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 | 766,176,379 | 19,513 | alexey-milovidov | @jyz0309 Thank you!
> if it is too easy(or make you plague) to become a contributor through this pr,I will close this pull request.
It's totally alright :) | 2021-01-23T20:45:38 | 2021-01-23T20:45:38 | {} | MEMBER |
ClickHouse | ClickHouse | 766,480,664 | 19,352 | godliness | found the roadmap list: https://github.com/ClickHouse/ClickHouse/issues/17623 | 2021-01-25T01:16:17 | 2021-01-25T01:16:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,497,821 | 19,403 | JunjianS | > aggregation_memory_efficient_merge_threads
> Also need to enable `distributed_aggregation_memory_efficient` first. It's not enabled by default.
@den-crane @alexey-milovidov thanks ，i set below settings,
set distributed_aggregation_memory_efficient = 1;
set aggregation_memory_efficient_merge_threads=2;
... | 2021-01-25T02:23:34 | 2021-01-25T02:25:04 | {} | NONE |
ClickHouse | ClickHouse | 766,597,174 | 19,349 | nnsgmsone | Thanks, I understand | 2021-01-25T07:05:25 | 2021-01-25T07:05:25 | {} | NONE |
ClickHouse | ClickHouse | 765,731,144 | 19,426 | alexey-milovidov | It's being implemented by @MaxTheHuman (just started). | 2021-01-22T22:53:27 | 2021-01-22T22:53:27 | {} | MEMBER |
ClickHouse | ClickHouse | 766,194,837 | 19,108 | alexey-milovidov | It may happen after adding "sigaltstack".
If some thread expects to have alternative signal stack but it has not been set up correctly, we will have this error.
> No other messages from thread 56 in log.
We have only tail of the server log, maybe we need a full log to figure out what thread 56 is. | 2021-01-23T22:52:13 | 2021-01-23T22:52:13 | {} | MEMBER |
ClickHouse | ClickHouse | 766,392,665 | 19,531 | winoenix | It seems that the default database comes to Atomic from Ordinary? | 2021-01-24T16:48:53 | 2021-01-24T16:48:53 | {} | NONE |
ClickHouse | ClickHouse | 766,417,812 | 19,466 | azat | I have few more fixes, that's fine, will submit separately | 2021-01-24T19:28:00 | 2021-01-24T19:28:00 | {} | MEMBER |
ClickHouse | ClickHouse | 765,672,997 | 19,393 | alexey-milovidov | Ok, I see. | 2021-01-22T20:41:59 | 2021-01-22T20:41:59 | {} | MEMBER |
ClickHouse | ClickHouse | 765,675,574 | 19,459 | azat | I have similar patch, likely I did not submit PR yet, but I noticed one more UB:
```
SELECT toIntervalMinute(9223372036854775807); -- INT64_MIN+1
SELECT toIntervalMinute(9223372036854775808); -- INT64_MIN
```
Looks like it is OK to add `NO_SANITIZE_UNDEFINED` for this function.
https://clickhouse-test-repor... | 2021-01-22T20:47:42 | 2021-01-22T20:47:42 | {} | MEMBER |
ClickHouse | ClickHouse | 765,676,187 | 19,459 | alexey-milovidov | Could you please submit the PR nevertheless?
I will integrate the changes...
Yes, it's Ok to add NO_SANITIZE_UNDEFINED. | 2021-01-22T20:49:10 | 2021-01-22T20:49:10 | {} | MEMBER |
ClickHouse | ClickHouse | 766,311,634 | 19,527 | kitaisreal | @ucasFL could you please add documentation to this `sign` function ? | 2021-01-24T08:36:57 | 2021-01-24T08:36:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,399,096 | 18,346 | mauidude | Following up on this... anyone know? | 2021-01-24T17:30:19 | 2021-01-24T17:30:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,521,322 | 19,403 | JunjianS | i updated below settings, the query result came out.@_@
set max_memory_usage=300000000000
set max_bytes_before_external_group_by=15000000000 #(this value is max_memory_usage/2 before, is max_memory_usage/20 now )
set distributed_aggregation_memory_efficient = 1;
1629365 rows in set. Elapsed: 116.427 sec. Processe... | 2021-01-25T03:48:18 | 2021-01-25T03:51:07 | {} | NONE |
ClickHouse | ClickHouse | 766,564,889 | 19,524 | alexey-milovidov | Still no segfaults. | 2021-01-25T06:00:30 | 2021-01-25T06:00:30 | {} | MEMBER |
ClickHouse | ClickHouse | 765,693,898 | 19,459 | azat | >Could you please submit the PR nevertheless?
Sure - #19466
*P.S. I've removed bits that you have already addressed* | 2021-01-22T21:27:28 | 2021-01-22T21:27:40 | {} | MEMBER |
ClickHouse | ClickHouse | 765,831,001 | 19,277 | tisonkun | closed by https://github.com/ClickHouse-Extras/libpq/commit/f1b8fe5716f1ba3954897421a964fd3d1de70f02 & #19377 | 2021-01-23T02:47:06 | 2021-01-23T02:47:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,177,655 | 19,494 | alexey-milovidov | Possible reason is the usage of old clang version (9.0). We are using clang 11 in CI and for releases.
Nevertheless, you can disable the problematic code inside `src/Parsers/New/` (it is optional) as a workaround. | 2021-01-23T20:55:14 | 2021-01-23T20:55:14 | {} | MEMBER |
ClickHouse | ClickHouse | 766,566,581 | 19,349 | alexey-milovidov | Hypothesis: the query without deduplication returns more records and it takes more time to transfer the result. | 2021-01-25T06:03:50 | 2021-01-25T06:03:50 | {} | MEMBER |
ClickHouse | ClickHouse | 766,600,539 | 19,346 | alesapin | Ok, let's merge and check. | 2021-01-25T07:11:13 | 2021-01-25T07:11:13 | {} | MEMBER |
ClickHouse | ClickHouse | 766,619,923 | 19,407 | templarzq | ok,i'll check the results. | 2021-01-25T07:49:23 | 2021-01-25T07:49:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 765,653,489 | 19,412 | alesapin | ubsan reported https://github.com/ClickHouse/ClickHouse/issues/19457 | 2021-01-22T20:03:55 | 2021-01-22T20:03:55 | {} | MEMBER |
ClickHouse | ClickHouse | 766,112,367 | 17,847 | fibersel | Sorry for a such long break, but finally I measured first new algorithm. @alexey-milovidov does that representation satisfies you requirements?If it's so, I'll submit code for measuring and benchmarks for other data types later.
Current measurement is for Titles.
| | compression (mb/s) | decompression... | 2021-01-23T16:18:35 | 2021-01-23T16:18:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,594,964 | 19,030 | sliontc | Yes, I figured it out that the clickhouse service have to be restarted finally. | 2021-01-25T07:01:27 | 2021-01-25T07:01:27 | {} | NONE |
ClickHouse | ClickHouse | 766,680,104 | 18,945 | qoega | 01459_manual_write_to_replicas https://clickhouse-test-reports.s3.yandex.net/0/9df296b55e6143acca2eba35036059686e3057ea/functional_stateless_tests_(release,_databaseordinary).html | 2021-01-25T09:29:20 | 2021-01-25T09:29:20 | {} | MEMBER |
ClickHouse | ClickHouse | 765,651,042 | 19,412 | alesapin | `test_materialize_mysql_database/test.py::test_select_without_columns_8_0` known flaky. | 2021-01-22T19:59:02 | 2021-01-22T19:59:02 | {} | MEMBER |
ClickHouse | ClickHouse | 765,863,216 | 18,788 | sundy-li | Server crashed , cc @hexiaoting
```
select map( 'aa', 4, '44' , 5) as c, mapKeys(c)
Query id: 0d737d61-ae11-4665-ae34-d70fa0237342
[ubuntu] 2021.01.23 04:12:52.627166 [ 22526 ] {0d737d61-ae11-4665-ae34-d70fa0237342} <Fatal> : Logical error: 'Empty column was returned by function mapKeys'.
[ubuntu] 2021.01... | 2021-01-23T04:16:06 | 2021-01-23T04:16:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,274,623 | 17,847 | alexey-milovidov | Also I think we can first integrate a bunch of interesting libraries... and then we will find something interesting. | 2021-01-24T01:58:52 | 2021-01-24T01:58:52 | {} | MEMBER |
ClickHouse | ClickHouse | 766,311,109 | 19,386 | kitaisreal | @spz0507 assigned. | 2021-01-24T08:31:41 | 2021-01-24T08:31:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,563,568 | 19,349 | nnsgmsone | @alexey-milovidov Can you help me solve my confusion? | 2021-01-25T05:57:42 | 2021-01-25T05:57:42 | {} | NONE |
ClickHouse | ClickHouse | 766,569,206 | 19,349 | nnsgmsone |
```
executeQuery: Read 100000000 rows, 1.40 GiB in 0.235573548 sec., 424495877 rows/sec., 5.93 GiB/sec.
executeQuery: Read 100000000 rows, 1.40 GiB in 0.887121595 sec., 112724118 rows/sec., 1.57 GiB/sec.
```
Do you mean that this time also includes the transmission time, if so, it is fine.
@alexey-milovidov | 2021-01-25T06:09:00 | 2021-01-25T06:09:00 | {} | NONE |
ClickHouse | ClickHouse | 766,609,708 | 19,407 | qoega | You can check that this tests are green without your changes. It is highly likely they are related to your changes. | 2021-01-25T07:29:43 | 2021-01-25T07:29:43 | {} | MEMBER |
ClickHouse | ClickHouse | 766,663,058 | 19,532 | qoega | Fuzzer fail is related
``` sql
SELECT lcm(NULL, NULL), lcm(nan, 2), lcm(NULL, 65537)
```
```
../src/Functions/lcm.cpp:56:77: runtime error: nan is outside the range of representable values of type 'long'
2021.01.25 03:18:20.586502 [ 155 ] {} <Fatal> BaseDaemon: ########################################
2021.01... | 2021-01-25T09:01:58 | 2021-01-25T09:01:58 | {} | MEMBER |
ClickHouse | ClickHouse | 765,830,825 | 19,400 | tisonkun | @kitaisreal thanks for you information. Build with master is green on my environment the same with #19277 reported. Closing the issue and this PR... | 2021-01-23T02:45:49 | 2021-01-23T02:45:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,058,302 | 18,861 | nautaa | hi, when I added libcxx-pretty-printers, there is a problem. It is failed on stack overflow-error. How do you use it?
@hexiaoting
```
File "/home/nauta/Downloads/libcxx-pretty-printers/src/libcxx/v1/printers.py", line 844 in invoke
File "/home/nauta/Downloads/libcxx-pretty-printers/src/libcxx/v1/printers.py", li... | 2021-01-23T13:32:51 | 2021-01-23T13:32:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,601,328 | 19,409 | songenjie | ./clickhouse client -h. ** --port *** -u default -m --password=** --log-level=trace --send_logs_level=trace
case 1:
Connecting to *** as user default.
Connecting. Database: (not specified). User: default
Code: 32. DB::Exception: Attempt to read after eof
Uninitializing subsystem: Logging Subsystem
case 2:
... | 2021-01-25T07:12:38 | 2021-01-25T07:12:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 765,603,617 | 19,212 | sevirov | @Avogar I fixed it. | 2021-01-22T18:26:58 | 2021-01-22T18:26:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,311,032 | 19,466 | azat | >Integration tests (asan) — fail: 18, passed: 11, error: 20
```
test_materialize_mysql_database
```
Looks like it maybe fixed by #19344 | 2021-01-24T08:30:48 | 2021-01-24T08:30:48 | {} | MEMBER |
ClickHouse | ClickHouse | 766,504,220 | 19,524 | alexey-milovidov | It passed again. Let's rerun. | 2021-01-25T02:46:11 | 2021-01-25T02:46:11 | {} | MEMBER |
ClickHouse | ClickHouse | 766,665,478 | 19,532 | qoega | Build error:
```
2021-01-25 01:53:42,652 Will build ClickHouse pkg with cmd: 'docker run --network=host --rm --volume=/place/sandbox-data/ramdrive/878396698/ClickHouse/build_output_folder:/output --volume=/place/sandbox-data/ramdrive/878396698/ClickHouse/docker/packager/../..:/build --volume=/place/sandbox-data/ramdr... | 2021-01-25T09:05:46 | 2021-01-25T09:05:46 | {} | MEMBER |
ClickHouse | ClickHouse | 765,683,946 | 19,108 | alexey-milovidov | https://clickhouse-test-reports.s3.yandex.net/19385/117708abc616a777014566cca6f1a00a3dae4109/fuzzer_debug/server.log | 2021-01-22T21:06:21 | 2021-01-22T21:06:21 | {} | MEMBER |
ClickHouse | ClickHouse | 765,687,252 | 19,444 | nikitamikhaylov | > > 00158_cache_dictionary_has | FAIL | 600.38
>
>
>
> Timeout :( @nikitamikhaylov
>
>
>
>
>
>
I told earlier that it is very slow test, because it uses ClickHouse as a source for dictionary. But I didn’t change the “size” of this test since then. | 2021-01-22T21:13:39 | 2021-01-22T21:13:39 | {} | MEMBER |
ClickHouse | ClickHouse | 765,696,466 | 18,446 | azat | >Fix constant folding in EXPLAIN SYNTAX when the result is unknown (i.e. for subqueries)
@akuzm but this it not only for `EXPLAIN SYNTAX`, but also during subqueries analysis
Origina PR contains the following description:
```
Subqueries analyzed with only_analyze is set to true, and this will
simply replace ... | 2021-01-22T21:30:48 | 2021-01-22T21:30:48 | {} | MEMBER |
ClickHouse | ClickHouse | 765,967,774 | 19,493 | kitaisreal | accurateCastOrNull function was not implemented in 20.12 | 2021-01-23T12:40:32 | 2021-01-23T12:40:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,077,823 | 19,512 | olgarev | @vitlibar Hi, will you review this, pls?
Unfortunately, I have not enough privileges for testing new syntax.
May be, we should add some examples for multiple names?
After checking and fixing mistakes I will edit the russian pages.
Part 2 for the same PR is here https://github.com/ClickHouse/ClickHouse/pull/194... | 2021-01-23T13:47:21 | 2021-01-23T13:47:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,177,357 | 19,497 | alexey-milovidov | Looks like you are using modified version of ClickHouse. The official release has:
```
ClickHouse client version 21.1.2.15 (official build).
SELECT buildId()
┌─buildId()────────────────────────────────┐
│ E40526A12E9A8F3819A18694F6B798F10C624D5C │
└──────────────────────────────────────────┘
```
Where d... | 2021-01-23T20:52:30 | 2021-01-23T20:52:59 | {} | MEMBER |
ClickHouse | ClickHouse | 766,276,447 | 19,497 | sundy-li | > Looks like you are using modified version of ClickHouse. The official release has:
>
> ```
> ClickHouse client version 21.1.2.15 (official build).
>
> SELECT buildId()
>
> ┌─buildId()────────────────────────────────┐
> │ E40526A12E9A8F3819A18694F6B798F10C624D5C │
> └───────────────────────────────────────... | 2021-01-24T02:15:48 | 2021-01-24T02:15:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,395,165 | 19,523 | alexey-milovidov | @den-crane I'm in doubt with this change. | 2021-01-24T17:04:57 | 2021-01-24T17:04:57 | {} | MEMBER |
ClickHouse | ClickHouse | 766,433,384 | 19,517 | MaxJPowers | Thanks, this makes a lot of sense I hadn't found anything to do with generating a table using the Dictionary engine but now my tables are loaded fine. It seems that when I do dictGet('cacheDictionary','random0',toUInt64(2)) it loads the entirety of the table, is this usual? (I've turned off lazy_load). | 2021-01-24T21:14:59 | 2021-01-24T21:14:59 | {} | NONE |
ClickHouse | ClickHouse | 766,159,512 | 19,489 | bharatnc | > Thank you for your PR
@vdimir you're welcome! | 2021-01-23T18:46:30 | 2021-01-23T18:46:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,195,685 | 19,525 | olgarev | @alexey-milovidov Will you review this, pls? | 2021-01-23T23:00:43 | 2021-01-23T23:00:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,418,082 | 19,466 | alexey-milovidov | Ok. Also I see that in some places NO_SANITIZE_UNDEFINED becomes irrelevant. | 2021-01-24T19:29:56 | 2021-01-24T19:29:56 | {} | MEMBER |
ClickHouse | ClickHouse | 766,436,289 | 18,346 | den-crane |
```sql
create table items ( id Int64, country String, state String) Engine=MergeTree order by tuple();
insert into items select number , toString(cityHash64(number)) ,
toString(cityHash64(number)) from numbers(200000000);
SELECT id,any(country) as country, any(state) as state
FROM items GROUP BY id format... | 2021-01-24T21:35:10 | 2021-01-24T21:35:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,472,113 | 19,542 | l1t1 | connect by client, the system.numbers_mt result is aslo wrong, `select * from system.numbers limit 10000000000` is right
```sql
:~/ch21$ ./clickhouse client -mA
ClickHouse client version 21.1.1.5745 (official build).
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 21.1.1 revisi... | 2021-01-25T00:41:43 | 2021-01-25T00:45:06 | {} | NONE |
ClickHouse | ClickHouse | 766,505,063 | 19,542 | alexey-milovidov | `select * from system.numbers_mt limit 10000000000`
Does not necessarily gives you numbers from 0 to 10000000000-1, because the order of data processing is non-deterministic.
Write `SELECT * FROM numbers_mt(10000000000)` instead. | 2021-01-25T02:49:15 | 2021-01-25T02:49:15 | {} | MEMBER |
ClickHouse | ClickHouse | 766,577,711 | 16,336 | JackyWoo | I hive the same issue with version 20.6.6.7.
@Fanduzi does it work with higher ch version? | 2021-01-25T06:27:40 | 2021-01-25T06:37:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,612,036 | 19,108 | qoega | Another one? https://clickhouse-test-reports.s3.yandex.net/19529/61e93d15a8f36702920238a93bfeee64f80715c7/functional_stateless_tests_(antlr_debug).html#fail1 | 2021-01-25T07:34:06 | 2021-01-25T07:34:06 | {} | MEMBER |
ClickHouse | ClickHouse | 765,629,681 | 19,403 | den-crane | try `set aggregation_memory_efficient_merge_threads=2` | 2021-01-22T19:17:03 | 2021-01-22T19:17:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 765,676,923 | 19,360 | alexey-milovidov | It has the infamous error message:
```
E subprocess.CalledProcessError: Command '['docker-compose', '--project-directory', '/ClickHouse/tests/integration/test_polymorphic_parts', '--project-name', 'roottestpolymorphicparts', '--file', '/ClickHouse/tests/integration/test_polymorphic_parts/_instances/no... | 2021-01-22T20:50:42 | 2021-01-22T20:50:42 | {} | MEMBER |
ClickHouse | ClickHouse | 765,685,674 | 19,444 | alexey-milovidov | > 00158_cache_dictionary_has | FAIL | 600.38
Timeout :( @nikitamikhaylov
| 2021-01-22T21:10:04 | 2021-01-22T21:10:04 | {} | MEMBER |
ClickHouse | ClickHouse | 765,689,348 | 19,444 | alexey-milovidov | It's a good time to adjust the test... could you please send a PR? | 2021-01-22T21:18:26 | 2021-01-22T21:18:26 | {} | MEMBER |
ClickHouse | ClickHouse | 765,708,606 | 19,451 | azat | And AFAICS MergeTree engine is safe | 2021-01-22T22:00:07 | 2021-01-22T22:00:17 | {} | MEMBER |
ClickHouse | ClickHouse | 766,155,975 | 19,489 | vdimir | Thank you for your PR | 2021-01-23T18:21:25 | 2021-01-23T18:21:25 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 766,511,325 | 19,542 | l1t1 | numbers_mt still has overflow problem ,toInt128 is much slower
```sql
:) SELECT max(number) FROM numbers_mt(10000000000);
SELECT max(number)
FROM numbers_mt(10000000000)
Query id: dd355d04-528a-4ec8-812b-8d99b4895a10
┌─max(number)─┐
│ 9999999999 │
└─────────────┘
1 rows in set. Elapsed: 2.512 sec. Pro... | 2021-01-25T03:11:29 | 2021-01-25T03:11:29 | {} | NONE |
ClickHouse | ClickHouse | 766,608,089 | 19,407 | templarzq | @qoega 00107_totals_after_having The reference file result does not match the sql result. | 2021-01-25T07:26:23 | 2021-01-25T07:26:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,608,545 | 19,407 | templarzq | @qoega and 00919_sum_aggregate_states_constants has problem too. | 2021-01-25T07:27:22 | 2021-01-25T07:27:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,619,047 | 19,551 | alesapin | Same as https://github.com/ClickHouse/ClickHouse/pull/18970 | 2021-01-25T07:47:47 | 2021-01-25T07:47:47 | {} | MEMBER |
ClickHouse | ClickHouse | 765,651,880 | 19,412 | alesapin | Fuzzer debug:
`2021.01.22 16:28:15.128031 [ 82 ] {} <Fatal> Application: Child process was terminated by signal 11.`
And nothing more. | 2021-01-22T20:00:42 | 2021-01-22T20:00:42 | {} | MEMBER |
ClickHouse | ClickHouse | 766,193,422 | 19,521 | alexey-milovidov | Atomic database engine is production grade and is used by default.
Please remove the mentions that it's experimental from the docs. | 2021-01-23T22:40:19 | 2021-01-23T22:40:19 | {} | MEMBER |
ClickHouse | ClickHouse | 766,434,343 | 19,524 | alexey-milovidov | Looks like it's working but there are no segfaults. Will restart... | 2021-01-24T21:22:04 | 2021-01-24T21:22:04 | {} | MEMBER |
ClickHouse | ClickHouse | 765,706,404 | 19,403 | alexey-milovidov | Also need to enable `distributed_aggregation_memory_efficient` first. It's not enabled by default. | 2021-01-22T21:54:36 | 2021-01-22T21:54:36 | {} | MEMBER |
ClickHouse | ClickHouse | 765,811,751 | 19,409 | den-crane | > If this happens, I can't connect to the current Clickhouse server with clickhouse client
>
it explains nothing. It does not help.
Need logs from CH server | 2021-01-23T01:33:48 | 2021-01-23T01:33:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,158,098 | 17,847 | alexey-milovidov | @fibersel Yes, that's what we need, except that measurement should be done for each of the data files.
(we should choose a set of about 5..10 data files that most likely have very different data distribution / data locality)
| 2021-01-23T18:37:27 | 2021-01-23T18:37:27 | {} | MEMBER |
ClickHouse | ClickHouse | 766,175,608 | 19,440 | alexey-milovidov | No longer supported. | 2021-01-23T20:39:16 | 2021-01-23T20:39:16 | {} | MEMBER |
ClickHouse | ClickHouse | 766,177,844 | 19,477 | alexey-milovidov | It has parsed FROM as function.
Note that our parser does not have keywords at all...
I can add a quirk to the parser, so FROM can never be parsed as a function. | 2021-01-23T20:57:05 | 2021-01-23T20:57:16 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 766,393,917 | 19,258 | kitaisreal | @MaxJPowers please check my answer here https://github.com/ClickHouse/ClickHouse/issues/19517. | 2021-01-24T16:57:34 | 2021-01-24T16:57:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,487,412 | 19,542 | bohutang | There are parallel workers under the numbers_mt source generator, it not works well with limit clause.
` set max_threads=1` will be work, but slowly.
| 2021-01-25T01:44:17 | 2021-01-25T01:44:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,494,356 | 18,813 | tkionshao | > @tkionshao Did you check the newer version?
Still happened, but moved the image snapshot to other host is working.
I think this is our environment issue, thanks for help. | 2021-01-25T02:10:15 | 2021-01-25T02:10:15 | {} | NONE |
ClickHouse | ClickHouse | 766,678,931 | 18,945 | qoega | 01605_adaptive_granularity_block_borders
https://clickhouse-test-reports.s3.yandex.net/0/303db08231f19ad5739551991fc23a63012a2773/fast_test.html | 2021-01-25T09:27:26 | 2021-01-25T09:27:26 | {} | MEMBER |
ClickHouse | ClickHouse | 765,728,040 | 19,441 | alexey-milovidov | True bug, it may cause crash. | 2021-01-22T22:44:29 | 2021-01-22T22:44:29 | {} | MEMBER |
ClickHouse | ClickHouse | 766,374,276 | 19,517 | kitaisreal | Hello, @MaxJPowers, thanks for the example.
Currently Cache, ComplexCache, SSDCache, SSDComplexCache dictionaries access and cache received keys only during source access with `dictGet`, `dictHas` commands.
When you SELECT * FROM cache_dictionary_table, if no keys were fetched from source yet, there will be no re... | 2021-01-24T16:09:05 | 2021-01-24T16:09:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,394,949 | 19,531 | winoenix | > That's true. Atomic database is default from version 20.10.
>
> It gives the following benefits: lock-free table CREATE/DROP/RENAME, allows EXCHANGE and REPLACE queries.
>
> Drawbacks: some external tools may not be ready for data layout on filesystem.
>
> You can also create database with ENGINE specified e... | 2021-01-24T17:03:34 | 2021-01-24T17:03:34 | {} | NONE |
ClickHouse | ClickHouse | 766,691,079 | 19,344 | tavplubix | AST fuzzer (ASan) - #19108
AST fuzzer (UBSan) - #19566 | 2021-01-25T09:46:44 | 2021-01-25T09:46:44 | {} | MEMBER |
ClickHouse | ClickHouse | 766,087,838 | 19,513 | jyz0309 | I am new to ClickHouse Community. When I was reading the source code and wanted to try to contribute code to ClickHouse, I find this README.md date error, so I fix it... if it is too easy(or make you plague) to become a contributor through this pr,I will close this pull request. And try to contribute code to ClickHouse... | 2021-01-23T14:27:49 | 2021-01-23T14:30:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,305,331 | 19,375 | azat | >AST fuzzer (UBSan) — ../src/Functions/FunctionDateOrDateTimeAddInterval.h:92:26: runtime error: signed integer overflow: -9223372036854775807 * 60 cannot be repr
#19466 | 2021-01-24T07:31:50 | 2021-01-24T07:31:50 | {} | MEMBER |
ClickHouse | ClickHouse | 766,394,558 | 19,531 | alexey-milovidov | That's true. Atomic database is default from version 20.10.
It gives the following benefits: lock-free table CREATE/DROP/RENAME, allows EXCHANGE and REPLACE queries.
Drawbacks: some external tools may not be ready for data layout on filesystem.
You can also create database with ENGINE specified explicitly, e.g... | 2021-01-24T17:01:17 | 2021-01-24T17:01:17 | {} | MEMBER |
ClickHouse | ClickHouse | 766,635,111 | 18,236 | kitaisreal | > `getColumn` is copy-pasted in each Dictionary class. Maybe we can move it to the interface?
It is impossible to move into interface but I reduced boilerplate with helper classes in DictionaryHelpers.h. | 2021-01-25T08:16:11 | 2021-01-25T08:16:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 765,651,524 | 16,841 | SaltTan | I got a few more segfaults on 20.12.5, all from MergeTreeSequentialSource threads, but with different stack traces.
I suspect the hardware now. | 2021-01-22T19:59:59 | 2021-01-22T19:59:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 765,704,859 | 19,467 | alexey-milovidov | It's just a preparation to fix the error:
```
Too large size (18446618729448839312) passed to allocator. It indicates an error.: While processing encrypt('aes-128-ecb', [1, -1, 0, NULL], 'text')
``` | 2021-01-22T21:51:02 | 2021-01-22T21:51:02 | {} | MEMBER |
ClickHouse | ClickHouse | 765,807,617 | 19,409 | songenjie | If this happens, I can't connect to the current Clickhouse server with clickhouse client
> can you show ping (latency) from one clickhouse server to another ?
>
> try to add setting connect_timeout_with_failover_ms=1000
>
> ```
> cat /etc/clickhouse-server/users.d/connect_timeout_with_failover_ms.xml
>
>
... | 2021-01-23T01:11:19 | 2021-01-23T01:15:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 765,878,439 | 19,392 | sasha-id | Cool, I can reference values within SELECT clause in CH.
But still this part `SELECT * FROM items ORDER BY group_id desc, time desc LIMIT 100 BY group_id` will scan whole table regardless the LIMIT, the bigger dataset the slower it will be.
| 2021-01-23T06:43:53 | 2021-01-23T06:43:53 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 765,884,107 | 19,450 | azat | >AST fuzzer (UBSan) — ../src/Core/Types.h:163:58: runtime error: signed integer overflow: -1967580000 + -655840000 cannot be represented in type 'int'
Looks like inlining should help - https://github.com/ClickHouse/ClickHouse/pull/19466/commits/9fb55fed77b8057510fc387aa2287f8627adf8ed (added to #19466) | 2021-01-23T07:45:03 | 2021-01-23T07:45:23 | {} | MEMBER |
ClickHouse | ClickHouse | 765,968,555 | 19,490 | kitaisreal | accurateCastOrNull function was not implemented in 20.10 | 2021-01-23T12:41:06 | 2021-01-23T12:41:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,313,904 | 19,527 | ucasfl | > @ucasFL could you please add documentation to this `sign` function ?
Ok. | 2021-01-24T08:59:07 | 2021-01-24T08:59:07 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 766,485,494 | 19,543 | den-crane | MergeTree() -- requires table `order by ` section, all rows in MergeTree.
You can use a dummy order -- tuple() function.
```sql
CREATE TABLE trips_mergetree
ENGINE = MergeTree()
order by tuple() <<<<----
as select * from trips order by dropoff_datetime;
```
But better to use proper partitioning and proper... | 2021-01-25T01:36:13 | 2021-01-25T01:36:42 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,645,485 | 19,396 | qoega | > It's 10G or 10Gi, not 10GB.
> That's reasonable, because it's applicable for any numbers.
There is a nice place to improve error text for a user. | 2021-01-25T08:32:00 | 2021-01-25T08:32:00 | {} | MEMBER |
ClickHouse | ClickHouse | 766,652,443 | 19,517 | kitaisreal | @MaxJPowers could you please provide example of your cache dictionary and source configuration ? Thanks. | 2021-01-25T08:44:07 | 2021-01-25T08:44:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 782,831,380 | 17,939 | alexey-milovidov | `neighbor` function is tricky to use - it can obtain neighbour values only within a block.
But the concept of "block" is hidden from user. It also depends on internal details of data processing. And for higher-order functions these details are very specific. The behaviour can be explained but the user cannot rely on... | 2021-02-21T10:04:08 | 2021-02-21T10:04:22 | {} | MEMBER |
ClickHouse | ClickHouse | 782,853,017 | 21,030 | alexey-milovidov | The error message is correct.
See
```
SELECT [9223372036854775807, -9223372036854775808]
``` | 2021-02-21T12:50:27 | 2021-02-21T12:50:27 | {} | MEMBER |
ClickHouse | ClickHouse | 782,908,939 | 20,839 | alexey-milovidov | Not important enough. | 2021-02-21T19:09:12 | 2021-02-21T19:09:12 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.