48:56Z | 2024-06-28T13:10:50Z | https://github.com/MonetDB/MonetDB/issues/6329 | 753,590,251 | 6,329 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-30 15:32:20 +0200
From: Richard Hughes <<richard.monetdb>>
To: SQL devs <<bugs-sql>>
Version: 11.25.21 (Dec2016-SP4)
CC: @njnes
Last updated: 2017-08-02 10:21:34 +0200
## Comment 25357
Date: 2017-05-30 15:32:20 +0200
From: Richard Hughes <<richard.monetdb>>
Build is Jul2017 [24dd2ff6c348](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=24dd2ff6c348) Reproducible: always.
Starting with a new, blank database:
*****open new mclient instance 'sql1' here*****
sql1>create table sub1 (i int);
sql1>create table sub2 (i int);
sql1>insert into sub1 values (10),(20);
sql1>insert into sub2 values (30),(40);
sql1>analyze sys.sub1;
sql1>analyze sys.sub2;
sql1>alter table sub1 set read only;
sql1>alter table sub2 set read only;
sql1>create merge table mt as select * from sub1 with no data;
sql1>alter table mt add table sub1;
sql1>alter table mt add table sub2;
sql1>plan select * from mt where i between 25 and 45;
+--------------------------------------------------------+
| rel |
+========================================================+
| project ( |
| | select ( |
| | | table(sys.sub2) [ mt.i, mt.%TID% NOT NULL ] COUNT |
| | ) [ int "25" <= mt.i <= int "45" ] |
| ) [ mt.i ] |
+--------------------------------------------------------+
*****open new mclient instance 'sql2' here*****
sql2>plan select * from mt where i between 25 and 45;
+--------------------------------------------------------+
| rel |
+========================================================+
| project ( |
| | select ( |
| | | table(sys.sub2) [ mt.i, mt.%TID% NOT NULL ] COUNT |
| | ) [ int "25" <= mt.i <= int "45" ] |
| ) [ mt.i ] |
+--------------------------------------------------------+
sql1>alter table sub1 set read write;
sql1>plan select * from mt where i between 25 and 45;
+-----------------------------------------------------------------------------+
| rel |
+=============================================================================+
| union ( |
| | project ( |
| | | select ( |
| | | | table(sys.sub1) [ sub1.i as mt.i, sub1.%TID% NOT NULL as mt.%TID% ] C |
: OUNT :
| | | ) [ int "25" <= mt.i <= int "45" ] |
| | ) [ mt.i ], |
| | project ( |
| | | select ( |
| | | | table(sys.sub2) [ sub2.i as mt.i, sub2.%TID% NOT NULL as mt.%TID% ] C |
: OUNT :
| | | ) [ int "25" <= mt.i <= int "45" ] |
| | ) [ mt.i ] |
| ) [ mt.i ] |
+-----------------------------------------------------------------------------+
sql2>plan select * from mt where i between 25 and 45;
+--------------------------------------------------------+
| rel |
+========================================================+
| project ( |
| | select ( |
| | | table(sys.sub2) [ mt.i, mt.%TID% NOT NULL ] COUNT |
| | ) [ int "25" <= mt.i <= int "45" ] |
| ) [ mt.i ] |
+--------------------------------------------------------+
I think this is a fix:
diff -r [24dd2ff6c348](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=24dd2ff6c348) sql/storage/store.c
--- a/sql/storage/store.c Tue May 30 13:25:19 2017 +0200
+++ b/sql/storage/store.c Tue May 30 14:25:14 2017 +0100
@@ -3402,6 +3402,7 @@
ft->base.wtime = ft->base.rtime = 0;
ft->cleared = 0;
+ ft->access = pft->access;
ok = reset_changeset( tr, &ft->columns, &pft->columns, &ft->base, (resetf) &reset_column, (dupfunc) &column_dup);
if (ok == LOG_OK)
ok = reset_changeset( tr, &ft->idxs, &pft->idxs, &ft->base, (resetf) &reset_idx, (dupfunc) &idx_dup);
Note that when preparing this bug report I managed to get the following after applying the above fix:
sql2>plan select * from mt where i between 25 and 45;
+-----------------------------------------------------------------------------+
| rel |
+=============================================================================+
| union ( |
| | project ( |
| | | select ( |
| | | | table(sys.sub2) [ sub2.i as mt.i, sub2.%TID% NOT NULL as mt.%TID% ] C |
: OUNT :
| | | ) [ int "25" <= mt.i <= int "45" ] |
| | ) [ mt.i ], |
| | project ( |
| | | select ( |
| | | | table(sys.sub2) [ sub2.i as mt.i, sub2.%TID% NOT NULL as mt.%TID% ] C |
: OUNT :
| | | ) [ int "25" <= mt.i <= int "45" ] |
| | ) [ mt.i ] |
| ) [ mt.i ] |
+-----------------------------------------------------------------------------+
Notice that it's reading from sub2 twice. I've been unable to reproduce that, but I suspect that there might be a missing line in reset_table() something like:
reset_changeset( tr, &ft->tables, &pft->tables, &ft->base, (resetf) NULL, (dupfunc) &table_of_tables_dup);
I had a couple of goes at reproducing this secondary fault but gave up fairly quickly. I'm not using merge tables therefore I'm not interested (see bug #6326).
## Comment 25407
Date: 2017-07-04 13:10:12 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [ec563c09d093](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=ec563c09d093) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=ec563c09d093](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=ec563c09d093)
Changeset description:
use a changeset of 'part' information in stead of
directly tables. This to solve order dependence (ie
is merge or replica table created before or after
the partition/member). This solves bug #6328
## Comment 25410
Date: 2017-07-05 12:19:09 +0200
From: Richard Hughes <<richard.monetdb>>
Thanks Niels.
Just to maintain the historical record, changeset [ec563c09d093](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=ec563c09d093) fixed bug #6325. This bug was fixed by part of changeset [5aa1c0ab39a2](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=5aa1c0ab39a2)
## Comment 25540
Date: 2017-08-02 10:21:34 +0200
From: @sjoerdmullender
Fixed in Jul2017-SP1 release.
| Transactional/multi-connection issues with merge tables | https://api.github.com/repos/MonetDB/MonetDB/issues/6328/comments | 0 | 2020-11-30T15:48:53Z | 2024-06-27T12:59:50Z | https://github.com/MonetDB/MonetDB/issues/6328 | 753,590,212 | 6,328 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-30 12:25:31 +0200
From: @kutsurak
To: Merovingian devs <<bugs-merovingian>>
Version: 11.25.21 (Dec2016-SP4)
Last updated: 2017-07-17 16:07:46 +0200
## Comment 25355
Date: 2017-05-30 12:25:31 +0200
From: @kutsurak
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:53.0) Gecko/20100101 Firefox/53.0
Build Identifier:
If you configure from source using the --program_suffix or --program_prefix flags, monetdbd cannot start any database, because it is forking with the monetdbd executable instead of the mserver5 executable.
Reproducible: Always
### Steps to Reproduce:
1. configure --program_prefix=testprefix_
2. make install
3. testprefix_monetdbd create /path/to/dbfarm
4. testprefix_monetdbd start /path/to/dbfarm
5. testprefix_monetdb create testdb
6. testprefix_monetdb start testdb
### Actual Results:
starting database 'testdb'... FAILED
start: starting 'testdb' failed: database 'testdb' appears to shut itself down after starting, check monetdbd's logfile (merovingian.log) for possible hints
$ less /path/to/dbfarm/merovignian.log
[...]
2017-05-30 11:33:10 MSG merovingian[16423]: starting database 'testdb', up min/avg/max: 0s/0s/0s, crash average: 0.00 0.00 0.00 (0-0=0)
2017-05-30 11:33:11 MSG testdb[16559]: arguments: /path/to/executable/testprefix_monetdbd --dbpath=/tmp/foo/testdb --set merovingian_uri=mapi:monetdb://zarquon.zarniwoop.org:50000/testdb --set mapi_open=false --set mapi_port=0 --set mapi_usock=/tmp/foo/testdb/.mapi.sock --set monet_vault_key=/tmp/foo/testdb/.vaultkey --set gdk_nr_threads=4 --set max_clients=64 --set sql_optimizer=default_pipe --set monet_daemon=yes
2017-05-30 11:33:11 MSG testdb[16559]: usage: monetdbd [ command [ command-options ] ] <dbfarm>
2017-05-30 11:33:11 MSG testdb[16559]: where command is one of:
2017-05-30 11:33:11 MSG testdb[16559]: create, start, stop, get, set, version or help
2017-05-30 11:33:11 MSG testdb[16559]: use the help command to get help for a particular command
2017-05-30 11:33:11 MSG testdb[16559]: The dbfarm to operate on must always be given to
2017-05-30 11:33:11 MSG testdb[16559]: monetdbd explicitly.
2017-05-30 11:33:11 ERR testdb[16559]: monetdbd: unknown command: --dbpath=/tmp/foo/testdb
2017-05-30 11:33:13 MSG merovingian[16423]: database 'testdb' (16559) has exited with exit status 1
[...]
### Expected Results:
The database should start.
## Comment 25356
Date: 2017-05-30 13:28:23 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [24dd2ff6c348](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=24dd2ff6c348) made by Panagiotis Koutsourakis <kutsurak@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=24dd2ff6c348](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=24dd2ff6c348)
Changeset description:
Change slightly the way the path to the executable for mserver5 is found
This fixes bug #6327
| The daemon does not respect the actual name of the mserver5 executable | https://api.github.com/repos/MonetDB/MonetDB/issues/6327/comments | 0 | 2020-11-30T15:48:50Z | 2024-06-27T12:59:49Z | https://github.com/MonetDB/MonetDB/issues/6327 | 753,590,177 | 6,327 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-25 16:39:54 +0200
From: Richard Hughes <<richard.monetdb>>
To: SQL devs <<bugs-sql>>
Version: 11.23.13 (Jun2016-SP2)
CC: @njnes
Last updated: 2019-04-30 12:36:02 +0200
## Comment 25349
Date: 2017-05-25 16:39:54 +0200
From: Richard Hughes <<richard.monetdb>>
Created attachment 552
Patch v1 (Jun2016)
The attached patch adds a new relational optimizer pass which removes any unioned selects which cannot contribute any rows to the result set. It applies to Jun2016 [7e686221367d;](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=7e686221367d;) if there's interest from other people then I'll rebase it to something newer.
This is similar to, but a superset of, the merge table functionality implemented in rel_merge_table_rewrite(). I looked at using merge tables in my application, but the following problems exist:
1) Bug #6235
2) I'd have to change quite a few bits of my application (not a show-stopper, but it seemed more productive to spend the time in Monet instead)
3) There's no way to make it work with aligned partitions
That last point needs some explanation. My application partitions tables by date, but has multiple associated tables in each date window, call them 'a' and 'b'. It is much more efficient to execute "(a1 join b1) union (a2 join b2) union (a3 join b3)" than "(a1 union a2 union a3) join (b1 union b2 union b3)" because the row count of the intermediate tables is much smaller. Merge tables, however, can only be used to implement the latter. The attached patch works for either.
A minimal example of this patch at work is:
sql>create table sub1 (i int);
sql>create table sub2 (i int);
sql>insert into sub1 values (10),(20);
sql>insert into sub2 values (30),(40);
sql>alter table sub1 set read only;
sql>alter table sub2 set read only;
sql>analyze sys.sub1;
sql>analyze sys.sub2;
sql>plan select i from sub1 where i between 25 and 50 union all select i from sub2 where i between 25 and 50;
+----------------------------------------+
| rel |
+========================================+
| project ( |
| | select ( |
| | | table(sys.sub2) [ sub2.i ] COUNT |
| | ) [ int "25" <= sub2.i <= int "50" ] |
| ) [ sub2.i ] |
+----------------------------------------+
There clearly need to be a whole bunch of tests written for this change, but I'm putting it out there now as an RFC. This is the first time I've messed around in this area of the MonetDB code base, so I'd appreciate it if somebody could tell me if I've done it right.
> Attached file: [remove-union-partitions.patch](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6326_remove-union-partitions.patch_552) (text/plain, 4238 bytes)
> Description: Patch v1 (Jun2016)
## Comment 25350
Date: 2017-05-26 14:24:52 +0200
From: Richard Hughes <<richard.monetdb>>
I should mention why partition elimination is important to me. There are two reasons:
1) Combining indexes. We have a number of queries like "select ... from unionview where t between timestamp '2017-5-14' and timestamp '2017-5-21' and x=42;". In the absence of partition elimination the quickest way to execute this is to filter on 't' first because that gets the row count down fastest and has the best locality in association with other queries; for this reason we actually write the above query as "... and x in (42)" to ensure that the optimizer sees that clause as less selective than the time. With partition elimination we can effectively use indexes on both t and x: out-of-range 't' tables are initially eliminated then the 'x' filter becomes easily the most selective without compromising locality.
2) MAL interpreter performance. I did this to myself due to Bug #3548: most of our plans are around the 20,000-30,000 instruction mark, and 80,000 instructions is not unusual. Benchmarking suggests that the MAL interpreter can dispatch on the order of 20,000-40,000 instructions per second on our hardware (assuming those instructions have no actual work to do) which means that our queries take upwards of a second to return even if they're easy. Eliminating partitions much earlier saves both instruction execution time and MAL generation/optimization time.
The combination of both of these effects allows us to achieve sub-100ms response times in some common cases and generally saves a second or so on every single query.
## Comment 26861
Date: 2019-01-30 12:31:29 +0100
From: MonetDB Mercurial Repository <<hg>>
Changeset [2c50b80d80c7](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=2c50b80d80c7) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=2c50b80d80c7](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=2c50b80d80c7)
Changeset description:
used patch from feature request Bug #6326, added cleanup of the
statistics on changing the access (read only vs read write)
| Eliminate unused UNION tables | https://api.github.com/repos/MonetDB/MonetDB/issues/6326/comments | 0 | 2020-11-30T15:48:47Z | 2024-06-27T12:59:48Z | https://github.com/MonetDB/MonetDB/issues/6326 | 753,590,122 | 6,326 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-23 16:53:29 +0200
From: Richard Hughes <<richard.monetdb>>
To: SQL devs <<bugs-sql>>
Version: 11.23.13 (Jun2016-SP2)
CC: @njnes
Last updated: 2017-08-02 10:21:28 +0200
## Comment 25347
Date: 2017-05-23 16:53:29 +0200
From: Richard Hughes <<richard.monetdb>>
Created attachment 551
Reproduction recipe
Reproduced on both default branch [a84d813e7f57](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=a84d813e7f57) and Jun2016 [7a344a54d712](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=7a344a54d712)
To reproduce:
1) Create a new, blank database named 'newdb'
2) Run the attached (change the IP address at the top if non-loopback)
### Expected results:
Script prints [(0,)]
### Actual results:
Traceback (most recent call last):
File "merge-multiconn.py", line 26, in <module>
print(query(c2, 'select count(*) from mt'))
monetdb.exceptions.OperationalError: MERGE or REPLICA TABLE should have at least one table associated
Changing 'c1' to be autocommit makes it work.
Running the final select on 'c1' instead of 'c2' also works.
Restarting mserver5 makes all future selects work.
> Attached file: [merge-multiconn.py](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_6325_merge-multiconn.py_551) (text/plain, 690 bytes)
> Description: Reproduction recipe
## Comment 25348
Date: 2017-05-24 14:27:13 +0200
From: Richard Hughes <<richard.monetdb>>
This change makes my test pass:
diff -r [a84d813e7f57](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=a84d813e7f57) sql/storage/store.c
--- a/sql/storage/store.c Fri May 19 16:40:54 2017 +0200
+++ b/sql/storage/store.c Wed May 24 13:22:50 2017 +0100
@@ -2392,7 +2392,6 @@
if (tr->parent == gtrans)
ot->columns.nelm = NULL;
}
- /*
if (ot->tables.set) {
for (n = ot->tables.set->h; n; n = n->next) {
sql_table *pt = n->data;
@@ -2404,7 +2403,6 @@
if (tr->parent == gtrans)
ot->tables.nelm = NULL;
}
- */
if (ot->idxs.set) {
for (n = ot->idxs.set->h; n; n = n->next) {
sql_idx *i = n->data;
...but that diff scores a full eleven out of ten on the frightening-o-meter. Why was it commented out? Those lines were commented out when they were first added in [1d95d284c771](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=1d95d284c771)
## Comment 25416
Date: 2017-07-05 22:23:46 +0200
From: @njnes
fixed by keeping a list of members in each merge table (parts) instead directly tables
## Comment 25537
Date: 2017-08-02 10:21:28 +0200
From: @sjoerdmullender
Fixed in Jul2017-SP1 release.
| Merge table unusable in other connections | https://api.github.com/repos/MonetDB/MonetDB/issues/6325/comments | 0 | 2020-11-30T15:48:43Z | 2024-06-27T12:59:47Z | https://github.com/MonetDB/MonetDB/issues/6325 | 753,590,072 | 6,325 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-22 18:50:11 +0200
From: @PedroTadim
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-07-17 16:07:35 +0200
## Comment 25346
Date: 2017-05-22 18:50:11 +0200
From: @PedroTadim
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.36
Build Identifier:
The mserver gets a SIGABRT while performing the SELECT query described bellow. This happened in the latest changeset of Jul2017 branch.
Reproducible: Always
### Steps to Reproduce:
1. CREATE TABLE tab0(pk INTEGER PRIMARY KEY, col0 INTEGER, col1 FLOAT, col2 TEXT, col3 INTEGER, col4 FLOAT, col5 TEXT);
2. SELECT col2 FROM tab0 AS cor0 WHERE + col4 NOT IN ( + 0 + + col1 );
### Actual Results:
The server gets a SIGABRT while performing the query.
### Expected Results:
An empty result set.
Assertion error:
mserver5: rel_bin.c:677: exp_bin: Assertion `0' failed.
Backtrace:
0 0x00007ffff4b6591f in raise () from /lib64/libc.so.6
1 0x00007ffff4b6751a in abort () from /lib64/libc.so.6
2 0x00007ffff4b5dda7 in __assert_fail_base () from /lib64/libc.so.6
3 0x00007ffff4b5de52 in __assert_fail () from /lib64/libc.so.6
4 0x00007fffe8d0fd84 in exp_bin (be=0x7fffb00cb0c0, e=0x7fffb00e3240, left=0x7fffb00e67f0, right=0x7fffb00e6a70, grp=0x0, ext=0x0, cnt=0x0, sel=0x0) at rel_bin.c:677
5 0x00007fffe8d15567 in rel2bin_semijoin (be=0x7fffb00cb0c0, rel=0x7fffb00e32d0, refs=0x7fffb00e58d0) at rel_bin.c:1925
6 0x00007fffe8d2214e in subrel_bin (be=0x7fffb00cb0c0, rel=0x7fffb00e32d0, refs=0x7fffb00e58d0) at rel_bin.c:4773
7 0x00007fffe8d17ac1 in rel2bin_project (be=0x7fffb00cb0c0, rel=0x7fffb00e38b0, refs=0x7fffb00e58d0, topn=0x0) at rel_bin.c:2418
8 0x00007fffe8d22208 in subrel_bin (be=0x7fffb00cb0c0, rel=0x7fffb00e38b0, refs=0x7fffb00e58d0) at rel_bin.c:4789
9 0x00007fffe8d224ff in output_rel_bin (be=0x7fffb00cb0c0, rel=0x7fffb00e38b0) at rel_bin.c:4854
10 0x00007fffe8d3c02e in sql_relation2stmt (be=0x7fffb00cb0c0, r=0x7fffb00e38b0) at sql_gencode.c:472
11 0x00007fffe8d3c38d in backend_dumpstmt (be=0x7fffb00cb0c0, mb=0x7fffb00cb680, r=0x7fffb00e38b0, top=1, add_end=0, query=0x7fffb010cb60 "select col2 from tab0 as cor0 where + col4 not in ( + 0 + + col1 );")
at sql_gencode.c:513
12 0x00007fffe8d02f83 in SQLparser (c=0x7fffea3cc338) at sql_scenario.c:1119
13 0x00007ffff7a1aab7 in runPhase (c=0x7fffea3cc338, phase=1) at mal_scenario.c:517
14 0x00007ffff7a1abfb in runScenarioBody (c=0x7fffea3cc338) at mal_scenario.c:550
15 0x00007ffff7a1ad9e in runScenario (c=0x7fffea3cc338) at mal_scenario.c:579
16 0x00007ffff7a1c844 in MSserveClient (dummy=0x7fffea3cc338) at mal_session.c:462
17 0x00007ffff7a1c334 in MSscheduleClient (command=0x7fffb00008d0 "0", challenge=0x7fffc54f5df0 "HEbVkmmQg", fin=0x7fffb0002990, fout=0x7fffbc002b70, protocol=PROTOCOL_9, blocksize=8190,
compute_column_widths=0) at mal_session.c:353
18 0x00007ffff7abd5b4 in doChallenge (data=0x7fffbc0008d0) at mal_mapi.c:269
19 0x00007ffff76a242b in thread_starter (arg=0x7fffbc004c60) at gdk_system.c:485
20 0x00007ffff4efd6ca in start_thread () from /lib64/libpthread.so.0
21 0x00007ffff4c37f7f in clone () from /lib64/libc.so.6
## Comment 25366
Date: 2017-05-31 17:34:18 +0200
From: @njnes
fixed, the simplify math optimizer failed to pass a column alias
## Comment 25367
Date: 2017-05-31 17:34:49 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [42e07c317685](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=42e07c317685) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=42e07c317685](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=42e07c317685)
Changeset description:
fixes for bug #6324
| Sqlitelogictest crash in a IN query (8th) | https://api.github.com/repos/MonetDB/MonetDB/issues/6324/comments | 0 | 2020-11-30T15:48:41Z | 2024-06-27T12:59:46Z | https://github.com/MonetDB/MonetDB/issues/6324 | 753,590,036 | 6,324 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-22 14:52:50 +0200
From: Richard Hughes <<richard.monetdb>>
To: GDK devs <<bugs-common>>
Version: 11.23.13 (Jun2016-SP2)
Last updated: 2017-07-17 16:07:45 +0200
## Comment 25345
Date: 2017-05-22 14:52:50 +0200
From: Richard Hughes <<richard.monetdb>>
Build is Jun2016 [7a344a54d712](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=7a344a54d712) (but I believe the issue still exists in the default branch).
I've got an mserver5 instance stuck here:
(gdb) bt
0 0x00007f59ce6b0893 in select () at ../sysdeps/unix/syscall-template.S:81
1 0x00007f59cfd1f5f9 in MT_sleep_ms (ms=ms@entry=4) at gdk_posix.c:1175
2 0x00007f59cfc809d9 in incref (lock=1, logical=0, i=13630)
at gdk_bbp.c:2451
3 BBPincref (i=i@entry=13630, logical=logical@entry=0) at gdk_bbp.c:2536
4 0x00007f59d02c06d3 in BATdescriptor (i=13630) at ../../../gdk/gdk.h:2586
5 CMDbbp (ID=0xf937950, NS=0xf937970, TT=0xf937990, CNT=0xf9379b0,
REFCNT=0xf9379d0, LREFCNT=0xf9379f0, LOCATION=0xf937a10, HEAT=0xf937a30,
DIRTY=0xf937a50, STATUS=0xf937a70, KIND=0xf937a90) at bbp.c:437
6 0x00007f59d022c6e4 in malCommandCall (stk=<optimized out>,
pci=<optimized out>) at mal_interpreter.c:165
7 0x00007f59d022d7ab in runMALsequence (cntxt=0x0, mb=0x1812fe30,
startpc=0, stoppc=-1, stk=0xf9378a0, env=0x0, pcicaller=0x0)
at mal_interpreter.c:670
8 0x00007f59d022ec2b in callMAL (cntxt=0x0, cntxt@entry=0x7f59c940d4d0,
mb=0x0, mb@entry=0x1812fe30, env=0xf937930, argv=0xa0, debug=-128 '\200')
at mal_interpreter.c:436
9 0x00007f59c8ca300c in SQLexecutePrepared (c=0x7f59c940d4d0,
be=be@entry=0x6c46410, q=0x5e163c0) at sql_execute.c:370
10 0x00007f59c8ca34ea in SQLengineIntern (c=0x7f59c940d4d0, be=0x6c46410)
at sql_execute.c:435
11 0x00007f59d0248217 in runPhase (phase=4, c=0x7f59c940d4d0)
at mal_scenario.c:531
12 runScenarioBody (c=c@entry=0x7f59c940d4d0) at mal_scenario.c:575
13 0x00007f59d0248d9d in runScenario (c=c@entry=0x7f59c940d4d0)
at mal_scenario.c:595
14 0x00007f59d02492e0 in MSserveClient (dummy=dummy@entry=0x7f59c940d4d0)
at mal_session.c:457
15 0x00007f59d0249946 in MSscheduleClient (
command=command@entry=0x1fe95db0 "",
challenge=challenge@entry=0x7f535dc1ce80 "BjJUfUrNT", fin=0x8fed910,
fout=fout@entry=0x7f59b81db8c0) at mal_session.c:342
16 0x00007f59d02cab96 in doChallenge (data=<optimized out>) at mal_mapi.c:205
17 0x00007f59cfd1e0af in thread_starter (arg=<optimized out>)
at gdk_system.c:485
18 0x00007f59ce982064 in start_thread (arg=0x7f535dc1d700)
at pthread_create.c:309
19 0x00007f59ce6b762d in clone ()
at ../sysdeps/unix/sysv/linux/x86_64/clone.S:111
(gdb) p BBP[0][13630]
$1 = {cache = {0x0, 0x0}, logical = {0x6a59990 "tmp_32476",
0x16a6cc70 "tmpr_32476"}, bak = {0x6a59990 "tmp_32476", 0x0}, next = {
59751, 0}, desc = 0x0, physical = 0x5ece700 "03/24/32476", options = 0x0,
refs = 1, lrefs = 0, lastused = 272413352, status = 2048}
This is the first (and so far only) time this has happened. I don't have logging of what queries led up to the incident.
The only place I can find to set BBPrec::status=2048 is BBPinsert(), so my hypothesis is:
BATnewstorage()
calls BATcreatedesc()
calls BBPinsert()
either HEAPalloc() or ATOMheap() fails, so BATnewstorage() returns before calling BBPcacheit().
Similar bailouts seem possible in BATcreatedesc() and VIEWcreate_() (and possibly some other places that I haven't found).
This bug, therefore, is a request that you check your error handling in these locations to ensure that the BBP is left in a stable state upon errors.
There is nothing in merovingian.log (other than normal client connection logging) anywhere near the time the problem started, so I'm not entirely happy with the above hypothesis; if you can come up with a better theory (or find a likely error path which logs nothing) then I'd be grateful. If you'd like any more information out of my core dump then let me know.
## Comment 25351
Date: 2017-05-29 10:33:55 +0200
From: @sjoerdmullender
If you still have this in a debugger, can you share the output of
thread apply all bt
## Comment 25352
Date: 2017-05-29 10:57:10 +0200
From: @sjoerdmullender
(In reply to Sjoerd Mullender from comment 1)
> If you still have this in a debugger, can you share the output of
>
> thread apply all bt
Belay this request. After studying the situation (and your theory) a bit more, I think you're on to something.
## Comment 25354
Date: 2017-05-29 17:12:32 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [e0f18665e346](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=e0f18665e346) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=e0f18665e346](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=e0f18665e346)
Changeset description:
When failing BAT creation after a successful BBPinsert(), we need to BBPclear().
This fixes bug #6323.
| Deadlock calling sys.bbp() | https://api.github.com/repos/MonetDB/MonetDB/issues/6323/comments | 0 | 2020-11-30T15:48:37Z | 2024-06-27T12:59:45Z | https://github.com/MonetDB/MonetDB/issues/6323 | 753,589,984 | 6,323 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-18 15:55:22 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: akkaran046, @njnes
Last updated: 2018-12-31 07:04:54 +0100
## Comment 25343
Date: 2017-05-18 15:55:22 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:52.0) Gecko/20100101 Firefox/52.0
Build Identifier:
Removing the LIMIT clause makes it work
Reproducible: Always
### Steps to Reproduce:
select
subq_0.c3 as c2
from
(select
sample_0.proj4text as c3,
sample_0.auth_name as c7
from
sys.spatial_ref_sys as sample_0
where true
limit 134) as subq_0
where (true)
or ((select pc from sys.tracelog)
is not NULL);
### Actual Results:
0 0x00007ffff496d765 in raise () from /lib64/libc.so.6
1 0x00007ffff496f36a in abort () from /lib64/libc.so.6
2 0x00007ffff4965f97 in __assert_fail_base () from /lib64/libc.so.6
3 0x00007ffff4966042 in __assert_fail () from /lib64/libc.so.6
4 0x00007ffff76b975f in BATgroup_internal (groups=0x7fffd441c728,
extents=0x7fffd441c730, histo=0x7fffd441c738, b=0x7fffb0003140, s=0x0,
g=0x7fffb4008960, e=0x0, h=0x0, subsorted=0)
at /export/scratch1/home/mk/default//package/gdk/gdk_group.c:480
5 0x00007ffff76cf1bf in BATgroup (groups=0x7fffd441c728, extents=0x7fffd441c730,
histo=0x7fffd441c738, b=0x7fffb0003140, s=0x0, g=0x7fffb4008960, e=0x0, h=0x0)
at /export/scratch1/home/mk/default//package/gdk/gdk_group.c:1101
6 0x00007ffff7a9df24 in GRPsubgroup5 (ngid=0x1cc88a0, next=0x1cc88c0,
nhis=0x1cc88e0, bid=0x1cc8820, sid=0x0, gid=0x1cc8840, eid=0x0, hid=0x0)
at /export/scratch1/home/mk/default//package/monetdb5/modules/kernel/group.c:40
7 0x00007ffff7a9e16f in GRPsubgroup2 (ngid=0x1cc88a0, next=0x1cc88c0,
nhis=0x1cc88e0, bid=0x1cc8820, gid=0x1cc8840)
at /export/scratch1/home/mk/default//package/monetdb5/modules/kernel/group.c:75
8 0x00007ffff79f38ca in malCommandCall (stk=0x1cc7ec0, pci=0x1cc5fb0)
at /export/scratch1/home/mk/default//package/monetdb5/mal/mal_interpreter.c:93
9 0x00007ffff79f6585 in runMALsequence (cntxt=0x7fffef926020, mb=0x1cb5300,
startpc=37, stoppc=38, stk=0x1cc7ec0, env=0x0, pcicaller=0x0)
at /export/scratch1/home/mk/default//package/monetdb5/mal/mal_interpreter.c:668
10 0x00007ffff79fb869 in DFLOWworker (T=0x7ffff7daa680 <workers+384>)
at /export/scratch1/home/mk/default//package/monetdb5/mal/mal_dataflow.c:403
11 0x00007ffff4d025ca in start_thread () from /lib64/libpthread.so.0
12 0x00007ffff4a3c0ed in clone () from /lib64/libc.so.6
## Comment 25363
Date: 2017-05-31 17:00:10 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [99f4886d5b45](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=99f4886d5b45) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=99f4886d5b45](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=99f4886d5b45)
Changeset description:
add test for bug #6322
## Comment 25374
Date: 2017-06-07 13:09:47 +0200
From: @njnes
fixed a bug in the push select optimizer
## Comment 25403
Date: 2017-06-29 16:52:55 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [0592a67375ad](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=0592a67375ad) made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=0592a67375ad](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=0592a67375ad)
Changeset description:
Exclude test push_subslice.Bug-6322 when it does not HAVE_GEOM
as the test query depends on sys.spatial_ref_sys (which is created in 40_geom.sql)
## Comment 25892
Date: 2017-11-16 15:31:23 +0100
From: MonetDB Mercurial Repository <<hg>>
Changeset [eb7ad3a6176d](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=eb7ad3a6176d) made by Martin van Dinther <martin.van.dinther@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=eb7ad3a6176d](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=eb7ad3a6176d)
Changeset description:
Corrected file name of limit.bug-6322.sql into limit.Bug-6322.sql as defined in All file
Added dependency condition HAVE_GEOM
Reduced output to 14 rows (was 134)
Added missing test output files
## Comment 26762
Date: 2018-12-31 07:04:54 +0100
From: martin garix <<akkaran046>>
google.com
| Crash on disjunction with LIMIT (sqlsmith) | https://api.github.com/repos/MonetDB/MonetDB/issues/6322/comments | 0 | 2020-11-30T15:48:35Z | 2024-06-27T12:59:44Z | https://github.com/MonetDB/MonetDB/issues/6322 | 753,589,941 | 6,322 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-16 16:16:32 +0200
From: @swingbit
To: GDK devs <<bugs-common>>
Version: 11.25.21 (Dec2016-SP4)
Last updated: 2017-10-26 14:01:31 +0200
## Comment 25342
Date: 2017-05-16 16:16:32 +0200
From: @swingbit
User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/57.0.2987.98 Safari/537.36
Build Identifier:
This is very similar to Bug #6317 (fixed in the meantime).
It is again about a slow aggregation on a two-column table, but it extends it with the join that generates the table to aggregate upon, and one more column.
Data to bootstrap the example can be found here:
https://drive.google.com/open?id=0BwkuZ5tb4WEnaHdYR2dyOWtpdGM
With that data:
-- Generate the table to aggregate upon (result is 280M tuples):
sql>CREATE TABLE tmp as SELECT ap.a1 as a1, bp.a1 AS a3, ap.prob * bp.prob as prob FROM ap, bp WHERE ap.a2 = bp.a2 WITH DATA;
operation successful (26.3s)
-- Aggregate, summing the prob column:
sql>CREATE TABLE t1 AS SELECT a1, a3, sum(prob) FROM tmp GROUP BY a1, a3 WITH DATA;
operation successful (35.0s)
No problem here.
If I rewrite the same query with the join as a subselect:
sql>CREATE TABLE t1 AS SELECT a1, a3, sum(prob) FROM (SELECT ap.a1 as a1, bp.a1 AS a3, ap.prob * bp.prob as prob FROM ap, bp WHERE ap.a2 = bp.a2) AS tmp GROUP BY a1, a3 WITH DATA;
this takes again forever (that is, I needed to stop it).
The interesting fact is that it is again (as in Bug #6317) stuck in group.subgroupdone().
Also interesting, the same query with only grouping but without the actual aggregation sum(prob) succeeds with no problem:
sql>CREATE TABLE t1 AS SELECT a1, a3 FROM (SELECT ap.a1 as a1, bp.a1 AS a3, ap.prob * bp.prob as prob FROM ap, bp WHERE ap.a2 = bp.a2) AS tmp GROUP BY a1, a3 WITH DATA;
operation successful (47.1s)
Reproducible: Always
$ mserver5 --version
MonetDB 5 server v11.25.22 (64-bit, 128-bit integers)
This is an unreleased version
Copyright (c) 1993-July 2008 CWI
Copyright (c) August 2008-2017 MonetDB B.V., all rights reserved
Visit http://www.monetdb.org/ for further information
Found 15.6GiB available memory, 8 available cpu cores
Libraries:
libpcre: 8.40 2017-01-11 (compiled with 8.40)
openssl: OpenSSL 1.0.2j 26 Sep 2016 (compiled with OpenSSL 1.0.2j-fips 26 Sep 2016)
libxml2: 2.9.3 (compiled with 2.9.3)
Compiled by: roberto@photon.hq.spinque.com (x86_64-unknown-linux-gnu)
Compilation: gcc -O3 -fomit-frame-pointer -pipe -DDISABLE_SPINQUE -Werror -Wall -Wextra -W -Werror-implicit-function-declaration -Wpointer-arith -Wdeclaration-after-statement -Wundef -Wformat=2 -Wno-format-nonliteral -Winit-self -Winvalid-pch -Wmissing-declarations -Wmissing-format-attribute -Wmissing-prototypes -Wold-style-definition -Wpacked -Wunknown-pragmas -Wvariadic-macros -fstack-protector-all -Wstack-protector -Wpacked-bitfield-compat -Wsync-nand -Wjump-misses-init -Wmissing-include-dirs -Wlogical-op -Wunreachable-code -D_FORTIFY_SOURCE=2
Linking : /usr/bin/ld -m elf_x86_64
## Comment 25353
Date: 2017-05-29 16:51:17 +0200
From: @sjoerdmullender
On my desktop with 16 GB of memory I can reproduce the problem. On my laptop with 32 GB of memory, both queries run reasonably fast.
It seems the problem is that for some reason the query with the sum requires more memory than the one without the sum. I don't understand why, though. The plans of both queries are identical up to and including the call to group.subgroupdone, so you'd expect basically the same behavior for both up to this point. But that's not what we're getting. The two plans are executed differently, possibly because of what follows later in the plan.
The query without sum stays within 13 GB or so in ther VIRT column of top, whereas the query with sum grows to 18 GB. The sum query therefore trashes the disk and thus becomes very slow.
## Comment 25612
Date: 2017-09-05 11:32:22 +0200
From: @sjoerdmullender
With the current Jul2017 code, the problem seems to have been fixed. All three variations run in about a minute:
sql>CREATE TABLE t1 AS SELECT a1, a3, sum(prob) FROM tmp GROUP BY a1, a3 WITH DATA;
operation successful (1m 8s)
sql>drop table t1;
operation successful (19.492ms)
sql>CREATE TABLE t1 AS SELECT a1, a3, sum(prob) FROM (SELECT ap.a1 as a1, bp.a1 AS a3, ap.prob * bp.prob as prob FROM ap, bp WHERE ap.a2 = bp.a2) AS tmp GROUP BY a1, a3 WITH DATA;
operation successful (1m 14s)
sql>drop table t1;
operation successful (17.256ms)
sql>CREATE TABLE t1 AS SELECT a1, a3 FROM (SELECT ap.a1 as a1, bp.a1 AS a3, ap.prob * bp.prob as prob FROM ap, bp WHERE ap.a2 = bp.a2) AS tmp GROUP BY a1, a3 WITH DATA;
operation successful (53.0s)
This was probably fixed by changeset [90a3b702ff32](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=90a3b702ff32) and the tweaks to the BATgroup code that followed.
Roberto, can you confirm?
## Comment 25616
Date: 2017-09-06 09:28:48 +0200
From: @swingbit
It seems fixed indeed. Thanks.
| Two-column aggregation on join result extremely slow. | https://api.github.com/repos/MonetDB/MonetDB/issues/6321/comments | 0 | 2020-11-30T15:48:31Z | 2024-06-27T12:59:43Z | https://github.com/MonetDB/MonetDB/issues/6321 | 753,589,897 | 6,321 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-16 10:15:01 +0200
From: @kutsurak
To: Merovingian devs <<bugs-merovingian>>
Version: 11.25.21 (Dec2016-SP4)
Last updated: 2017-06-12 15:14:45 +0200
## Comment 25333
Date: 2017-05-16 10:15:01 +0200
From: @kutsurak
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:53.0) Gecko/20100101 Firefox/53.0
Build Identifier:
The daemon uses the file .merovignian_properties in the db directory to set the arguments passed to the mserver5 process. The daemon ignores all the entries in that file except the last one that passes multiple times to the server.
Reproducible: Always
### Steps to Reproduce:
Assuming bash:
1. monetdbd create /tmp/dbfarm
2. monetdbd start /tmp/dbfarm
3. monetdb create db
4. for i in {1..5}; do monetdb set prop$i=val$i db; done
5. monetdb start db
6. ps -ww -fC mserver5
### Actual Results:
mserver5
--dbpath=/tmp/dbfarm/db
--set merovingian_uri mapi:monetdb://zarquon.zarniwoop.org:50000/db
--set mapi_open false
--set mapi_port 0
--set mapi_usock /tmp/stable_farm/foo/.mapi.sock
--set monet_vault_key /tmp/dbfarm/db/.vaultkey
--set gdk_nr_threads 4
--set max_clients 64
--set sql_optimizer default_pipe
--set prop5 val5
--set prop5 val5
--set prop5 val5
--set prop5 val5
--set prop5 val5
--set monet_daemon yes
### Expected Results:
mserver5
--dbpath=/tmp/dbfarm/db
--set merovingian_uri mapi:monetdb://zarquon.zarniwoop.org:50000/db
--set mapi_open false
--set mapi_port 0
--set mapi_usock /tmp/stable_farm/foo/.mapi.sock
--set monet_vault_key /tmp/dbfarm/db/.vaultkey
--set gdk_nr_threads 4
--set max_clients 64
--set sql_optimizer default_pipe
--set prop1 val1
--set prop2 val2
--set prop3 val3
--set prop4 val4
--set prop5 val5
--set monet_daemon yes
## Comment 25335
Date: 2017-05-16 11:17:49 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [f088e35112c5](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=f088e35112c5) made by Panagiotis Koutsourakis <kutsurak@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=f088e35112c5](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=f088e35112c5)
Changeset description:
Make sure we are using ALL properties we just read from the file
This fixes bug #6320
| The daemon ignores all but the last entry in the .merovignian_properties file when spawning mserver5 | https://api.github.com/repos/MonetDB/MonetDB/issues/6320/comments | 0 | 2020-11-30T15:48:28Z | 2024-06-27T12:59:42Z | https://github.com/MonetDB/MonetDB/issues/6320 | 753,589,855 | 6,320 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-16 09:59:24 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2017-07-17 16:07:28 +0200
## Comment 25332
Date: 2017-05-16 09:59:24 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:52.0) Gecko/20100101 Firefox/52.0
Build Identifier:
The following is causing a crash
Reproducible: Always
### Steps to Reproduce:
select
subq_0
from
sys.keys ,
lateral (select ref_2.id
from tmp.keys
where ((select role_id from sys.user_role) is NULL)
or (false)) as subq_0;
### Actual Results:
Thread 1 "mserver5" received signal SIGSEGV, Segmentation fault.
0x00007fffee055026 in rel_column_ref (sql=0x1bb9630, rel=0x7fffffffccb0,
column_r=0x1c90410, f=2)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:1065
1065 while(gb->l && !is_groupby(gb->op))
(gdb) where
0 0x00007fffee055026 in rel_column_ref (sql=0x1bb9630, rel=0x7fffffffccb0,
column_r=0x1c90410, f=2)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:1065
1 0x00007fffee064808 in rel_value_exp2 (sql=0x1bb9630, rel=0x7fffffffccb0,
se=0x1c90410, f=2, ek=..., is_last=0x7fffffffcbcc)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4345
2 0x00007fffee0652c5 in rel_value_exp (sql=0x1bb9630, rel=0x7fffffffccb0,
se=0x1c90410, f=2, ek=...)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4489
3 0x00007fffee0653aa in column_exp (sql=0x1bb9630, rel=0x7fffffffccb0,
column_e=0x1c90490, f=2)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4502
4 0x00007fffee065611 in rel_column_exp (sql=0x1bb9630, rel=0x7fffffffccb0,
column_e=0x1c90490, f=2)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4546
5 0x00007fffee0661d8 in rel_select_exp (sql=0x1bb9630, rel=0x1c93de0,
sn=0x1c90c10, ek=...)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4741
6 0x00007fffee067275 in rel_query (sql=0x1bb9630, rel=0x0, sq=0x1c90c10,
toplevel=1, ek=..., apply=8)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4990
7 0x00007fffee068b55 in rel_subquery (sql=0x1bb9630, rel=0x0, sq=0x1c90c10,
ek=..., apply=8)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:5355
8 0x00007fffee051391 in rel_subquery_optname (sql=0x1bb9630, rel=0x0,
query=0x1c90c10)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:259
9 0x00007fffee0547ac in table_ref (sql=0x1bb9630, rel=0x0, tableref=0x1c90c10,
lateral=0)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:937
10 0x00007fffee066dcc in rel_query (sql=0x1bb9630, rel=0x0, sq=0x1c90e00,
toplevel=1, ek=..., apply=8)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4920
11 0x00007fffee068b55 in rel_subquery (sql=0x1bb9630, rel=0x0, sq=0x1c90e00,
ek=..., apply=8)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:5355
12 0x00007fffee068c66 in rel_selects (sql=0x1bb9630, s=0x1c90e00)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:5376
13 0x00007fffee04fffe in rel_semantic (sql=0x1bb9630, s=0x1c90e00)
## Comment 25334
Date: 2017-05-16 11:11:41 +0200
From: @mlkersten
select
subq_0
from
sys.keys ,
lateral (select ref_2.id
from tmp.keys as ref_2
where ((select role_id from sys.user_role) is NULL)
or (false)) as subq_0;
## Comment 25362
Date: 2017-05-31 16:57:39 +0200
From: @njnes
solved, ie added more protection when we lookup the group by expression needed for the proper error message in column_ref.
| Server crash on LATERAL (sqlsmith) | https://api.github.com/repos/MonetDB/MonetDB/issues/6319/comments | 0 | 2020-11-30T15:48:25Z | 2024-06-27T12:59:41Z | https://github.com/MonetDB/MonetDB/issues/6319 | 753,589,812 | 6,319 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-16 09:58:31 +0200
From: @kutsurak
To: Merovingian devs <<bugs-merovingian>>
Version: 11.25.21 (Dec2016-SP4)
Last updated: 2017-06-12 15:14:47 +0200
## Comment 25331
Date: 2017-05-16 09:58:31 +0200
From: @kutsurak
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:53.0) Gecko/20100101 Firefox/53.0
Build Identifier:
When attempting to start a database that has more than 42 properties in the .merovignian_properties file, the daemon crashes.
Reproducible: Always
### Steps to Reproduce:
Assuming bash:
1. monetdbd create /tmp/dbfarm
2. monetdbd start /tmp/dbfarm
3. monetdb create db
4. for i in {1..42}; do monetdb set prop$i=val$i foo; done
5. monetdb start db
### Actual Results:
starting database 'db'... FAILED
start: incomplete response from monetdbd
(monetdbd has actually crashed)
### Expected Results:
The database should start
## Comment 25336
Date: 2017-05-16 11:17:56 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [8cb80fc92ece](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=8cb80fc92ece) made by Panagiotis Koutsourakis <kutsurak@monetdbsolutions.com> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=8cb80fc92ece](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=8cb80fc92ece)
Changeset description:
Make sure we don't touch memory that doesn't belong to us
The list argument of the function readConfFileFull has been allocated
statically, and now has length of 64. The convention in the existing code is
that the last entry is a sentinel, so in total we can use 63 entries. Make sure
we don't write past the last entry.
This fixes bug #6318
| Daemon crashes if .merovignian_properties for a database contains more than 42 entries | https://api.github.com/repos/MonetDB/MonetDB/issues/6318/comments | 0 | 2020-11-30T15:48:22Z | 2024-06-27T12:59:40Z | https://github.com/MonetDB/MonetDB/issues/6318 | 753,589,763 | 6,318 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-15 21:29:41 +0200
From: @swingbit
To: GDK devs <<bugs-common>>
Version: 11.25.21 (Dec2016-SP4)
Last updated: 2017-08-15 16:10:23 +0200
## Comment 25328
Date: 2017-05-15 21:29:41 +0200
From: @swingbit
User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.36
Build Identifier:
I have a 280M tuples, two-column table of integers, with the following details:
sql>\d bigaggr
CREATE TABLE "sys"."bigaggr" (
"a1" INTEGER,
"a2" INTEGER
);
sql>select count(*) from bigaggr;
+-----------+
| L3 |
+===========+
| 279401374 |
+-----------+
1 tuple (30.333ms)
sql>select count(distinct a1) from bigaggr;
+--------+
| L3 |
+========+
| 153033 |
+--------+
1 tuple (804.289ms)
sql>select count(distinct a2) from bigaggr;
+-------+
| L3 |
+=======+
| 41365 |
+-------+
1 tuple (762.654ms)
On this table, I perform the following aggregation:
CREATE TABLE x1 AS
SELECT a1, a2
FROM bigaggr
GROUP BY a1,a2
WITH DATA;
This runs for hours at 100% CPU, about 5-6 GB resident memory (13GB virtual), until it eats up about 80GB of disk space and finally crashes.
Stethoscope shows that group.subgroup() correctly delivers 153033 groups in just a few seconds.
Then, groupsubgroupdone() runs forever.
Swapping the two columns in the GROUP BY clause does reflect in the MAL plan, but shows the same behaviour: the secondary aggregation fails.
Notice that the same query where GROUP BY is replaced with ORDER BY succeeds in just 50 seconds.
The dump of the table is available here (672MB compressed):
https://drive.google.com/open?id=0BwkuZ5tb4WEnOEdKRk85eXpXMXc
N.B.: sometimes, with many small groups, mitosis can hurt aggregations badly, but it is not to blame in this case. I have tried with no_mitosis_pipe, with the same results.
Reproducible: Always
$ mserver5 --version
MonetDB 5 server v11.25.22 (64-bit, 128-bit integers)
This is an unreleased version
Copyright (c) 1993-July 2008 CWI
Copyright (c) August 2008-2017 MonetDB B.V., all rights reserved
Visit http://www.monetdb.org/ for further information
Found 15.6GiB available memory, 8 available cpu cores
Libraries:
libpcre: 8.40 2017-01-11 (compiled with 8.40)
openssl: OpenSSL 1.0.2j 26 Sep 2016 (compiled with OpenSSL 1.0.2j-fips 26 Sep 2016)
libxml2: 2.9.3 (compiled with 2.9.3)
Compiled by: roberto@photon.hq.spinque.com (x86_64-unknown-linux-gnu)
Compilation: gcc -O3 -fomit-frame-pointer -pipe -DDISABLE_SPINQUE -Werror -Wall -Wextra -W -Werror-implicit-function-declaration -Wpointer-arith -Wdeclaration-after-statement -Wundef -Wformat=2 -Wno-format-nonliteral -Winit-self -Winvalid-pch -Wmissing-declarations -Wmissing-format-attribute -Wmissing-prototypes -Wold-style-definition -Wpacked -Wunknown-pragmas -Wvariadic-macros -fstack-protector-all -Wstack-protector -Wpacked-bitfield-compat -Wsync-nand -Wjump-misses-init -Wmissing-include-dirs -Wlogical-op -Wunreachable-code -D_FORTIFY_SOURCE=2
Linking : /usr/bin/ld -m elf_x86_64
## Comment 25330
Date: 2017-05-16 00:32:28 +0200
From: @swingbit
A correction: swapping the two columns in the GROUP BY clause, thus executing
CREATE TABLE x1 AS
SELECT a1, a2
FROM bigaggr
GROUP BY a2,a1
WITH DATA;
does complete, in 85 minutes!
## Comment 25337
Date: 2017-05-16 13:37:44 +0200
From: @sjoerdmullender
I can see that the second call to subgroup does indeed take (just about) forever, but I don't see a growth in the database. Growth is anyway unexpected since there is no join involved which might explode the size of the datebase.
As to the size of the server, that is to be expected. The inputs of the second subgroup are two 280M row columns of 4 and 8 bytes wide respectively (the column being grouped (int) and the result of the first subgroup (oid)). The result of the call is a new group column (280M rows of oid) plus an extent and histogram (both 8 byte wide and with one row per group). There is also a temporary hash structure with is roughly 280M * 2 * 4 bytes.
With about 260M groups in the result, this gives about 12G of memory being involved in this call.
A fix for the speed issue is on its way.
## Comment 25338
Date: 2017-05-16 13:39:55 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [f5629e7c3ee7](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=f5629e7c3ee7) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=f5629e7c3ee7](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=f5629e7c3ee7)
Changeset description:
By doing what we say, subgrouping can be a lot faster.
This should fix bug #6317.
## Comment 25339
Date: 2017-05-16 14:16:16 +0200
From: @swingbit
Yep, it now completes in 22 seconds. Thanks!
I'll try to have a look at why I saw that database growth. But this is fixed.
## Comment 25340
Date: 2017-05-16 14:49:55 +0200
From: @swingbit
For completeness, the title of this bug report is perhaps misleading.
I did experience a crash due to database growth, but it was caused by something else (which I think I've just found and am going to report soon).
So this was just extremely slow.
## Comment 25341
Date: 2017-05-16 14:52:27 +0200
From: @sjoerdmullender
Can't you edit the title? Right next to the title there should be an "edit" link.
## Comment 25579
Date: 2017-08-15 16:10:23 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [90a3b702ff32](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=90a3b702ff32) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=90a3b702ff32](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=90a3b702ff32)
Changeset description:
Use a clever trick to reduce the overlap of value and group ID in subgroup.
See the comment.
This fix keeps the performance of query 16 (see changeset
cdf01e261bc6) and brings back the performance for bug #6317.
| Two-column integer aggregation extremely slow | https://api.github.com/repos/MonetDB/MonetDB/issues/6317/comments | 0 | 2020-11-30T15:48:19Z | 2024-06-27T12:59:39Z | https://github.com/MonetDB/MonetDB/issues/6317 | 753,589,694 | 6,317 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-13 23:04:14 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: 11.25.21 (Dec2016-SP4)
CC: @njnes
Last updated: 2017-06-12 15:14:35 +0200
## Comment 25318
Date: 2017-05-13 23:04:14 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:53.0) Gecko/20100101 Firefox/53.0
Build Identifier:
The error reported is '!illegal argument'
Unclear why this message is raised.
Reproducible: Always
### Steps to Reproduce:
select role_id,
cast(coalesce(role_id, login_id) as int)
from sys.user_role
limit 145;
## Comment 25323
Date: 2017-05-14 21:55:10 +0200
From: @njnes
Seems to work now on Dec2016
| Coalesc and limit error (sqlsmith) | https://api.github.com/repos/MonetDB/MonetDB/issues/6316/comments | 0 | 2020-11-30T15:48:16Z | 2024-06-27T12:59:38Z | https://github.com/MonetDB/MonetDB/issues/6316 | 753,589,662 | 6,316 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-13 22:47:03 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: 11.25.21 (Dec2016-SP4)
CC: @njnes
Last updated: 2017-06-12 15:14:41 +0200
## Comment 25317
Date: 2017-05-13 22:47:03 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:53.0) Gecko/20100101 Firefox/53.0
Build Identifier:
The implementation of EXIST is incomplete
Reproducible: Always
### Steps to Reproduce:
select
case when EXISTS ( select hashes from sys.tablestoragemodel )
then (true)
else (false)
end
from sys.tables;
## Comment 25319
Date: 2017-05-13 23:16:16 +0200
From: @mlkersten
Also reported on this query
insert into sys.privileges values (
1,
31,
87,
case when EXISTS (
select
(select io from sys.querylog_calls)
as c0
from
sys.rejects as ref_0
where ref_0.fldid is NULL) then 56 else 56 end
,
case when (select dependency_type_name from sys.dependency_types)
is not NULL then 7 else 7 end
)
ERROR = !exist operator on type int missing
## Comment 25322
Date: 2017-05-14 21:53:18 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [2a80648779e3](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=2a80648779e3) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=2a80648779e3](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=2a80648779e3)
Changeset description:
fix bug #6315, ie add missing exists functions.
## Comment 25324
Date: 2017-05-14 21:55:56 +0200
From: @njnes
fixed in Dec2016. added missing exists functions
| Exist operator on type bigint missing (sqlsmith) | https://api.github.com/repos/MonetDB/MonetDB/issues/6315/comments | 0 | 2020-11-30T15:48:12Z | 2024-06-27T12:59:37Z | https://github.com/MonetDB/MonetDB/issues/6315 | 753,589,603 | 6,315 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-13 19:24:35 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: 11.25.21 (Dec2016-SP4)
CC: @njnes
Last updated: 2017-06-12 15:14:38 +0200
## Comment 25316
Date: 2017-05-13 19:24:35 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:53.0) Gecko/20100101 Firefox/53.0
Build Identifier:
A crash occurred in the query shown below.
Perhaps related to the name resolution in laterals.
Test file added to the testweb.
(gdb) print *sql
$1 = {errstr = '\000' <repeats 8191 times>, sa = 0x1c7d4d0, qc = 0x1afeee0, clientid = 0, scanner = {rs = 0x60e8e0, ws = 0x60a160, log = 0x0, yynext = 0, yylast = 365,
yysval = 846, yyval = 388, yycur = 847, yybak = 10 '\n', as = 0, key = 578, started = 1, mode = LINE_N, schema = 0x1c91a40 "sys", errstr = 0x0}, params = 0x0,
forward = 0x0, vars = 0x1aed420, topvars = 10, sizevars = 32, frame = 1, use_views = 1, args = 0x1b08740, argc = 1, argmax = 32, sym = 0x1c923a0, no_mitosis = 0,
user_id = 3, role_id = 3, last_id = -1, rowcnt = -1, timezone = 0, cache = 100, caching = 100, history = 0, reply_size = -1, sizeheader = 0, debug = 0, emode = 0 '\000',
emod = 0 '\000', session = 0x1b089c0, type = 0, pushdown = 1, label = 10, cascade_action = 0x0, opt_stats = {0, 0, 0, 0, 0, 0, 0, 0}, result_id = 0, results = 0x0}
(gdb) print *rel
$2 = (sql_rel *) 0x1c95680
(gdb) print *column_r
$3 = {token = 47, type = type_list, data = {i_val = 29949856, l_val = 29949856, sval = 0x1c8ffa0 "\300\377\310\001", lval = 0x1c8ffa0, sym = 0x1c8ffa0, typeval = {
type = 0x1c8ffa0, digits = 3183328701, scale = 3183328701}}}
0 0x00007fffee295ede in rel_column_ref (sql=0x1bb8ea0, rel=0x7fffffffc5e0, column_r=0x1c90000, f=2)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:1065
1 0x00007fffee2a56c0 in rel_value_exp2 (sql=0x1bb8ea0, rel=0x7fffffffc5e0, se=0x1c90000, f=2, ek=..., is_last=0x7fffffffc4fc)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4345
2 0x00007fffee2a617d in rel_value_exp (sql=0x1bb8ea0, rel=0x7fffffffc5e0, se=0x1c90000, f=2, ek=...)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4489
3 0x00007fffee2a6262 in column_exp (sql=0x1bb8ea0, rel=0x7fffffffc5e0, column_e=0x1c90090, f=2) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4502
4 0x00007fffee2a64c9 in rel_column_exp (sql=0x1bb8ea0, rel=0x7fffffffc5e0, column_e=0x1c90090, f=2)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4546
5 0x00007fffee2a7090 in rel_select_exp (sql=0x1bb8ea0, rel=0x1c95680, sn=0x1c908a0, ek=...) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4741
6 0x00007fffee2a80f4 in rel_query (sql=0x1bb8ea0, rel=0x0, sq=0x1c908a0, toplevel=1, ek=..., apply=8)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4980
7 0x00007fffee2a99d4 in rel_subquery (sql=0x1bb8ea0, rel=0x0, sq=0x1c908a0, ek=..., apply=8) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:5345
8 0x00007fffee292249 in rel_subquery_optname (sql=0x1bb8ea0, rel=0x0, query=0x1c908a0) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:259
9 0x00007fffee295664 in table_ref (sql=0x1bb8ea0, rel=0x0, tableref=0x1c908a0, lateral=0) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:937
10 0x00007fffee2a7c84 in rel_query (sql=0x1bb8ea0, rel=0x0, sq=0x1c90e80, toplevel=1, ek=..., apply=8)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4919
11 0x00007fffee2a99d4 in rel_subquery (sql=0x1bb8ea0, rel=0x0, sq=0x1c90e80, ek=..., apply=8) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:5345
12 0x00007fffee292249 in rel_subquery_optname (sql=0x1bb8ea0, rel=0x0, query=0x1c90e80) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:259
13 0x00007fffee295664 in table_ref (sql=0x1bb8ea0, rel=0x0, tableref=0x1c90e80, lateral=0) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:937
14 0x00007fffee2a8a9c in rel_joinquery_ (sql=0x1bb8ea0, rel=0x0, tab1=0x1c90e80, natural=0, jointype=jt_inner, tab2=0x1c91100, js=0x1c912f0)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:5130
15 0x00007fffee2a959e in rel_joinquery (sql=0x1bb8ea0, rel=0x0, q=0x1c913d0) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:5264
16 0x00007fffee292703 in query_exp_optname (sql=0x1bb8ea0, r=0x0, q=0x1c913d0) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:339
17 0x00007fffee295686 in table_ref (sql=0x1bb8ea0, rel=0x0, tableref=0x1c913d0, lateral=0) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:939
18 0x00007fffee2a8a9c in rel_joinquery_ (sql=0x1bb8ea0, rel=0x0, tab1=0x1c913d0, natural=0, jointype=jt_inner, tab2=0x1c91e90, js=0x1c920b0)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:5130
19 0x00007fffee2a959e in rel_joinquery (sql=0x1bb8ea0, rel=0x0, q=0x1c92190) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:5264
20 0x00007fffee292703 in query_exp_optname (sql=0x1bb8ea0, r=0x0, q=0x1c92190) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:339
21 0x00007fffee295686 in table_ref (sql=0x1bb8ea0, rel=0x0, tableref=0x1c92190, lateral=0) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:939
22 0x00007fffee2a7c84 in rel_query (sql=0x1bb8ea0, rel=0x0, sq=0x1c923a0, toplevel=1, ek=..., apply=8)
at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:4919
23 0x00007fffee2a99d4 in rel_subquery (sql=0x1bb8ea0, rel=0x0, sq=0x1c923a0, ek=..., apply=8) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:5345
24 0x00007fffee2a9ae5 in rel_selects (sql=0x1bb8ea0, s=0x1c923a0) at /export/scratch1/home/mk/default//package/sql/server/rel_select.c:5366
25 0x00007fffee290eb6 in rel_semantic (sql=0x1bb8ea0, s=0x1c923a0) at /export/scratch1/home/mk/default//package/sql/server/rel_semantic.c:211
26 0x00007fffee1d6bc3 in sql_symbol2relation (c=0x1bb8ea0, sym=0x1c923a0) at /export/scratch1/home/mk/default//package/sql/backends/monet5/sql.c:117
Reproducible: Always
### Steps to Reproduce:
select
(select id from sys.schemas) as c0
from
(select subq_0.c3 as c4
from
lateral (select
ref_0.ndim as c3
from
tmp._tables as sample_1
where ((select index_type_id from sys.index_types) is not NULL)
or (true)) as subq_0
where (select n_regionkey from sys.nation) is not NULL) as subq_1
inner join sys.idxs as ref_1
on (subq_1.c2 = ref_1.id )
inner join sys.netcdf_attrs as sample_15
inner join sys.dependency_types as ref_20
on (sample_15.att_name = ref_20.dependency_type_name )
right join sys.privilege_codes as sample_16
on (sample_15.att_type = sample_16.privilege_code_name )
on (ref_1.type = sample_16.privilege_code_id )
where sample_16.privilege_code_id is NULL;
## Comment 25321
Date: 2017-05-14 14:33:43 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [f3896bec855d](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=f3896bec855d) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=f3896bec855d](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=f3896bec855d)
Changeset description:
fixed bug #6314, ie recursive lateral
## Comment 25325
Date: 2017-05-14 21:56:44 +0200
From: @njnes
fixed in dec016. Also recursive lateral is now handled
| Lateral crash report (sqlsmith) | https://api.github.com/repos/MonetDB/MonetDB/issues/6314/comments | 0 | 2020-11-30T15:48:09Z | 2024-06-27T12:59:36Z | https://github.com/MonetDB/MonetDB/issues/6314 | 753,589,558 | 6,314 |
[
"MonetDB",
"MonetDB"
] | Date: 2017-05-13 14:14:44 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: 11.25.21 (Dec2016-SP4)
CC: @njnes
Last updated: 2017-07-17 16:07:42 +0200
## Comment 25315
Date: 2017-05-13 14:14:44 +0200
From: @mlkersten
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:53.0) Gecko/20100101 Firefox/53.0
Build Identifier:
It seems that the type resolution of the following predicate is incorrect.
where ((select r_regionkey from sys.region) is NULL)
or ((select index_type_name from sys.index_types) is not NULL);
Reproducible: Always
### Steps to Reproduce:
select auxiliary
from sys.tablestoragemodel
where ((select r_regionkey from sys.region) is NULL)
or ((select index_type_name from sys.index_types) is not NULL);
## Comment 25320
Date: 2017-05-13 23:25:09 +0200
From: @mlkersten
Another disjoint query shows similar problems.
select
sample_69.f_table_catalog as c2
from
sys.storagemodel as sample_65
left join sys.geometry_columns as sample_69
on (sample_65.type = sample_69.f_table_catalog )
where
((select default_schema from sys.users) is NULL)
or (cast(coalesce((select side_effect from sys.functions) ,
sample_65.revsorted) as boolean) is not NULL);
!types boolean(1,0) and int(32,0) are not equal for column 'L140'
## Comment 25360
Date: 2017-05-31 16:46:02 +0200
From: MonetDB Mercurial Repository <<hg>>
Changeset [595f2fed54eb](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=595f2fed54eb) made by Niels Nes <niels@cwi.nl> in the MonetDB repo, refers to this bug.
For complete details, see [http//devmonetdborg/hg/MonetDB?cmd=changeset;node=595f2fed54eb](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=http//devmonetdborg/hg/MonetDB?cmd=changeset;node=595f2fed54eb)
Changeset description:
fixes for bug #6313, ie compare types of a union on the parts which need
to be unioned (ie not the intermediate selects + subquery crossproducts or joins)
## Comment 25361
Date: 2017-05-31 16:54:10 +0200
From: @njnes
fixed, ie check union types on the base tables not on the intermediates needed for the or expression.
| Null type resolution in disjunction fails (sqlsmith) | https://api.github.com/repos/MonetDB/MonetDB/issues/6313/comments | 0 | 2020-11-30T15:48:06Z | 2024-06-27T12:59:35Z | https://github.com/MonetDB/MonetDB/issues/6313 | 753,589,520 | 6,313 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.