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 | 647,193,584 | 11,841 | HooliBox | there are about 2000 distinct s_info_windcode.
Thanks, the suggestion make a lot of sense. | 2020-06-21T23:10:47 | 2020-06-21T23:10:47 | {} | NONE |
ClickHouse | ClickHouse | 647,387,062 | 11,797 | qoega | >description for this pull request need overwrite,sorry for my mistake.
It can be done without PR close. Just use three dots near description. | 2020-06-22T09:05:05 | 2020-06-22T09:07:28 | {} | MEMBER |
ClickHouse | ClickHouse | 647,459,953 | 11,849 | filimonov | No i think it will not work :\ Dot has a special meaning in used XML parser. | 2020-06-22T11:33:18 | 2020-06-22T11:33:18 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,740,771 | 11,374 | alexey-milovidov | `SELECT toDateTime64('2011-11-11 11:11:11.1234567890123456789', 0)` | 2020-06-22T20:00:54 | 2020-06-22T20:00:54 | {} | MEMBER |
ClickHouse | ClickHouse | 647,817,662 | 11,622 | fpacifici | Hi @filimonov, thanks for the reply. I tried to assign aliases to the column and pass the aliases to the parameters of the transform function but I still get the same issue.
Is this what you meant ?
```
curl -X POST http://MY_DB:8123/ -d '
SELECT event_id,
array('\''1'\'') AS p_ids,
array... | 2020-06-22T23:23:14 | 2020-06-22T23:23:14 | {} | NONE |
ClickHouse | ClickHouse | 647,203,833 | 11,774 | azat | @alexey-milovidov what do you think, should it be enabled by default for freebsd (once it will work) ? Or the behavior that the PR has right now is more suitable (disable by default, but can be enabled with `-DENABLE_JEMALLOC=1`) until more testing will done ? | 2020-06-22T00:32:38 | 2020-06-22T00:32:38 | {} | MEMBER |
ClickHouse | ClickHouse | 647,461,174 | 11,850 | wallace-clickhouse | OK, verified! It's fine.Thanks!
Can this directory '/etc/clickhouse-client/config.xml' be configured to an other directory? | 2020-06-22T11:36:07 | 2020-06-22T11:36:07 | {} | NONE |
ClickHouse | ClickHouse | 647,775,149 | 8,624 | alexey-milovidov | @filimonov This number does not mean anything.
It can be too slow or too fast depending on hardware.
Nikita has tested with laptop SSD. | 2020-06-22T21:20:18 | 2020-06-22T21:20:18 | {} | MEMBER |
ClickHouse | ClickHouse | 647,797,130 | 11,875 | alexey-milovidov | @Enmk I highlighted all the bugs in comments. | 2020-06-22T22:17:50 | 2020-06-22T22:17:50 | {} | MEMBER |
ClickHouse | ClickHouse | 647,911,860 | 11,876 | xiemeilong | Resovle by change sql to below, this maybe not a bug.
```
CREATE MATERIALIZED VIEW IF NOT EXISTS test_hourly_mv
ENGINE = AggregatingMergeTree
PARTITION BY toYYYYMM(`__day`) ORDER BY (`__deviceId`, `__day`, `__hour`) AS
SELECT `__deviceId`,
toDate(`__eventTime`) AS `__day`,
toHour(`__eventTime`) AS `__hour... | 2020-06-23T05:12:00 | 2020-06-23T05:12:22 | {} | NONE |
ClickHouse | ClickHouse | 647,937,145 | 11,873 | KochetovNicolai | > BTW how about command to show processors pipeline graph in a graphviz format? I saw it in pipeline stuck exception, can it also be shown on demand?
Yes, I will add it too. | 2020-06-23T06:28:19 | 2020-06-23T06:28:19 | {
"rocket": 2
} | MEMBER |
ClickHouse | ClickHouse | 647,947,455 | 4,609 | filimonov | Implementation for that case in more cumbersome, and questionable
1) codec is smth we apply on the top of numerical column. Internally Nullable(UIntXX) is 2 columns, it is not very clear how it should be applied there.
2) what is the `Delta(1,Null)` ? | 2020-06-23T06:54:28 | 2020-06-23T06:54:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,164,040 | 11,715 | alexey-milovidov | > Yandex synchronization check (only for Yandex employees)
This is Ok. | 2020-06-21T18:27:44 | 2020-06-21T18:27:44 | {} | MEMBER |
ClickHouse | ClickHouse | 647,403,641 | 11,850 | filimonov | Just put a line `<port>4444</port>` inside /etc/clickhouse-client/config.xml
```
clickhouse-client
ClickHouse client version 20.5.1.3716 (official build).
Connecting to localhost:4444 as user default.
```` | 2020-06-22T09:36:38 | 2020-06-22T09:36:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,410,151 | 8,624 | filimonov | BTW - it would be really interesting to see some perf number, and/or comparison with aerospike (#5629) | 2020-06-22T09:48:55 | 2020-06-22T09:48:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,567,046 | 11,866 | alexey-milovidov | Ok.
> initializeAggregation('minState', b)
Should it have "State" or better to assume that the user wants to create state?
Also I want a function to create empty aggregation state.
Like `emptyAggregation('min', b)` where `b` is used only for type inference. | 2020-06-22T14:49:51 | 2020-06-22T14:50:10 | {} | MEMBER |
ClickHouse | ClickHouse | 647,696,381 | 11,401 | akuzm | In the experiment 2.3, many arithmetic queries are 10-20% slower, but why?
Locally I managed to calculate the same statistics for all metrics from `PerformanceEvents` from query log. Here is a list of those metrics that significantly changed for many arithmetic queries,
and the amounts of change:
```
akuzm@akuzm-ni... | 2020-06-22T18:23:53 | 2020-06-22T18:23:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,733,215 | 11,774 | azat | >ClickHouse build check — Some exception occurred :(
Will it be restarted? (I saw the same issue in another PR) | 2020-06-22T19:43:02 | 2020-06-22T19:43:02 | {} | MEMBER |
ClickHouse | ClickHouse | 647,911,098 | 11,873 | filimonov | BTW how about command to show processors pipeline graph in a graphviz format? I saw it in pipeline stuck exception, can it also be shown on demand? | 2020-06-23T05:09:24 | 2020-06-23T05:09:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,182,655 | 11,847 | alexey-milovidov | > Docs check — Found 1 errors
Broken on master @blinkov
> PVS check — Found 1 new errors, total 120 errors
Ok. | 2020-06-21T21:15:54 | 2020-06-21T21:15:54 | {} | MEMBER |
ClickHouse | ClickHouse | 647,185,333 | 11,806 | alexey-milovidov | I have deleted 20.5 branch. | 2020-06-21T21:41:03 | 2020-06-21T21:41:03 | {} | MEMBER |
ClickHouse | ClickHouse | 647,527,250 | 11,766 | den-crane | JDBC driver issue https://github.com/ClickHouse/clickhouse-jdbc/issues/290 | 2020-06-22T13:41:53 | 2020-06-22T13:41:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,681,977 | 11,858 | alexey-milovidov | I know why it has failed...
There is a race condition between program startup and clearing the password in command line. | 2020-06-22T17:53:43 | 2020-06-22T17:53:43 | {} | MEMBER |
ClickHouse | ClickHouse | 647,445,578 | 10,058 | alexey-milovidov | Merged in #11847. | 2020-06-22T11:01:35 | 2020-06-22T11:01:35 | {} | MEMBER |
ClickHouse | ClickHouse | 646,768,518 | 11,745 | alesapin | TODO:
- [x] 1. Add a test with `ALTER` during long select. (We are absolutely non blocking now)
- [x] 2. Add comments.
- [x] 3. Realize what's wrong with `01213_alter_rename_column_zookeeper` | 2020-06-19T17:20:12 | 2020-06-22T11:55:40 | {} | MEMBER |
ClickHouse | ClickHouse | 647,677,480 | 11,840 | azat | >We have to ask @KochetovNicolai for help or maybe you can investigate?
Interesting, fixed - 267a6c8639f3614bff3ff42385c2a1f7d7c35d41
P.S. Hm, I didn't look that it does not fails on upstream, egh. | 2020-06-22T17:44:00 | 2020-06-22T17:44:00 | {} | MEMBER |
ClickHouse | ClickHouse | 647,681,280 | 11,858 | alexey-milovidov | It is strange that https://clickhouse-test-reports.s3.yandex.net/11858/18ec373369d13a1d22f8c3faa680e49d9e2d1f8c/functional_stateless_tests_(thread)/test_run.txt.out.log failed, but looks not related. | 2020-06-22T17:52:18 | 2020-06-22T17:52:18 | {} | MEMBER |
ClickHouse | ClickHouse | 647,749,440 | 11,374 | alexey-milovidov | `SELECT toDateTime64('2011-11-11 11:11:11.-12345678901234567890', 0)` | 2020-06-22T20:21:03 | 2020-06-22T20:21:03 | {} | MEMBER |
ClickHouse | ClickHouse | 647,547,896 | 8,053 | akuzm | Probably the reason is type promotion. We have 'sumWithOverflow' that does not promote the type, but it's not in the list of supported SimpleAggregateFunctions. If we just add it to the list, it should work. | 2020-06-22T14:16:46 | 2020-06-22T14:16:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,553,498 | 11,850 | den-crane | > OK, verified! It's fine.Thanks!
> Can this directory '/etc/clickhouse-client/config.xml' be configured to an other directory?
yes it can, https://clickhouse.tech/docs/en/interfaces/cli/#configuration_files | 2020-06-22T14:26:40 | 2020-06-22T14:26:40 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,686,922 | 11,774 | alexey-milovidov | @azat It's better to make it by default, because ClickHouse will work bettter. If I remember correctly, FreeBSD is also using jemalloc by default but the version will be older and without patches for performance. | 2020-06-22T18:03:50 | 2020-06-22T18:03:50 | {} | MEMBER |
ClickHouse | ClickHouse | 647,778,852 | 11,783 | den-crane | >failed to respond
"Failed to respond" -- it's an JDBC driver issue. You should just repeat your insert in this case.
Or try another JDBC drivers https://github.com/blynkkk/clickhouse4j https://github.com/housepower/ClickHouse-Native-JDBC
>I just want to know where is performance bottlenect while
>CH process... | 2020-06-22T21:28:59 | 2020-06-22T21:28:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,786,215 | 11,861 | azat | >under what situation, the second construcation function takes effect?
Used for `remote`:
https://github.com/ClickHouse/ClickHouse/blob/e76941b52cbb722a5b6f4e7252df58ba14d8d910/src/TableFunctions/TableFunctionRemote.cpp#L179-L187 | 2020-06-22T21:48:08 | 2020-06-22T21:48:08 | {} | MEMBER |
ClickHouse | ClickHouse | 647,946,090 | 11,668 | bharatnc | Thank you @akuzm for your help with fixing the failing test. | 2020-06-23T06:51:11 | 2020-06-23T06:51:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,946,891 | 11,815 | alexey-milovidov | 01079_parallel_alter_detach_table_zookeeper timeout https://clickhouse-test-reports.s3.yandex.net/0/d147cd646a1e6cc41d42f92d57f016a5c49d04de/functional_stateless_tests_(thread).html#fail1 | 2020-06-23T06:53:13 | 2020-06-23T06:53:13 | {} | MEMBER |
ClickHouse | ClickHouse | 647,978,132 | 11,880 | boy-xiaozhang | > 错误消息不解自明：没有ZooKeeper，您将无法创建复制表
>
> 如果您在连接Zookeeper时遇到问题，请查看Clickhouse日志（`/var/log/clickhouse-server/clickhouse-server.log`）以查找原因。
Two machines, one can be created, one can not be created, Zookeeper,log does not print exceptions | 2020-06-23T07:59:33 | 2020-06-23T07:59:33 | {} | NONE |
ClickHouse | ClickHouse | 647,149,832 | 11,840 | alexey-milovidov | Something wrong with `01091_num_threads` test. | 2020-06-21T16:24:48 | 2020-06-21T16:24:48 | {} | MEMBER |
ClickHouse | ClickHouse | 647,150,022 | 11,840 | alexey-milovidov | #9483 | 2020-06-21T16:26:04 | 2020-06-21T16:26:04 | {} | MEMBER |
ClickHouse | ClickHouse | 647,200,360 | 11,669 | alexey-milovidov | What value will be good: 1 or 10 or 100? | 2020-06-22T00:09:23 | 2020-06-22T00:09:23 | {} | MEMBER |
ClickHouse | ClickHouse | 647,498,175 | 11,669 | akuzm | > P.S. I'm trying to keep PRs without cleanups/refactoring to reduce number of changes (lots of changes is hard to review), but looks like it is not a problem for you, so I will consider deviating from this rule in some cases
That's generally a good idea if there is a non-trivial amount of refactoring. One way to do... | 2020-06-22T12:49:43 | 2020-06-22T12:49:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,525,806 | 11,863 | alesapin | Only functional tests are important | 2020-06-22T13:39:26 | 2020-06-22T13:39:26 | {} | MEMBER |
ClickHouse | ClickHouse | 647,525,966 | 11,860 | alesapin | Only integration tests are important. | 2020-06-22T13:39:42 | 2020-06-22T13:39:49 | {} | MEMBER |
ClickHouse | ClickHouse | 647,742,248 | 11,868 | filimonov | Duplicate of https://github.com/ClickHouse/ClickHouse/issues/4432 | 2020-06-22T20:04:26 | 2020-06-22T20:05:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,968,200 | 11,878 | qoega | Can you add relevant test cases? | 2020-06-23T07:39:36 | 2020-06-23T07:39:36 | {} | MEMBER |
ClickHouse | ClickHouse | 647,163,019 | 2,494 | alexey-milovidov | Fixed by #11661 | 2020-06-21T18:19:40 | 2020-06-21T18:19:40 | {} | MEMBER |
ClickHouse | ClickHouse | 647,163,858 | 11,564 | alexey-milovidov | Yes, you're right. | 2020-06-21T18:26:13 | 2020-06-21T18:26:13 | {} | MEMBER |
ClickHouse | ClickHouse | 647,198,111 | 11,669 | azat | >I'm not sure how admin is supposed to choose the value of this setting.
via `config.xml`? | 2020-06-21T23:52:12 | 2020-06-21T23:52:12 | {} | MEMBER |
ClickHouse | ClickHouse | 647,363,454 | 11,852 | dependabot-preview[bot] | OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting `@dependabot ignore this major version` or `@dependabot ignore this minor version`.
If you change your mind, just re-... | 2020-06-22T08:20:04 | 2020-06-22T08:20:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,401,010 | 11,849 | filimonov | You can make different configurations per topic in config.xml, like that:
```xml
<yandex>
<kafka>
<!-- global configuration -->
</kafka>
<kafka_events>
<!-- configuration for topic 'events' -->
</kafka_events>
<kafka_users>
<!-- configuration for topic 'users' -->
... | 2020-06-22T09:31:36 | 2020-06-22T09:31:44 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,650,446 | 11,669 | azat | >That's generally a good idea if there is a non-trivial amount of refactoring.
And sometimes even trivial may be a problem (especially for backporting)
> One way to do it is maintain logically separate changes in separate commits -- it works well with the github interface.
Well, keeping separate changes in sep... | 2020-06-22T17:05:43 | 2020-06-22T17:05:43 | {} | MEMBER |
ClickHouse | ClickHouse | 647,161,115 | 11,815 | alexey-milovidov | #11835
#11834
| 2020-06-21T18:02:19 | 2020-06-21T18:02:19 | {} | MEMBER |
ClickHouse | ClickHouse | 647,203,340 | 11,774 | azat | >You can run built binary in Vagrant/Virtualbox/Qemu and test that it does not crash on trivial queries.
Yep, that was a good idea, right now - it does not works (at least for freebsd) | 2020-06-22T00:30:07 | 2020-06-22T00:30:07 | {} | MEMBER |
ClickHouse | ClickHouse | 647,229,903 | 11,783 | etah000 | Yes, I used jdbc to insert data one row each insert in 20 threads.
I just want to know where is performance bottlenect while CH processes a large number of small inserts in parallel.
We know that mysql can handle 10k single row insert each second, so you mean that it is the connecting protocol that limits the pe... | 2020-06-22T02:20:15 | 2020-06-22T02:20:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,363,435 | 11,851 | dependabot-preview[bot] | OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting `@dependabot ignore this major version` or `@dependabot ignore this minor version`.
If you change your mind, just re-... | 2020-06-22T08:20:03 | 2020-06-22T08:20:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,445,351 | 11,847 | alexey-milovidov | > Performance — 14 faster
As expected. | 2020-06-22T11:01:04 | 2020-06-22T11:01:04 | {} | MEMBER |
ClickHouse | ClickHouse | 647,492,768 | 8,624 | alexey-milovidov | @filimonov They are published in https://presentations.clickhouse.tech/hse_2020/3rd/SSD_Dictionary_full.pdf | 2020-06-22T12:40:17 | 2020-06-22T12:40:17 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 647,550,633 | 11,841 | den-crane | I could not reproduce it with debian10 - 4.19.0-6 / CH 20.4.5.36 / nproc --all - 16 | 2020-06-22T14:21:35 | 2020-06-22T14:21:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,650,656 | 11,840 | azat | > >But the way, this test is also flaky but it fails only occasionally.
>
>I don't see the failure, but it may require some tuning
@alexey-milovidov Ok, found, and fixed and now only 01091_num_threads fails. | 2020-06-22T17:06:10 | 2020-06-22T17:06:10 | {} | MEMBER |
ClickHouse | ClickHouse | 647,715,745 | 11,826 | alexey-milovidov | After set `send_logs_level` to something other than `none` there are too many progress messages in client:
```
milovidov-desktop :) SELECT toDateTime64('2019-09-16 19:20:12', 0)
SELECT toDateTime64('2019-09-16 19:20:12', 0)
↖ Progress: 0.00 rows, 0.00 B (0.00 rows/s., 0.00 B/s.)
┌─toDateTime64('2019-09-16 1... | 2020-06-22T19:02:11 | 2020-06-22T19:02:11 | {} | MEMBER |
ClickHouse | ClickHouse | 647,866,948 | 4,609 | mjodeh | any progress? | 2020-06-23T02:23:15 | 2020-06-23T02:23:15 | {} | NONE |
ClickHouse | ClickHouse | 647,164,090 | 11,715 | alexey-milovidov | ```
/build/build_docker/../src/Interpreters/ClusterProxy/executeQuery.cpp:55:56: error: converting integer literal to bool, use bool literal instead [modernize-use-bool-literals,-warnings-as-errors]
new_settings.optimize_skip_unused_shards = 0;
^
... | 2020-06-21T18:28:09 | 2020-06-21T18:28:09 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 647,201,935 | 11,669 | azat | It depends, and should be adjusted only after looking at system.clusters (that has error_count rate) + server logs (for issues around distributed queries).
I would say that 100 definitely is not good, but 1-10 may be a good idea.
Also rate of queries and time of replica error_count recover (distributed_replica_er... | 2020-06-22T00:21:39 | 2020-06-22T00:21:39 | {} | MEMBER |
ClickHouse | ClickHouse | 647,433,763 | 972 | lyfzwrthlvng | @imvs did you open a separate issue? Did you figure out why you were seeing duplicate data? Some days baack we also saw similar issue, unfortunately we restarted and issue was gone, so can't reproduct it since then. Wanted to get better understanding of the issue. | 2020-06-22T10:35:38 | 2020-06-22T10:35:38 | {} | NONE |
ClickHouse | ClickHouse | 647,446,432 | 11,498 | filimonov | Regarding our talk about the permissions of proxy users.
It still not clear for me if we should allow proxying per user or globally for all users. The use case when we need to limit the list of users in which name that proxy can talk is still not very clear for me.
Let's imagine:
We have user Alice logged in ... | 2020-06-22T11:03:26 | 2020-06-22T11:10:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,491,255 | 11,859 | alexey-milovidov | Let's support comments in form of
`# comment`
in this file and add description?
| 2020-06-22T12:37:28 | 2020-06-22T12:37:28 | {} | MEMBER |
ClickHouse | ClickHouse | 647,651,738 | 11,840 | alexey-milovidov | We have to ask @KochetovNicolai for help or maybe you can investigate? | 2020-06-22T17:08:28 | 2020-06-22T17:08:28 | {} | MEMBER |
ClickHouse | ClickHouse | 647,150,067 | 11,840 | alexey-milovidov | But the way, this test is also flaky but it fails only occasionally. | 2020-06-21T16:26:26 | 2020-06-21T16:26:26 | {} | MEMBER |
ClickHouse | ClickHouse | 647,407,129 | 11,804 | filimonov | Can you also add test cases from #3178 and from https://github.com/ClickHouse/ClickHouse/pull/5435#issuecomment-496486509
Probably those are the same. | 2020-06-22T09:43:03 | 2020-06-22T09:44:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 646,853,558 | 11,804 | akuzm | The naive approach doesn't work because of queries like this:
```
00167_settings_inside_query.sql:
SELECT min(number) FROM system.numbers WHERE toUInt64(number % 1000) IN (
SELECT DISTINCT blockSize() FROM system.numbers
SETTINGS max_block_size = 123, max_rows_to_read = 1000,
read_overflow_mo... | 2020-06-19T20:43:18 | 2020-06-22T12:22:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,975,915 | 11,745 | alesapin | Race condition in stress test. https://gist.github.com/alesapin/82c0f645d714eb74de255040c1d08fc2 | 2020-06-23T07:54:58 | 2020-06-23T07:54:58 | {} | MEMBER |
ClickHouse | ClickHouse | 647,498,811 | 11,859 | exprmntr | > Let's support comments in form of
> `# comment`
> in this file and add description?
It's already supported, any line without testcase name will be ignored. Description added. | 2020-06-22T12:50:51 | 2020-06-22T12:51:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,522,261 | 11,300 | blinkov | > Maybe there is some regression in stack trace formatting speed?
@alexey-milovidov doesn't look like it, I've restored the previous stacktrace-to-string code and it behaves the same in these tests. | 2020-06-22T13:33:06 | 2020-06-22T13:33:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,747,244 | 11,774 | alexey-milovidov | > Will it be restarted?
No. You can make one more commit to restart CI checks. | 2020-06-22T20:16:24 | 2020-06-22T20:16:35 | {} | MEMBER |
ClickHouse | ClickHouse | 647,756,550 | 11,872 | alexey-milovidov | > Functional stateful tests (thread)
GitHub is down:
> CalledProcessError: Command 'docker -H unix:///place/sandbox-data/tasks/1/8/706746281/docker_root/docker.sock run -i --rm ubuntu:14.04 ping6 github.com -c4' returned non-zero exit status 2.
https://downforeveryoneorjustme.com/github | 2020-06-22T20:36:15 | 2020-06-22T20:36:46 | {} | MEMBER |
ClickHouse | ClickHouse | 647,773,755 | 8,624 | filimonov | Just for the reference - aerospike gives up to ~1Mln req/sec | 2020-06-22T21:17:03 | 2020-06-22T21:17:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,824,639 | 11,872 | alexey-milovidov | Functional tests Ok. | 2020-06-22T23:47:40 | 2020-06-22T23:47:40 | {} | MEMBER |
ClickHouse | ClickHouse | 647,937,606 | 11,840 | azat | Now tests passed | 2020-06-23T06:29:32 | 2020-06-23T06:29:32 | {} | MEMBER |
ClickHouse | ClickHouse | 647,441,523 | 11,786 | alesapin | > Funny thing, but we can't unmaterialize column back:
>
> ```
> :) desc ttt
>
> DESCRIBE TABLE ttt
>
> ┌─name─┬─type───┬─default_type─┬─default_expression─┬─comment─┬─codec_expression─┬─ttl_expression─┐
> │ x │ UInt32 │ MATERIALIZED │ 3 │ │ │ │
> ... | 2020-06-22T10:52:31 | 2020-06-22T10:52:31 | {} | MEMBER |
ClickHouse | ClickHouse | 647,444,619 | 11,824 | alexey-milovidov | New test has failed in "Arcadia" because it does not support tests with cluster. | 2020-06-22T10:59:26 | 2020-06-22T10:59:26 | {} | MEMBER |
ClickHouse | ClickHouse | 647,506,339 | 11,826 | alexey-milovidov | This change will allow to stop run of functional test after first fatal error. | 2020-06-22T13:04:50 | 2020-06-22T13:04:50 | {} | MEMBER |
ClickHouse | ClickHouse | 647,682,257 | 11,815 | alexey-milovidov | Clear password in command line, race condition: https://clickhouse-test-reports.s3.yandex.net/11858/18ec373369d13a1d22f8c3faa680e49d9e2d1f8c/functional_stateless_tests_(thread)/test_run.txt.out.log | 2020-06-22T17:54:20 | 2020-06-22T17:54:20 | {} | MEMBER |
ClickHouse | ClickHouse | 647,692,372 | 11,774 | azat | >@azat It's better to make it by default, because ClickHouse will work bettter.
Ok, actually I already did this (with experimental message just in case) in the latest version (replace ifdef in headers with separate header for each os/arch) | 2020-06-22T18:15:35 | 2020-06-22T18:15:35 | {} | MEMBER |
ClickHouse | ClickHouse | 647,697,094 | 11,815 | alexey-milovidov | #11872 | 2020-06-22T18:25:25 | 2020-06-22T18:25:25 | {} | MEMBER |
ClickHouse | ClickHouse | 647,714,279 | 11,873 | KochetovNicolai | ```sql
:) explain select sum(number) from numbers(10) union all select sum(number) from numbers(10) order by sum(number)
EXPLAIN SELECT sum(number)
FROM numbers(10)
UNION ALL
SELECT sum(number)
FROM numbers(10)
ORDER BY sum(number) ASC
FORMAT TSV
Union
Expression (Projection)
Expression (Before ORD... | 2020-06-22T18:59:52 | 2020-06-22T18:59:52 | {} | MEMBER |
ClickHouse | ClickHouse | 647,981,069 | 11,745 | alesapin | > Race condition in stress test. https://gist.github.com/alesapin/82c0f645d714eb74de255040c1d08fc2
Fix https://github.com/ClickHouse/ClickHouse/pull/11745/commits/292016e04515f5cddf5a540baddab77952e2db83 | 2020-06-23T08:05:15 | 2020-06-23T08:05:15 | {} | MEMBER |
ClickHouse | ClickHouse | 647,185,251 | 11,669 | alexey-milovidov | @nvartolomei The calculated number of errors is already a sliding "window" with exponential smoothing. | 2020-06-21T21:40:04 | 2020-06-21T21:40:04 | {} | MEMBER |
ClickHouse | ClickHouse | 647,187,738 | 11,567 | oandrew | @filimonov Looks like you were trying to use Avro instead of AvroConfluent | 2020-06-21T22:05:14 | 2020-06-21T22:05:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,970,200 | 11,875 | qoega | Relevant tests:
```
[ RUN ] data_type/LeastSuperTypeTest.getLeastSupertype/20
unknown file: Failure
C++ exception with description "Scale 12 is too large for DateTime64. Maximum is up to nanoseconds (9)." thrown in SetUp().
[ FAILED ] data_type/LeastSuperTypeTest.getLeastSupertype/20, where GetParam() = Ty... | 2020-06-23T07:43:51 | 2020-06-23T07:43:51 | {} | MEMBER |
ClickHouse | ClickHouse | 647,972,474 | 11,219 | filimonov | @4ertus2 it is a regression (happened somewhere between 19.11 and 19.13), and since 20.3 is lts we should try to backport / adopt the fix. :\ | 2020-06-23T07:48:19 | 2020-06-23T07:48:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,148,193 | 11,669 | nvartolomei | I think a better way to balance things would be to track the rate of errors (instead of an absolute value) with a sliding window and allow a small variability in the rate of errors between replicas.
The ignore value added here seems to help only in the beginning before the err counter reaches that value and then the... | 2020-06-21T16:13:40 | 2020-06-21T16:13:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,159,600 | 11,841 | den-crane | Can you show an output of `select uniq(s_info_windcode) from quant.minute_bar` I mean some approximate number of possible s_info_windcode ? | 2020-06-21T17:47:50 | 2020-06-21T17:47:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,748,807 | 11,873 | alexey-milovidov | > MergingSorted (Merge sorted streams before ORDER BY)
MergeSorting (Merge sorted blocks before ORDER BY)
PartialSorting (Sort each block before ORDER BY)
Actually not before ORDER BY but for ORDER BY. | 2020-06-22T20:19:50 | 2020-06-22T20:19:50 | {} | MEMBER |
ClickHouse | ClickHouse | 647,885,536 | 11,565 | qixiaogang | @azat
https://github.com/ClickHouse/ClickHouse/issues/11565#issuecomment-643674741 simple round-robin will not work for my case, as my case across AZs. For query hit A AZ, we would like it go to replica AZ_A_shard1_replicas1 and A_shard1_replicas2 first if all 4 replica has same errors.
https://github.com/ClickHo... | 2020-06-23T03:27:26 | 2020-06-23T03:28:36 | {} | NONE |
ClickHouse | ClickHouse | 647,926,984 | 11,774 | azat | Now for non-linux setup it has a warning, but it uses number of available CPUs correctly on darwin/freebsd:
> <jemalloc>: No getcpu support: percpu_arena:percpu
> <jemalloc>: perCPU arena getcpu() not available. Setting narenas to 1.
| 2020-06-23T05:59:12 | 2020-06-23T05:59:12 | {} | MEMBER |
ClickHouse | ClickHouse | 647,151,085 | 11,840 | azat | >But the way, this test is also flaky but it fails only occasionally.
I don't see the failure, but it may require some tuning | 2020-06-21T16:33:19 | 2020-06-21T16:33:19 | {} | MEMBER |
ClickHouse | ClickHouse | 647,443,845 | 11,849 | hoseiney | it is good to have customized config for each topic
but we are using Debezium and topics name contain dot
for example : mysql.users
can we have this config ?
```
<kafka_mysql.users>
</kafka_mysql.users>
```
| 2020-06-22T10:57:41 | 2020-06-22T10:57:41 | {} | NONE |
ClickHouse | ClickHouse | 647,744,143 | 11,866 | filimonov | One of the usecases:
```sql
CREATE TABLE ...
(
last_value aggregateFunction(argMax, String, DateTime) DEFAULT arrayReduce('argMaxState', ['lack of value'], [0])
...
)
``` | 2020-06-22T20:09:08 | 2020-06-22T20:09:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 647,975,277 | 11,880 | filimonov | The error message is self-explaining: you can't create replicated table without ZooKeeper
If you have some issues connecting to zookeeper, check the clickhouse logs (`/var/log/clickhouse-server/clickhouse-server.log`) to find the reason. | 2020-06-23T07:53:51 | 2020-06-23T07:53:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 628,662,328 | 10,909 | den-crane | For replication you have to set `interserver_https_port`
https://github.com/ClickHouse/ClickHouse/blob/97f2a2213e754ba25dabba4bc8ddf507cd67660c/tests/integration/test_https_replication/configs/ssl_conf.xml
And CH still needs `tcp_port` do not remove it. | 2020-05-14T14:10:26 | 2020-05-14T14:10:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 628,717,307 | 10,914 | 4ertus2 | ```
CREATE TABLE hits
(
`EventDate` Date,
`CounterID` UInt32,
`UserID` UInt64
)
ENGINE = MergeTree()
ORDER BY (CounterID);
insert into hits values ('2001-01-01', 0, 0);
select l.EventDate from hits as l join hits as r on l.CounterID = r.CounterID where l.EventDate = today() and r.EventDate = t... | 2020-05-14T15:39:37 | 2020-05-14T16:42:47 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.