Datasets:
open-index
/
open-github-issues

Dataset card Data Studio Files
xet
 Community
1
owner
stringclasses
14 values
repo
stringclasses
14 values
id
int64
116k
4.61B
issue_number
int32
1
181k
author
stringlengths
1
39
body
stringlengths
1
262k
created_at
timestamp[us]date
2000-06-06 02:40:44
2026-06-02 20:16:18
updated_at
timestamp[us]date
2000-06-06 02:40:44
2026-06-02 20:22:56
reactions
unknown
author_association
stringclasses
5 values
ClickHouse
ClickHouse
711,964,628
16,107
CurtizJ
It's because `optimize_monotonous_functions_in_order_by` optimization drops `toDate` function as redundant and then we can't match `ORDER BY` expression with sorting key to apply `optimize_read_in_order` optimization. ```sql EXPLAIN SYNTAX SELECT * FROM test_order_by ORDER BY toDate(timestamp) ASC LIMIT 10 ┌─e...
2020-10-19T09:59:58
2020-10-19T10:06:22
{}
MEMBER
ClickHouse
ClickHouse
711,976,091
15,872
socketpair
``` /usr/bin/ld.gold: warning: extract/glibc-compatibility/longjmp.s.o: missing .note.GNU-stack section implies executable stack /usr/bin/ld.gold: warning: extract/glibc-compatibility/syscall.s.o: missing .note.GNU-stack section implies executable stack ``` ``` ./base/glibc-compatibility/musl/x86_64/longjmp.s ....
2020-10-19T10:08:21
2020-10-19T10:14:41
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,168,719
16,152
filimonov
Do you mean producer (sending the same messages several times on error) or consumer (flushing to materialized view)? Can you explain the exact scenario you expect to have?
2020-10-19T13:48:05
2020-10-19T13:48:05
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,357,018
16,037
alexey-milovidov
https://clickhouse.tech/codebrowser/html_report/ClickHouse/src/Storages/StorageReplicatedMergeTree.cpp.html#5804
2020-10-19T18:19:29
2020-10-19T18:19:29
{}
MEMBER
ClickHouse
ClickHouse
709,962,620
16,065
qinglok
`INSERT INTO table_1 SELECT ... FROM table_2 GROUP BY ...` This SQL has two parts. My problem is that the results from the query section are too large for the memory capacity. I'm going to try slicing way.
2020-10-16T10:25:38
2020-10-16T10:25:38
{}
NONE
ClickHouse
ClickHouse
709,963,048
16,041
hczhcz
If I understand it right, `tupleFlatten` is more close to a horizontal version of `arrayJoin` for tuples (not an `arrayFlatten` for tuples).
2020-10-16T10:26:40
2020-10-16T10:26:40
{}
CONTRIBUTOR
ClickHouse
ClickHouse
709,974,461
15,591
alesapin
but `CREATE TABLE test (x Int32, y Int32, s String) ENGINE = MergeTree PRIMARY KEY (x, y);` is not valid create query...
2020-10-16T10:52:55
2020-10-16T10:52:55
{}
MEMBER
ClickHouse
ClickHouse
711,104,046
16,001
alexey-milovidov
Minor fix.
2020-10-18T01:38:30
2020-10-18T01:38:30
{}
MEMBER
ClickHouse
ClickHouse
711,944,011
16,107
CurtizJ
But with `SET optimize_monotonous_functions_in_order_by = 0` `optimize_read_in_order` works :) I guess, something wrong is in rewriting ASTs.
2020-10-19T09:44:02
2020-10-19T09:46:12
{}
MEMBER
ClickHouse
ClickHouse
712,158,846
15,883
yanchaoguo
> @yanchaoguo The reason for this problem may be `binlog_checksum = NONE`, you can check it with `SHOW VARIABLES WHERE Variable_name ='binlog_checksum';` 好的 明天我试试
2020-10-19T13:32:09
2020-10-19T13:32:09
{}
NONE
ClickHouse
ClickHouse
709,951,667
15,073
azat
>I will create some performance tests in this pr later. AFAIU they may fail, since perf tests executed on the upstream/master and this PR, while obviously upstream/master does not have this storage. > A simple show of this engine with bitmap value, but the results may diff to varying filter conditions. Thanks!...
2020-10-16T10:02:03
2020-10-16T10:02:03
{}
MEMBER
ClickHouse
ClickHouse
710,035,008
16,034
akuzm
AST Fuzzer: https://github.com/ClickHouse/ClickHouse/issues/16081 Perf test broken in master. Not sure about the integration tests, will restart.
2020-10-16T13:09:20
2020-10-16T13:11:40
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,053,679
15,966
den-crane
2 CH + 3 zookeeper nodes is right. >Nope, it did not helped. Got this code: Show full CH logs. You can experience "Session expired (Session expired)." for a short time (several seconds) while CH reconnects to another ZK node.
2020-10-16T13:40:03
2020-10-16T13:40:20
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,571,911
16,034
akuzm
The integration tests (release) are timing out again, and this is a known problem.
2020-10-16T20:28:55
2020-10-16T20:28:55
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,647,759
6,997
den-crane
No way to reproduce the issue. No info.
2020-10-16T21:18:23
2020-10-16T21:18:23
{}
CONTRIBUTOR
ClickHouse
ClickHouse
711,104,033
16,004
alexey-milovidov
Minor fix.
2020-10-18T01:38:22
2020-10-18T01:38:22
{}
MEMBER
ClickHouse
ClickHouse
712,292,451
16,157
alexey-milovidov
Could you please provide the details?
2020-10-19T16:45:03
2020-10-19T16:45:03
{}
MEMBER
ClickHouse
ClickHouse
712,368,379
16,037
roohitavaf
But in step 3, we made the block visible to SELECTs. So we make a block visible before the final duplicate check? Does that mean at least for some time we might see duplicates?
2020-10-19T18:41:42
2020-10-19T18:41:42
{}
NONE
ClickHouse
ClickHouse
709,998,893
15,966
sovietotaku
Nope, it did not helped. Got this code: Code: 999. DB::Exception: Received from localhost:9000. DB::Exception: Session expired (Session expired). So, my config now is two Clickhouse nodes and 3 zookeeper nodes. Is it right? Or i need 3 Clickhouse nodes and 3 Zookeeper nodes?
2020-10-16T11:51:32
2020-10-16T11:51:32
{}
NONE
ClickHouse
ClickHouse
710,682,343
8,580
den-crane
could not reproduce with 20.11.1.4897
2020-10-16T22:34:28
2020-10-16T22:34:28
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,711,335
16,064
gaoyang-projects
**kafka_max_block_size = 1048576** from https://altinity.com/blog/clickhouse-kafka-engine-faq posted at 04 May 2020 `input_format_import_nested_json=1` already changed in system.settings and it's works . adding `input_format_import_nested_json=1` to create table scripts and rerun testcase, result is miss column data ...
2020-10-17T00:39:54
2020-10-17T01:05:54
{}
NONE
ClickHouse
ClickHouse
710,829,905
15,173
keliss
And it happens way too often. In system.merges I can see that the part being dropped is also being merged at the same time. It seems to me pointless to continue merging a part when it is no longer needed. And according to the logs, CH tries to stop the merges but looks like it fails to do so, which causes timeouts.
2020-10-17T11:11:00
2020-10-18T01:38:43
{}
NONE
ClickHouse
ClickHouse
711,412,144
11,841
azat
The problem is memory accounting, memory does released. @HooliBox can you check latest version? 20.9 should be fine, I cannot find anything that may affect this between 20.9 and upstream/master
2020-10-18T19:45:31
2020-10-18T19:45:31
{}
MEMBER
ClickHouse
ClickHouse
712,173,996
16,131
filimonov
That is expected: > When using GLOBAL ... JOIN, first the requestor server runs a subquery to calculate the right table. This temporary table is passed to each remote server, and queries are run on them using the temporary data that was transmitted. See https://clickhouse.tech/docs/en/sql-reference/statements/sel...
2020-10-19T13:56:16
2020-10-19T13:56:16
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,049,031
16,073
den-crane
solution is here https://github.com/ClickHouse/ClickHouse/issues/6063#issuecomment-513620887
2020-10-16T13:31:33
2020-10-16T13:31:33
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,203,243
16,085
den-crane
ERROR: type should be string, got "https://clickhouse.tech/docs/en/sql-reference/functions/date-time-functions/#formatdatetime\r\n```\r\nselect formatDateTime(today(), '%Y-%V-%u')\r\n┌─formatDateTime(today(), '%Y-%V-%u')─┐\r\n│ 2020-42-5 │\r\n└─────────────────────────────────────┘\r\n\r\n\r\n\r\nSELECT formatDateTime(today(), '%Y-%G-%V-%u')\r\n\r\n┌─formatDateTime(today(), '%Y-%G-%V-%u')─┐\r\n│ 2020-2020-42-5 │\r\n└────────────────────────────────────────┘\r\n```"
2020-10-16T16:44:20
2020-10-16T16:45:20
{ "+1": 1 }
CONTRIBUTOR
ClickHouse
ClickHouse
710,657,347
8,425
den-crane
Works now. At least after 20.3.19.4.
2020-10-16T21:28:44
2020-10-16T21:28:44
{}
CONTRIBUTOR
ClickHouse
ClickHouse
709,981,240
16,041
amosbird
> If I understand it right, `tupleFlatten` is more close to a horizontal version of `arrayJoin` for tuples (not an `arrayFlatten` for tuples). Exactly. BTW arrayJoin has a standard name called `unnested`.
2020-10-16T11:08:37
2020-10-16T11:08:37
{}
COLLABORATOR
ClickHouse
ClickHouse
711,009,124
8,331
Vxider
> > > what's the progress of this feature?CAN be Released in this Year? > > > > > > The basic implementation of WindowView is done. We are waiting for a code review now. > > We currently have another optimized version, and we will consider committing to the community after the basic version is merged. > > @Vxi...
2020-10-17T14:13:43
2020-10-17T14:13:43
{}
CONTRIBUTOR
ClickHouse
ClickHouse
711,131,137
1,469
l1t1
those functions are hard to understand, i prefer the standard window functions (https://github.com/ClickHouse/ClickHouse/issues/5132). as i know, both actian vector and sqlite realized them.
2020-10-18T07:53:35
2020-10-18T07:53:35
{ "+1": 3 }
NONE
ClickHouse
ClickHouse
711,154,720
15,967
Jokser
@macEar `<proxy>` configuration section should be placed under `<s3>` section. In your configs, it's placed at the same level as `<s3>`.
2020-10-18T11:33:37
2020-10-18T11:33:37
{ "+1": 1 }
CONTRIBUTOR
ClickHouse
ClickHouse
712,157,445
15,883
zhang2014
@yanchaoguo The reason for this problem may be `binlog_checksum = NONE`, you can check it with `SHOW VARIABLES WHERE Variable_name ='binlog_checksum';`
2020-10-19T13:29:52
2020-10-19T13:30:06
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,352,609
16,078
Akazz
Failure of Integration tests (release) (test_host_ip_change/test.py::test_ip_change_update_dns_cache) is a pretty worthless matter, in the context of this PR.
2020-10-19T18:11:22
2020-10-19T18:11:22
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,816,120
15,685
vivarum
@Avogar @alexey-milovidov, can't understand from the logs the cause of tests failure. Is it something with the patch or is it some connection problems?
2020-10-17T10:52:11
2020-10-17T10:52:11
{}
CONTRIBUTOR
ClickHouse
ClickHouse
711,155,434
16,114
Jokser
Yes, you can use the `s3` table function: `select * from s3('https://s3_endpoint/bucket/file', 'CSV', 'table format')` Btw, when you create a table with S3 engine it doesn't store any data on the local disk and you don't need to insert something into table first. It queries a data directly from S3.
2020-10-18T11:40:16
2020-10-18T14:24:01
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,157,167
16,040
kitaisreal
@akuzm Okay. Will fix that.
2020-10-19T13:29:22
2020-10-19T13:29:22
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,019,205
16,041
KochetovNicolai
Segfault: ``` SELECT tupleFlatten(argMax(tuple(* EXCEPT(number)), number)) + 3 FROM ( SELECT number, number + 1 AS n_1, number + 2 FROM numbers(10) ) GROUP BY number % 2 2020.10.16 15:35:16.314829 [ 17411 ] {} <Fatal> BaseDaemon: ######################################## 2...
2020-10-16T12:36:55
2020-10-16T12:36:55
{}
MEMBER
ClickHouse
ClickHouse
710,069,620
11,721
crakjie
@siradjev I'm also looking for this kind of processing. I think the following exemple could be a work around for sumMapFiltered ``` WITH arrayFilter(x -> ((x.2) = 'abc'), arrayZip(Array1, Array2, ArrayMetrics)) AS filteredArrays SELECT sumMap(filteredArrays.1, filteredArrays.3) FROM Table ``` It's filter both ar...
2020-10-16T14:07:07
2020-10-16T14:07:07
{}
NONE
ClickHouse
ClickHouse
712,171,202
16,134
filimonov
Please try increasing `max_ast_depth` setting
2020-10-19T13:51:58
2020-10-19T13:51:58
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,181,222
16,131
den-crane
@chang0391 CH version ?
2020-10-19T14:06:52
2020-10-19T14:06:52
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,041,176
16,080
den-crane
КМК это исправлено очень давно. Вам нужно обновится до 20.3.20.6. 20.3.20.6 это багфикс релиз, обновление довольно безопасно.
2020-10-16T13:21:01
2020-10-16T13:21:01
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,014,191
16,076
azat
>Introduced #8750 Actually it has been introduced in #11931
2020-10-16T12:26:13
2020-10-16T14:19:38
{ "+1": 1 }
MEMBER
ClickHouse
ClickHouse
710,740,345
15,428
amosbird
Sorry for the delay. We're aware of this issue https://github.com/ClickHouse/ClickHouse/issues/16108 and it's expected to be merged soon.
2020-10-17T03:42:22
2020-10-17T03:42:22
{}
COLLABORATOR
ClickHouse
ClickHouse
711,484,321
15,806
hexiaoting
@CurtizJ I have finished coding. Can you review it when free? Thank you.
2020-10-19T03:04:55
2020-10-19T03:04:55
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,137,983
16,086
akuzm
ast fuzzer https://github.com/ClickHouse/ClickHouse/issues/14974 perf test broken in master
2020-10-19T12:57:54
2020-10-19T12:57:54
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,184,039
16,130
den-crane
mysql -uptx -p -h**172.16.6.207** -P3306 select * from mysql('**172.16.6.207**:3306', Can't connect to MySQL server on '**192.168.5.207'** (from 172.16.6.210:53228) (in query: select * from mysql( **192.168.5.207:** 3306', Clickhouse version **20.8.2.3** version **20.9.3.45**
2020-10-19T14:09:15
2020-10-19T14:11:24
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,304,375
8,546
mga-chka
I think it has been fixed because on version 20.9.3.45 I have the following behavior (which looks logical) `SELECT length(topK(30)(A)) FROM topX ┌─length(topK(30)(A))─┐ │ 30 │ └──────────────┘ SELECT length(topK(30)(A)) FROM remote('localhost,127.0.0.1', currentDatabase(), to...
2020-10-19T17:06:30
2020-10-19T17:09:13
{}
NONE
ClickHouse
ClickHouse
710,054,734
15,670
den-crane
>Why does it try to find these parts if they were removed? Can you show `echo stat | nc <zookeeper ip> 2181` ? for each ZK.
2020-10-16T13:41:52
2020-10-16T13:42:15
{}
CONTRIBUTOR
ClickHouse
ClickHouse
711,103,979
16,009
alexey-milovidov
> test_role > test_grant_and_revoke Deadlocks, already fixed.
2020-10-18T01:37:27
2020-10-18T01:37:27
{}
MEMBER
ClickHouse
ClickHouse
711,104,038
16,002
alexey-milovidov
Minor fix.
2020-10-18T01:38:26
2020-10-18T01:38:26
{}
MEMBER
ClickHouse
ClickHouse
711,459,644
15,994
spihiker
After I executed the command SET DEFAULT ROLE...TO ... ,The user has enough privileges as i expected. But when i executed the command SHOW CURRENT ROLES ,the list has nothing.Why? jyw-test-221 :) set default role devgroup_3399,group_com to webuser; SET DEFAULT ROLE devgroup_3399, group_com TO webuser Ok. 0 ...
2020-10-19T01:23:39
2020-10-19T01:23:39
{}
NONE
ClickHouse
ClickHouse
712,215,922
15,702
liangli1213
@BohuTANG @zhang2014 , in my environment I have below requirement, I think this is normal one and very useful for HA. 1. Retry to connect to MySQL if failed, just like MySQL slave keep trying to connect to master every 60s if failed 2. Replication keep going on when use VIP switching from one MySQL node to another
2020-10-19T14:45:04
2020-10-19T14:45:04
{}
NONE
ClickHouse
ClickHouse
710,051,913
16,065
den-crane
If you insert into SummingMT / AggregatingMT Engines there are other ways. You can insert without groupby for example.
2020-10-16T13:36:54
2020-10-16T13:36:54
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,516,060
16,039
myrrc
After discussion the cons of this feature overweighted the pros.
2020-10-16T19:52:42
2020-10-16T19:52:42
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,718,871
16,105
alexey-milovidov
rand, now, dictGet - are nondeterministic functions of different kind. Maybe we can describe it somehow in Function interface, so dictGet will always work to skip unused shards, without a need for separate setting?
2020-10-17T01:18:04
2020-10-17T01:18:04
{}
MEMBER
ClickHouse
ClickHouse
667,659,099
4,881
nvartolomei
I found a few cases of "watch leak". One example is detaching tables, watches are not removed. Re-attaching a table multiple times will create 2 watches. On an usual server with some activity these go away on the first event on the watched paths. This might or might not be related to the original issue. I reckon the...
2020-08-02T10:58:33
2020-10-19T13:02:06
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,153,003
16,040
akuzm
Looks like these changes have broken the `parse_engine_file` perf test, and the processing of perf tests itself. Try to reproduce it locally -- start the server with your changes and run `ClickHouse/docker/test/performance-comparison/perf.py -- ClickHouse/test/performance/parse_engine_file.xml`. The docs for the ...
2020-10-19T13:22:42
2020-10-19T13:22:42
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,199,422
15,409
liangli1213
@adilyas @zhang2014 , would you please tell me what is the usage or syntax of this feature?
2020-10-19T14:27:54
2020-10-19T14:27:54
{}
NONE
ClickHouse
ClickHouse
712,356,388
16,037
alexey-milovidov
> But what if someone in some other replica add a block with the same hash after we checked the hash, i.e. after step 2? On step 4 it's also checked; atomically.
2020-10-19T18:18:19
2020-10-19T18:18:19
{}
MEMBER
ClickHouse
ClickHouse
709,954,509
16,010
KochetovNicolai
Build is broken in release branch Stress test is a ci issue.
2020-10-16T10:07:49
2020-10-16T10:07:49
{}
MEMBER
ClickHouse
ClickHouse
709,964,895
15,899
alesapin
Let's check in branch
2020-10-16T10:30:52
2020-10-16T10:30:52
{}
MEMBER
ClickHouse
ClickHouse
710,009,673
16,064
filimonov
Add `input_format_import_nested_json` to your Kafka table settings. ```diff CREATE TABLE IF NOT EXISTS kafka_queue ( timestamp UInt64, id Nullable(String), `localEndpoint.serviceName` Nullable(String) ) ENGINE = Kafka SETTINGS kafka_broker_list = '', kafka_topic_li...
2020-10-16T12:15:51
2020-10-16T12:15:51
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,649,765
8,690
den-crane
@tdingus is it still valid ? Did you check new CH versions?
2020-10-16T21:20:41
2020-10-16T21:20:41
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,678,302
8,643
den-crane
still relevant 20.11.1.4897
2020-10-16T22:19:39
2020-10-16T22:19:39
{}
CONTRIBUTOR
ClickHouse
ClickHouse
711,549,713
15,977
TCeason
It's working fine on my machine. ``` ClickHouse/tests/integration/test_materialize_mysql_database# pytest ./test.py ======================================================================== test session starts ======================================================================== platform linux -- Python 3.8.5,...
2020-10-19T04:42:46
2020-10-19T04:42:46
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,210,631
15,702
liangli1213
@BohuTANG , have you got a chance to reproduce it?
2020-10-19T14:38:26
2020-10-19T14:38:26
{}
NONE
ClickHouse
ClickHouse
712,343,004
16,037
roohitavaf
So, the order is 1. Write to temp file, and calculate the hash. 2. Check the hash for duplicates 3. Make file visible 4. Write block hash and metadata to Zookeeper. But what if someone in some other replica add a block with the same hash after we checked the hash, i.e. after step 2? Shouldn't be order l...
2020-10-19T18:00:50
2020-10-19T18:01:11
{}
NONE
ClickHouse
ClickHouse
709,967,480
15,823
kitaisreal
@alesapin Okay, thanks.
2020-10-16T10:36:45
2020-10-16T10:36:45
{}
CONTRIBUTOR
ClickHouse
ClickHouse
709,999,669
16,072
alexey-milovidov
You are using Apple Clang. To build ClickHouse, you should install `clang` from brew.
2020-10-16T11:53:18
2020-10-16T11:53:18
{}
MEMBER
ClickHouse
ClickHouse
710,067,912
15,670
keliss
Sure, here: 1. ``` Zookeeper version: 3.4.14-4c25d480e66aadd371de8bd2fd8da255ac140bcf, built on 03/06/2019 16:18 GMT Clients: some clients' ip's Latency min/avg/max: 0/0/25 Received: 39615904 Sent: 39615903 Connections: 8 Outstanding: 0 Zxid: 0xb00030dd5 Mode: follower Node count: 483223 ``` 2. ``` Z...
2020-10-16T14:04:11
2020-10-16T14:04:11
{}
NONE
ClickHouse
ClickHouse
710,758,077
16,105
azat
>rand, now, dictGet - are nondeterministic functions of different kind. Totally agree >Maybe we can describe it somehow in Function interface, so dictGet will always work to skip unused shards, without a need for separate setting? From one side I don't like adding a setting for this since most of the time it i...
2020-10-17T06:05:59
2020-10-17T06:05:59
{}
MEMBER
ClickHouse
ClickHouse
711,297,942
16,114
tooptoop4
@Jokser @alexey-milovidov 1. anything in the docs about it? 2. or how to access parquet file from S3? 3. also any idea how much slower these queries run compared to if the data were inserted onto local disk? 4. does it accept IAM roles or HTTP proxies to access S3?
2020-10-18T17:10:02
2020-10-18T18:48:36
{}
NONE
ClickHouse
ClickHouse
712,170,319
16,130
filimonov
> Can't connect to MySQL server on '192.168.5.207' (4) Check your firewall. It seems ClickHouse host can't establish connection to your mysql host.
2020-10-19T13:50:29
2020-10-19T13:50:29
{}
CONTRIBUTOR
ClickHouse
ClickHouse
709,965,288
15,823
alesapin
Ok, I'll add several tests by myself.
2020-10-16T10:31:48
2020-10-16T10:31:48
{}
MEMBER
ClickHouse
ClickHouse
709,995,094
15,670
keliss
Hi @den-crane , after upgrading to the latest version and restarting the CH, I see a lot of the same messages in the error log. Zgrep again shows that these parts were removed at some point. Error log: Removing locally missing part from ZooKeeper and queueing a fetch: 1594965600_1_6_1 Checking part 1594965600_1_6_1 ...
2020-10-16T11:42:29
2020-10-16T11:42:29
{}
NONE
ClickHouse
ClickHouse
710,074,315
16,083
alesapin
Fix is incorrect.
2020-10-16T14:15:16
2020-10-16T14:15:16
{}
MEMBER
ClickHouse
ClickHouse
710,237,248
16,094
nvartolomei
Context: Need a way of adding a new field for in memory parts #16033 cc @CurtizJ @alesapin
2020-10-16T17:07:31
2020-10-16T17:09:58
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,675,637
8,490
den-crane
Could not reproduce with current stable versions. Test added.
2020-10-16T22:10:20
2020-10-16T22:10:20
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,831,838
15,173
keliss
And these two messages are appearing from time to time, I'm a little worried about them (especially the latter, because subsequent alter succeeded and it looks like it was going to drop more than needed at the first attempt): ``` 2020.10.17 11:06:34.097839 [ 266902 ] {} <Error> DynamicQueryHandler: Code: 373, e.displ...
2020-10-17T11:14:26
2020-10-18T01:38:12
{}
NONE
ClickHouse
ClickHouse
712,103,626
16,148
socketpair
Unfortunately, changes are not enough: ![image](https://user-images.githubusercontent.com/1157412/96445726-ce5b9680-1229-11eb-9c44-3f27584e9b20.png) I need piece of advice how to fix better. Yes, I can enforce `-z noexecstack`, but seems it's a crutch. Seel also: https://wiki.ubuntu.com/SecurityTeam/Roadmap/E...
2020-10-19T11:51:01
2020-10-19T12:01:37
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,295,780
16,037
roohitavaf
What will happen if ClickHouse server crashes after step 3 and before step 4, i.e. before committing transaction to Zookeeper?
2020-10-19T16:51:07
2020-10-19T16:51:07
{}
NONE
ClickHouse
ClickHouse
709,950,252
16,071
hczhcz
A further issue: ``` select if(0, toDecimal64(123, 2), 456); -- ok select if(1, toDecimal64(123, 2), 456); -- ok select if(arrayJoin([0, 1]), toDecimal64(123, 2), 456); -- runtime error ```
2020-10-16T09:59:02
2020-10-16T09:59:02
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,014,583
15,385
nikitamikhaylov
Could you provide us a full table structure of source and destination tables? The I will try to write a test and fix it. Also you can try to use copier from 20.10 version. All columns from primary key are being wrapped with backticks there.
2020-10-16T12:27:04
2020-10-16T12:27:04
{}
MEMBER
ClickHouse
ClickHouse
710,018,444
12,741
UnamedRus
Double Distributed JOINs doesn't work correctly with row policy and allow_insecure_prewhere=1: ``` create table test_join ( _time DateTime, team LowCardinality(String) , dt Date DEFAULT toDate(_time), hash UInt64 DEFAULT cityHash64(concat(CAST(_time, 'String'), team)) ) ENGINE = MergeTree ...
2020-10-16T12:35:24
2020-10-16T12:47:15
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,050,241
16,073
den-crane
fix https://github.com/ClickHouse/ClickHouse/pull/8202 Support DROP DATABASE, DETACH TABLE, DROP TABLE and ATTACH TABLE for MySQL database engine. #8202 (Winter Zhang) you need to upgrade CH to 20.7+
2020-10-16T13:33:41
2020-10-16T13:33:41
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,366,908
16,075
tavplubix
Functional stateless tests - 01133_max_result_rows - flacky Stress test (thread) - Sanitizer assert - fixed in #15984
2020-10-16T18:19:56
2020-10-16T18:19:56
{}
MEMBER
ClickHouse
ClickHouse
710,823,886
15,173
keliss
@den-crane Hi, we've upgraded to 20.9.3.45-2, and this behavior seems to be somewhat alleviated and no complete deadlock happens (at least, we haven't encountered it yet), but here's what I was able to reproduce: ``` 2020.10.17 10:50:27.754895 [ 232093 ] {dd1f198e-b13b-4fbb-af0b-2c6874f40694} <Error> executeQuery: Co...
2020-10-17T11:02:26
2020-10-17T11:02:26
{}
NONE
ClickHouse
ClickHouse
710,864,532
16,105
azat
@alexey-milovidov also note that without #14731 there will be two dictGet for skipping unused shards: - once for query stage - second for query execution
2020-10-17T11:58:15
2020-10-17T11:58:15
{}
MEMBER
ClickHouse
ClickHouse
712,314,802
16,037
alexey-milovidov
On next startup, it will compare the state in ZooKeeper (as ground truth) with the local state, consider the written part as "unexpected" and rename it to the `detached` directory.
2020-10-19T17:25:42
2020-10-19T17:25:50
{}
MEMBER
ClickHouse
ClickHouse
709,979,137
15,856
nikitamikhaylov
> @nikitamikhaylov could you check this PR? Of course, but I am a little bit busy this week.
2020-10-16T11:03:42
2020-10-16T11:03:42
{}
MEMBER
ClickHouse
ClickHouse
710,046,048
16,050
den-crane
https://github.com/ClickHouse/ClickHouse/blob/master/CHANGELOG.md#experimental-feature-1 you need CH version 20.8+
2020-10-16T13:27:16
2020-10-16T13:27:16
{}
CONTRIBUTOR
ClickHouse
ClickHouse
710,076,789
15,994
vitlibar
Probably the `group_com` role was not current for `webuser`. You can check the current roles with the command `SHOW CURRENT ROLES` and see if the list of the current roles includes `group_com`.
2020-10-16T14:19:50
2020-10-16T14:20:25
{}
MEMBER
ClickHouse
ClickHouse
710,078,257
14,414
filimonov
`double free or corruption (out)` and sig_abort near ExpressionTransform / ExpressionActions https://github.com/ClickHouse/ClickHouse/pull/15199#issuecomment-709294693
2020-10-16T14:22:30
2020-10-16T14:22:30
{}
CONTRIBUTOR
ClickHouse
ClickHouse
711,224,815
15,932
azat
>It consumes ~4MB per request and accumulates up to the 3.6GB limit. I see a continuous growth in MemoryTracker, which, if you do not stop the requests (or switch to other request), will not be reset. Indeed, I can confirm this. I won't clutter up the comment with details, but if you are curios you can take a loo...
2020-10-18T15:29:16
2020-10-18T15:29:16
{ "+1": 1, "heart": 1 }
MEMBER
ClickHouse
ClickHouse
712,179,400
16,064
filimonov
> kafka_max_block_size = 1048576 from https://altinity.com/blog/clickhouse-kafka-engine-faq Yes, and that's ok. But you have: ``` select * from system.settings where changed max_block_size 1048576 ``` That was never recommended, and I was referring that. > input_format_import_nested_json=1 already cha...
2020-10-19T14:03:55
2020-10-19T14:03:55
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,213,045
13,675
den-crane
20.11.1.4897 relevant
2020-10-19T14:40:34
2020-10-19T14:40:34
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,313,611
11,247
alexey-milovidov
In recent versions we have added better diagnostics on free space. It will show you the filesystem path, the amount of total and free space and inodes.
2020-10-19T17:23:36
2020-10-19T17:23:36
{ "+1": 2 }
MEMBER
ClickHouse
ClickHouse
712,453,546
16,026
akuzm
Docker for woboq is broken, not related to this PR.
2020-10-19T21:30:09
2020-10-19T21:30:09
{}
CONTRIBUTOR
ClickHouse
ClickHouse
712,877,595
14,135
alesapin
A very strange thing happened in docker while binary installed using clickhouse install: ``` root@click:/# whoami root root@click:/# ls -laF /usr/bin/clickhouse -r-xr-xr-x 1 root root 299278344 Oct 20 17:00 /usr/bin/clickhouse* root@click:/# /usr/bin/clickhouse bash: /usr/bin/clickhouse: Operation not permitted ...
2020-10-20T14:05:06
2020-10-20T14:05:06
{ "+1": 1 }
MEMBER
ClickHouse
ClickHouse
712,995,768
15,977
tavplubix
> `DROP DATABASE test_database` hangs for unknown reason. Probably it's something similar to #13822. Let's try to capture stacktraces It's not related to your changes, deadlock on `MaterializeMySQL` database shutdown: ``` 35 ceee5aa3-9035-4b09-9894-6daa37013147 pthread_cond_wait /usr/lib/x86_64-linux-gnu/lib...
2020-10-20T16:45:10
2020-10-20T16:45:10
{}
MEMBER