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 | 670,792,095 | 13,478 | bohutang | (revised version)
This may be caused by the charset.
RESPONSE FROM MySQL to MySQL the charset is <b>latin1</b> and the Packet Length: 35:
```
... ...
MySQL Protocol
Packet Length: 59
Packet Number: 2
Catalog: def
Database: ch_test
Table: user_data
Original table: user_data
Name... | 2020-08-08T00:11:45 | 2020-08-08T14:32:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,941,683 | 13,478 | bohutang | I have a check with MySQL client under utf8 charset, the response payload is same as your CH_request.pcapng.
The client:
```
mysql -t --default-character-set utf8 -uroot -p123 -h172.17.0.2 -e "select secret from ch_test.user_data"
```
Payload traffic capature with:
```
sudo tcpdump -i docker0 -X -s 0 -w /tmp/ ... | 2020-08-08T15:29:00 | 2020-08-08T15:36:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,981,202 | 13,446 | alexey-milovidov | Query Fuzzer - it's about #13283 | 2020-08-08T22:39:36 | 2020-08-08T22:39:36 | {} | MEMBER |
ClickHouse | ClickHouse | 670,984,414 | 13,529 | alexey-milovidov | `shellcheck --check-sourced --external-sources --severity info --exclude SC1071,SC2086 *.sh ../1_stateful/*.sh` | 2020-08-08T23:23:53 | 2020-08-08T23:23:53 | {} | MEMBER |
ClickHouse | ClickHouse | 671,018,564 | 13,360 | mjodeh | when I execute `select databse, progress, elapsed, num_parts from system.merges` I see the same database in multiple rows, is it ok?
| database | progress | elapsed | num_parts |
| database1 │ 0.7091122321270097 │ 84.209726576 │ 17 │
│ database1 │ 0.9967329158725837 │ 80.834683612 │ ... | 2020-08-09T07:32:12 | 2020-08-09T07:32:12 | {} | NONE |
ClickHouse | ClickHouse | 670,190,383 | 13,416 | alexey-milovidov | > Have no idea how to write a test.
Insert by single row in one thread.
Truncate in other thread periodically.
Select in third thread.
Should reproduce easily under TSan. | 2020-08-06T21:02:45 | 2020-08-06T21:02:45 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 670,398,513 | 13,418 | ghuname | I hoped that I will be able to use the same {shard} variable placeholder because it is very convenient if you are creating tables (for example kafka, view, shard, distributive tables for many kafka streams, each stream these four tables) by using script automation.
Because the fact that cluster is specified at the ... | 2020-08-07T08:27:23 | 2020-08-07T08:27:23 | {} | NONE |
ClickHouse | ClickHouse | 670,434,311 | 13,318 | luoqg | I also have this problem。Continuous automatic restart
```
2020.08.07 17:42:29.060255 [ 10662 ] {} <Debug> Application: rlimit on number of file descriptors is 500000
2020.08.07 17:42:29.060271 [ 10662 ] {} <Debug> Application: Initializing DateLUT.
2020.08.07 17:42:29.060282 [ 10662 ] {} <Trace> Application: Init... | 2020-08-07T09:50:51 | 2020-08-07T10:59:31 | {} | NONE |
ClickHouse | ClickHouse | 670,467,792 | 12,741 | filimonov | Here is the test:
https://gist.github.com/filimonov/c84f48f91c155ab6afb91af33f49b084
query conditions:
```
WHERE id = 80000 and date = '2020-08-03' and team='team3'
```
row-level security conditions
```
(team = 'team3') OR ((team = 'team1') AND (user IN ('aaa', 'bbb', 'ccc', 'ddd'))
```
result
```
PR... | 2020-08-07T11:18:33 | 2020-08-07T11:21:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,610,593 | 13,450 | KochetovNicolai | So, we add counter to zookeeper. Looks like it won't brake any compatibility, which is cool. | 2020-08-07T16:51:33 | 2020-08-07T16:51:33 | {} | MEMBER |
ClickHouse | ClickHouse | 670,807,746 | 13,497 | alexey-milovidov | Need to check/update tests... | 2020-08-08T02:03:13 | 2020-08-08T02:03:13 | {} | MEMBER |
ClickHouse | ClickHouse | 670,932,428 | 13,033 | alexey-milovidov | Yes, it will be perfect. | 2020-08-08T14:00:27 | 2020-08-08T14:00:27 | {} | MEMBER |
ClickHouse | ClickHouse | 670,217,785 | 13,359 | filimonov | #8004 | 2020-08-06T22:14:17 | 2020-08-06T22:14:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,228,424 | 13,343 | akuzm | The reason the performance test fails is that now we can't read an empty file as TSVWithNamesAndTypes:
```
$ clickhouse-local --query "
create table t engine File(TSVWithNamesAndTypes, '1.tsv') as
select * from numbers(1) where number < 0;
select * from t;
"
Code: 117, e.displayText() = DB::Excep... | 2020-08-06T22:48:06 | 2020-08-06T22:48:06 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,268,595 | 13,228 | alexey-milovidov | The query
```
SELECT CAST('0.12345678901234567890123456789012345678', 'Decimal(38,37)');
```
gives different result in "Unbundled" mode. | 2020-08-07T01:17:57 | 2020-08-07T01:17:57 | {} | MEMBER |
ClickHouse | ClickHouse | 670,275,032 | 10,066 | alexey-milovidov | @alesapin Looks like it's fixed? | 2020-08-07T01:43:34 | 2020-08-07T01:43:34 | {} | MEMBER |
ClickHouse | ClickHouse | 670,412,462 | 13,399 | KochetovNicolai | I generally ok with this change, however using other formats still get error, e.g.:
```
$ clickhouse client -q "SELECT 1 AS x LIMIT 0 FORMAT JSONEachRow" | clickhouse client -q 'insert into tab format JSONEachRow'
Code: 108. DB::Exception: No data to insert
$ clickhouse client -q "SELECT 1 AS x LIMIT 0 FORMAT TSV... | 2020-08-07T08:59:24 | 2020-08-07T08:59:24 | {} | MEMBER |
ClickHouse | ClickHouse | 670,467,918 | 13,190 | blinkov | @Slach it was waiting for test results, but I haven't noticed that this PR didn't have the label that's supposed to launch them. | 2020-08-07T11:18:53 | 2020-08-07T11:18:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,471,098 | 13,453 | alexey-milovidov | How it should work according to SQL standard? | 2020-08-07T11:27:42 | 2020-08-07T11:27:42 | {} | MEMBER |
ClickHouse | ClickHouse | 670,731,168 | 13,076 | akuzm | 00933_test_fix_extra_seek_on_compressed_cache is in CI Fest #7
fuzzer failure is https://github.com/ClickHouse/ClickHouse/issues/13392 | 2020-08-07T22:04:26 | 2020-08-07T22:04:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,732,882 | 13,336 | akuzm | perf test is broken in master | 2020-08-07T22:05:39 | 2020-08-07T22:05:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,922,520 | 13,033 | vladimir-golovchenko | @alexey-milovidov As a sample of a function should be taken [Postgress date_trunc](https://www.postgresql.org/docs/9.1/functions-datetime.html#FUNCTIONS-DATETIME-TRUNC)?
The signature should look like this? -
```sql
date_trunc(datepart, time_or_data, [timezone])
/*
- datepart - the name of datepart or timepart... | 2020-08-08T12:18:10 | 2020-08-08T12:22:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 671,076,982 | 13,515 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/issues/13546 | 2020-08-09T17:08:40 | 2020-08-09T17:08:40 | {} | MEMBER |
ClickHouse | ClickHouse | 671,095,864 | 13,099 | alexey-milovidov | Performance test Ok. | 2020-08-09T20:13:46 | 2020-08-09T20:13:46 | {} | MEMBER |
ClickHouse | ClickHouse | 670,489,484 | 13,405 | excitoon | This PR looks okay but needs tuning of tests or configuration because now `maxConnections` is 25 instead of infinity. | 2020-08-07T12:20:49 | 2020-08-07T12:20:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,583,300 | 13,491 | KochetovNicolai | Waiting for perftests | 2020-08-07T15:49:26 | 2020-08-07T15:49:26 | {} | MEMBER |
ClickHouse | ClickHouse | 670,650,511 | 12,081 | alexey-milovidov | Version 20.6 is released as stable.
The delay of 25 days was due to extended testing in Yandex.Metrica and Yandex Banner System that has to be performed manually and has revealed some additional issues that were fixed.
Changelog is being prepared #13421 | 2020-08-07T18:27:58 | 2020-08-07T18:28:28 | {} | MEMBER |
ClickHouse | ClickHouse | 670,730,750 | 13,392 | akuzm | `SELECT has(materialize([9223372036854775807, NULL, 256]), materialize(NULL))` | 2020-08-07T22:04:10 | 2020-08-07T22:04:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,934,035 | 13,516 | alexey-milovidov | > AST-based query fuzzer — Logical error: 'Block structure mismatch in PipelineExecuting stream: different columns:
@akuzm I cannot reproduce the issue found by Query Fuzzer. | 2020-08-08T14:16:15 | 2020-08-08T14:16:15 | {} | MEMBER |
ClickHouse | ClickHouse | 671,144,721 | 2,675 | jywjyw | Though "kafka_skip_broken_messages=65535" is effective, but if kafka receive a NULL msg, the CH will consume the bad msg repeatly. I found this bug in Ver 20.3. | 2020-08-10T03:14:40 | 2020-08-10T03:14:40 | {} | NONE |
ClickHouse | ClickHouse | 670,229,026 | 13,417 | akuzm | Broken build and perf test is just our CI being dead today, I'll merge. | 2020-08-06T22:50:16 | 2020-08-06T22:50:16 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,588,370 | 13,493 | RikeVoltz | Also, it needs to be investigated if the Resample combinator works with the DateTime type at all. | 2020-08-07T16:00:49 | 2020-08-07T16:00:49 | {} | NONE |
ClickHouse | ClickHouse | 670,704,189 | 13,404 | excitoon | I've just found out that proxy resolver is not being checked for host filter as well. :-( | 2020-08-07T20:40:13 | 2020-08-07T20:40:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,472,773 | 13,392 | alexey-milovidov | Segfault:
```
SELECT NULL = countEqual(materialize([(number - 2.) - 1., NULL AS x]), materialize(x))
FROM system.numbers
LIMIT 1
``` | 2020-08-07T11:32:19 | 2020-08-07T11:32:19 | {} | MEMBER |
ClickHouse | ClickHouse | 670,530,109 | 13,147 | akuzm | Perf test broken in master. | 2020-08-07T13:57:44 | 2020-08-07T13:57:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,791,984 | 8,986 | alexey-milovidov | I have to clarify that I'm still afraid merging it because I'm not sure about the "version" interface for tables. We have to think more... | 2020-08-08T00:11:05 | 2020-08-08T00:11:05 | {} | MEMBER |
ClickHouse | ClickHouse | 670,806,716 | 12,003 | alexey-milovidov | #13511
#13513 | 2020-08-08T01:53:22 | 2020-08-08T01:53:22 | {} | MEMBER |
ClickHouse | ClickHouse | 670,983,493 | 3,011 | alexey-milovidov | Fixed in recent releases. | 2020-08-08T23:10:17 | 2020-08-08T23:10:17 | {} | MEMBER |
ClickHouse | ClickHouse | 670,985,571 | 13,515 | alexey-milovidov | @amosbird It is also needed to simplify https://clickhouse.tech/docs/en/getting-started/install/#from-binaries-non-linux | 2020-08-08T23:43:09 | 2020-08-08T23:43:09 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 671,076,787 | 13,515 | alexey-milovidov | test_storage_rabbitmq - bad test. | 2020-08-09T17:06:30 | 2020-08-09T17:06:30 | {} | MEMBER |
ClickHouse | ClickHouse | 671,076,821 | 13,515 | alexey-milovidov | 00933_test_fix_extra_seek_on_compressed_cache - bad test. | 2020-08-09T17:06:50 | 2020-08-09T17:06:50 | {} | MEMBER |
ClickHouse | ClickHouse | 671,076,840 | 13,515 | alexey-milovidov | 01193_metadata_loading - bad test. | 2020-08-09T17:07:00 | 2020-08-09T17:07:00 | {} | MEMBER |
ClickHouse | ClickHouse | 669,390,759 | 13,390 | qoega | > If you want to allow admin to change access rights of anyone except admin himself, the current implementation of GRANT/REVOKE commands doesn't support that.
Lets convert it to feature request then? | 2020-08-05T18:41:46 | 2020-08-09T18:32:05 | {} | MEMBER |
ClickHouse | ClickHouse | 670,253,106 | 13,378 | alexey-milovidov | Some unneeded submodules remain, but they are small. | 2020-08-07T00:17:05 | 2020-08-07T00:17:05 | {} | MEMBER |
ClickHouse | ClickHouse | 670,458,641 | 13,190 | Slach | @blinkov could you merge this PR? or we need to wait someone else? | 2020-08-07T10:54:14 | 2020-08-07T10:54:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,590,108 | 13,448 | qoega | Failed as it tries to get `latest` images.
```
Pulling golang1 (yandex/clickhouse-mysql-golang-client:)...
manifest for yandex/clickhouse-mysql-golang-client:latest not found
```
@alesapin Can you push new images to docker hub with latest tag to check tests are OK? | 2020-08-07T16:04:27 | 2020-08-07T16:04:27 | {} | MEMBER |
ClickHouse | ClickHouse | 671,095,704 | 13,099 | alexey-milovidov | 00933_test_fix_extra_seek_on_compressed_cache - known bad test. | 2020-08-09T20:12:14 | 2020-08-09T20:12:14 | {} | MEMBER |
ClickHouse | ClickHouse | 670,387,153 | 13,446 | qoega | There should be no schema in netloc
https://stackoverflow.com/a/53993037
https://tools.ietf.org/html/rfc1808.html#ref-2
``` sql
SELECT netloc('http://paul@www.example.com:80/') AS Netloc;
2020-08-07 05:03:43 -paul@www.example.com:80
2020-08-07 05:03:43 +http://paul@www.example.com:80
``` | 2020-08-07T08:00:32 | 2020-08-07T08:00:32 | {} | MEMBER |
ClickHouse | ClickHouse | 670,440,144 | 13,451 | vladimir-golovchenko | Could you approve that all changes are correct to I can propagate these changes to docs of other languages? | 2020-08-07T10:05:10 | 2020-08-07T10:05:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 671,077,643 | 13,539 | alexey-milovidov | Functional tests Ok. | 2020-08-09T17:14:40 | 2020-08-09T17:14:40 | {} | MEMBER |
ClickHouse | ClickHouse | 671,092,537 | 13,542 | filimonov | Check AvroConfluent format | 2020-08-09T19:37:45 | 2020-08-09T19:37:45 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,279,309 | 13,327 | templarzq | ok,thanks. | 2020-08-07T02:00:57 | 2020-08-07T02:00:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,430,914 | 13,399 | hczhcz | A consideration in this PR is, an empty file is not a legal Parquet/Arrow file. When we pipe the result to somewhere else, it breaks. In comparison, an empty TSV/CSV/JSONL file is okay. (I am not sure about some other formats.)
| 2020-08-07T09:42:11 | 2020-08-07T09:42:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,470,704 | 13,453 | alexey-milovidov | I have checked how it works in other DBMS just in case.
MySQL:
`SELECT * FROM (SELECT 1 AS x, 1 AS x) t`
> Duplicate column name 'x'
`SELECT * FROM (SELECT 1, 1) t`
> Duplicate column name '1'
`SELECT * FROM (SELECT avg(x), avg(x) FROM (SELECT 1 AS x) t1) t2`
> Duplicate column name 'avg(x)'
Postgres:... | 2020-08-07T11:26:39 | 2020-08-07T11:26:47 | {} | MEMBER |
ClickHouse | ClickHouse | 670,547,119 | 13,360 | mjodeh | Thanks for reply @den-crane
> What table persists in system.merges ? What is the insertion rate into this table?
There is no single table which persists in system.merges since every db has only 1 table and the insertion rate is 1 insertion per db every 10 seconds, system.merges keeps changing but always shows ... | 2020-08-07T14:32:58 | 2020-08-07T14:34:09 | {} | NONE |
ClickHouse | ClickHouse | 670,701,609 | 12,527 | vitlibar | Fixed. | 2020-08-07T20:33:05 | 2020-08-07T20:33:05 | {} | MEMBER |
ClickHouse | ClickHouse | 670,981,373 | 13,491 | alexey-milovidov | It shows 10..17% improvement of memory usage, is it real? @akuzm | 2020-08-08T22:41:43 | 2020-08-08T22:41:43 | {} | MEMBER |
ClickHouse | ClickHouse | 670,419,409 | 13,406 | qoega | Can you reproduce it with single TSV file? If so can you share its contents or minimize and obfuscate example to reproduce it in tests? | 2020-08-07T09:14:47 | 2020-08-07T09:14:47 | {} | MEMBER |
ClickHouse | ClickHouse | 670,526,670 | 13,147 | akuzm | Fuzz test: https://github.com/ClickHouse/ClickHouse/issues/13477 | 2020-08-07T13:50:20 | 2020-08-07T13:50:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,613,987 | 13,337 | KochetovNicolai | ` 7 faster` - that's unusual | 2020-08-07T16:59:36 | 2020-08-07T16:59:36 | {} | MEMBER |
ClickHouse | ClickHouse | 670,995,211 | 9,624 | CurtizJ | Build was broken in cf255ae. | 2020-08-09T02:03:32 | 2020-08-09T02:03:32 | {} | MEMBER |
ClickHouse | ClickHouse | 671,092,421 | 13,543 | filimonov | Use 20.5 and newer and JSONAsString format. After that parse the string with JSON functions | 2020-08-09T19:36:41 | 2020-08-09T19:36:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,382,976 | 10,066 | alesapin | I'm not sure. We still can get here https://github.com/clickhouse/ClickHouse/blob/master/src/Storages/MergeTree/IMergeTreeDataPart.cpp#L266 for the temporary part and remove the same path during `DROP`. | 2020-08-07T07:50:49 | 2020-08-07T07:50:49 | {} | MEMBER |
ClickHouse | ClickHouse | 670,528,913 | 13,147 | akuzm | Stress test failure under msan is in netloc https://github.com/ClickHouse/ClickHouse/issues/13335 | 2020-08-07T13:55:19 | 2020-08-07T13:55:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,534,387 | 13,451 | vladimir-golovchenko | @alexey-milovidov I updated the docs of other languages.
FYI the doc in Russian had already been updated and it is a little different from others in *clang* related sentence:
> Компилятор - `gcc`. На данный момент (август 2020), код собирается версией 9.3. (Также код может быть собран `clang` версий 10 и 9) | 2020-08-07T14:05:38 | 2020-08-07T14:05:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,980,975 | 13,451 | alexey-milovidov | We use clang-10 in CI but versions 9 and 8 are also seems to work. | 2020-08-08T22:36:02 | 2020-08-08T22:36:02 | {} | MEMBER |
ClickHouse | ClickHouse | 671,143,936 | 13,248 | alexey-milovidov | This simple query works on 19.14 but segfaults on 20.3..20.8:
`SELECT count() FROM test.hits WHERE (CounterID <= 1025) AND (CounterID > 107931)` | 2020-08-10T03:09:42 | 2020-08-10T03:09:42 | {} | MEMBER |
ClickHouse | ClickHouse | 670,275,390 | 5,182 | alexey-milovidov | #13444 | 2020-08-07T01:45:08 | 2020-08-07T01:45:08 | {} | MEMBER |
ClickHouse | ClickHouse | 670,405,393 | 13,332 | millin | cc @blinkov | 2020-08-07T08:43:37 | 2020-08-07T08:43:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,520,002 | 13,360 | den-crane | >1 insertion per server not per database
yes. But this number 1 part / per second / per server is highly approximate value.
It depends on many parameters, number of columns, partitioning schema, volume of data in bytes, disks.
`optimize table` makes everything worse. It causes unplanned merges and overloads I/O... | 2020-08-07T13:35:52 | 2020-08-07T13:35:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,528,862 | 13,335 | akuzm | msan failure in stress test: `SELECT toIntervalMonth((CAST((netloc('=nD')) AS DateTime))); ` | 2020-08-07T13:55:13 | 2020-08-07T13:55:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,983,849 | 13,497 | alexey-milovidov | @amosbird Let's check, maybe the task https://github.com/ClickHouse/ClickHouse/issues/10977 is also resolved by this PR? | 2020-08-08T23:15:35 | 2020-08-08T23:15:35 | {} | MEMBER |
ClickHouse | ClickHouse | 671,079,122 | 13,539 | alexey-milovidov | TestFlows: @traceon confirmed that there is an issue with tests. | 2020-08-09T17:27:27 | 2020-08-09T17:27:27 | {} | MEMBER |
ClickHouse | ClickHouse | 670,272,477 | 13,281 | alexey-milovidov | https://llvm.org/doxygen/LoopIdiomRecognize_8cpp_source.html | 2020-08-07T01:33:14 | 2020-08-07T01:33:14 | {} | MEMBER |
ClickHouse | ClickHouse | 670,382,082 | 13,424 | spihiker | I also have the same problem .But the SQL is different. | 2020-08-07T07:48:40 | 2020-08-07T07:48:40 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 670,472,324 | 13,399 | alexey-milovidov | @KochetovNicolai It's unusual to insert empty dataset and it may indicate an error.
Let's keep this exception for convenience. | 2020-08-07T11:31:06 | 2020-08-07T11:31:06 | {} | MEMBER |
ClickHouse | ClickHouse | 670,777,659 | 8,482 | alexey-milovidov | Blocked by #13281. | 2020-08-07T23:05:18 | 2020-08-07T23:05:18 | {} | MEMBER |
ClickHouse | ClickHouse | 670,807,301 | 13,502 | alexey-milovidov | It does not work on ARM: https://clickhouse-builds.s3.yandex.net/13502/df79f415ce24e650244150bc399ac083def3d05d/clickhouse_build_check/report.html#fail1 | 2020-08-08T01:59:15 | 2020-08-08T01:59:30 | {} | MEMBER |
ClickHouse | ClickHouse | 670,218,945 | 7,666 | alexey-milovidov | @KochetovNicolai Ok, let's do it.
I see that the `isDeterministicInScopeOfQuery` property is only used for `JIT` and `RedundantFunctionsInOrderByVisitor`. | 2020-08-06T22:17:49 | 2020-08-06T22:17:49 | {} | MEMBER |
ClickHouse | ClickHouse | 670,461,424 | 13,318 | alexey-milovidov | Is there any message in `dmesg` about clickhouse or OOM? | 2020-08-07T11:01:31 | 2020-08-07T11:01:48 | {} | MEMBER |
ClickHouse | ClickHouse | 670,472,111 | 13,416 | KochetovNicolai | > Should reproduce easily under TSan.
I cannot do it easily. Test I tried:
```bash
#!/usr/bin/env bash
CURDIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)
. "$CURDIR"/../shell_config.sh
set -e
$CLICKHOUSE_CLIENT -q "drop table if exists mem"
$CLICKHOUSE_CLIENT -q "create table mem (x UInt64) engine = M... | 2020-08-07T11:30:32 | 2020-08-07T11:30:32 | {} | MEMBER |
ClickHouse | ClickHouse | 670,520,722 | 13,416 | Akazz | I would suggest to try slightly different timings:
```
function f {
for i in $(seq 1 1000); do
$CLICKHOUSE_CLIENT -q "select count() >= 0 from (select * from mem settings max_block_size=1)" | grep -v '^1$' || true
done
}
function g {
for i in $(seq 1 1000); do
$CLICKHOUSE_CLIENT -q "insert into... | 2020-08-07T13:37:26 | 2020-08-07T14:30:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,617,774 | 13,494 | alesapin | Only push to dockerhub important. | 2020-08-07T17:08:22 | 2020-08-07T17:08:22 | {} | MEMBER |
ClickHouse | ClickHouse | 671,013,862 | 13,526 | alanpaulkwan | While we're at it, although I know there's plenty of reasons why Clickhouse uses a proprietary format, I would suggest that Clickhouse is a little bit slow right now when dealing with Parquet. DuckDB is much faster for example on the same file. I have a 96 million row with a 33k group by (just count()). The proprietary... | 2020-08-09T06:35:16 | 2020-08-09T06:36:05 | {} | NONE |
ClickHouse | ClickHouse | 670,261,823 | 13,335 | alexey-milovidov | Even a simple use case does not work:
```
SELECT netloc('<');
``` | 2020-08-07T00:53:07 | 2020-08-07T00:53:07 | {} | MEMBER |
ClickHouse | ClickHouse | 670,348,252 | 10,851 | zhang2014 | > Additionally it may be useful to introduce a setting of MaterializedMySQL database, that allow selecting from materialized tables when master is not available. We probably should enable this setting by default to make behaviour more consistent with ReplicatedMergeTree, but I'm not sure.
@tavplubix Done. I added th... | 2020-08-07T06:20:10 | 2020-08-07T06:21:11 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,417,592 | 10,851 | zhang2014 | > Segfault on ClickHouse shutdown when MySQL is not available: https://gist.github.com/tavplubix/c84d5b422b1aabff3a2f9d43ce64765c
Done. | 2020-08-07T09:10:44 | 2020-08-07T09:10:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 670,933,270 | 13,511 | alexey-milovidov | test_rabbitmq - bad test #12508 | 2020-08-08T14:08:58 | 2020-08-08T14:08:58 | {} | MEMBER |
ClickHouse | ClickHouse | 670,981,032 | 13,449 | alexey-milovidov | 00933_test_fix_extra_seek_on_compressed_cache - known bad test. | 2020-08-08T22:36:55 | 2020-08-08T22:36:55 | {} | MEMBER |
ClickHouse | ClickHouse | 671,079,247 | 12,508 | alexey-milovidov | https://clickhouse-builds.s3.yandex.net/0/be38b0fe753b03e0e88db3588f59cfb12e80ebcd/clickhouse_build_check/build_log_743800965_1596934349.txt
Build failed without build log. It's inner CI problem. Maybe something wrong with docker or host. Just wait for build restart. | 2020-08-09T17:28:43 | 2020-08-09T17:28:43 | {} | MEMBER |
ClickHouse | ClickHouse | 670,381,737 | 13,409 | alesapin | Fast test works! | 2020-08-07T07:47:54 | 2020-08-07T07:47:54 | {} | MEMBER |
ClickHouse | ClickHouse | 670,387,199 | 10,851 | tavplubix | Segfault on ClickHouse shutdown when MySQL is not available: https://gist.github.com/tavplubix/c84d5b422b1aabff3a2f9d43ce64765c | 2020-08-07T08:00:38 | 2020-08-07T08:00:38 | {} | MEMBER |
ClickHouse | ClickHouse | 670,471,899 | 13,453 | jrogov | Not really sure if it's dictated by the standard, since in Postgres I get this:
```psql
# select * from (select sum(id), sum(id) from users) s;
sum | sum
------+------
8923 | 8923
``` | 2020-08-07T11:29:57 | 2020-08-07T11:29:57 | {} | NONE |
ClickHouse | ClickHouse | 670,473,279 | 13,449 | alexey-milovidov | It's also Ok to backport this change because it fixes usability bug. | 2020-08-07T11:33:56 | 2020-08-07T11:33:56 | {} | MEMBER |
ClickHouse | ClickHouse | 670,681,226 | 12,508 | alexey-milovidov | > Use after free in sql_fuzzy test:
Fixed in #13234. | 2020-08-07T19:39:49 | 2020-08-07T19:39:49 | {} | MEMBER |
ClickHouse | ClickHouse | 670,849,998 | 12,513 | adrian17 | Still reproducible on 20.6.3. | 2020-08-08T09:24:34 | 2020-08-08T09:24:34 | {} | NONE |
ClickHouse | ClickHouse | 670,932,654 | 13,510 | alexey-milovidov | https://clickhouse-test-reports.s3.yandex.net/13510/6663f21fdd7c07b2ab549732d2f406133d839803/stress_test_(thread)/stderr.log
https://github.com/ClickHouse/ClickHouse/pull/13416 | 2020-08-08T14:02:50 | 2020-08-08T14:02:50 | {} | MEMBER |
ClickHouse | ClickHouse | 670,981,708 | 13,099 | alexey-milovidov | @akuzm It shows 22..24% performance improvement, but for unknown reason, performance test does not show it as confident result. | 2020-08-08T22:46:26 | 2020-08-08T22:46:26 | {} | MEMBER |
ClickHouse | ClickHouse | 666,633,985 | 12,991 | alexey-milovidov | ```
2020.07.28 18:03:53.757221 [ 54 ] {6e2c65f5-f212-4b29-88df-076e5ee755a6} <Trace> StorageKafka (kafka): Polled batch of 1 messages. Offset position: [ ]
2020.07.28 18:03:53.757275 [ 54 ] {6e2c65f5-f212-4b29-88df-076e5ee755a6} <Error> StorageKafka (kafka): Consumer error: Broker: Unknown topic or partition
2020.0... | 2020-07-30T19:30:53 | 2020-07-30T19:30:53 | {} | MEMBER |
ClickHouse | ClickHouse | 666,724,808 | 13,055 | alexey-milovidov | No changes according to the existing perf tests. | 2020-07-30T21:44:45 | 2020-07-30T21:44:45 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.