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 | 863,899,454 | 21,619 | vdimir | Solved, ref https://github.com/ClickHouse/ClickHouse/pull/23312 | 2021-06-18T09:28:46 | 2021-06-18T09:28:46 | {} | MEMBER |
ClickHouse | ClickHouse | 863,964,753 | 25,450 | alexey-milovidov | The version 21.3.3 does not contain all the patches. Should be updated to 21.3.12 or 21.6.3. | 2021-06-18T11:18:53 | 2021-06-18T11:18:53 | {} | MEMBER |
ClickHouse | ClickHouse | 863,968,998 | 7,288 | alexey-milovidov | Did not implemented. | 2021-06-18T11:26:23 | 2021-06-18T11:26:35 | {} | MEMBER |
ClickHouse | ClickHouse | 864,504,937 | 25,512 | huangmr0811 | Could you help me? @alexey-milovidov
Thanks so much! | 2021-06-20T06:06:14 | 2021-06-20T06:06:14 | {} | NONE |
ClickHouse | ClickHouse | 864,519,965 | 25,427 | azat | >Marker check Pending — Waiting to start Functional stateless tests (ANTLR debug)
Seems that they won't be run? | 2021-06-20T08:39:51 | 2021-06-20T08:39:51 | {} | MEMBER |
ClickHouse | ClickHouse | 863,632,945 | 22,159 | alexey-milovidov | 3565 threads should be tolerable. | 2021-06-17T23:55:28 | 2021-06-17T23:55:28 | {} | MEMBER |
ClickHouse | ClickHouse | 863,634,783 | 22,159 | alexey-milovidov | >│ ParalInputsProc │ 881 │
It should use less amount of threads since https://github.com/ClickHouse/ClickHouse/pull/23683, version 21.5. | 2021-06-18T00:01:16 | 2021-06-18T00:01:16 | {} | MEMBER |
ClickHouse | ClickHouse | 863,639,674 | 11,193 | alexey-milovidov | > Most common reason is bad schema (too granular partitioning, too aggressive compression, etc) creating too many merges, or making them slow.
+1 Absolutely true. | 2021-06-18T00:13:29 | 2021-06-18T00:13:29 | {} | MEMBER |
ClickHouse | ClickHouse | 863,640,073 | 11,193 | alexey-milovidov | > Whats the significant difference between max_replicated_merges_in_queue and background_pool_size?
The first is the limit on scheduled merges and the second is on currently running merges. | 2021-06-18T00:14:34 | 2021-06-18T00:14:34 | {} | MEMBER |
ClickHouse | ClickHouse | 863,763,561 | 24,867 | bharatnc | @nikitamikhaylov I think all comments have been addressed, let me know if it looks fine to be merged.
Test failures seem unrelated. | 2021-06-18T05:19:30 | 2021-06-18T05:21:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 863,862,659 | 25,421 | alesapin | A lot of tests failed. I don't understand why. Need to investigate. | 2021-06-18T08:32:12 | 2021-06-18T08:32:12 | {} | MEMBER |
ClickHouse | ClickHouse | 863,876,122 | 25,421 | alesapin | Keeper failures:
```
==================================== ERRORS ====================================
___________________ ERROR at setup of test_restart_multinode ___________________
[gw0] linux -- Python 3.8.5 /usr/bin/python3
@pytest.fixture(scope="module")
def started_cluster():
try:
> ... | 2021-06-18T08:53:04 | 2021-06-18T08:53:04 | {} | MEMBER |
ClickHouse | ClickHouse | 863,943,446 | 23,911 | vdimir | Resolved with #24885 | 2021-06-18T10:39:56 | 2021-06-18T10:39:56 | {} | MEMBER |
ClickHouse | ClickHouse | 863,997,226 | 24,341 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-06-18T12:18:58 | 2021-06-18T12:18:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 864,082,599 | 25,484 | vdimir | Or maybe better to add check here and do not allow types other that UInt8?
Or add conversion?
https://github.com/ClickHouse/ClickHouse/blob/e4d80f53fa25e9738a11d68a3be592ec5bd15c2d/src/Interpreters/ExpressionAnalyzer.cpp#L972-L974 | 2021-06-18T14:32:23 | 2021-06-18T14:32:54 | {} | MEMBER |
ClickHouse | ClickHouse | 864,357,493 | 25,427 | azat | >[01551_context_uaf](https://clickhouse-test-reports.s3.yandex.net/25427/07a463589592e123bc017f67dd949f14a577b556/functional_stateless_tests_(release).html#fail1)
```
2021-06-17 23:26:20 [90c14ab606f2] 2021.06.17 23:26:20.610776 [ 7852 ] {833fc10a-7fa6-4b73-aa84-17b427ffa439} <Error> executeQuery: Code: 76, e.displ... | 2021-06-19T05:13:59 | 2021-06-19T05:13:59 | {} | MEMBER |
ClickHouse | ClickHouse | 864,461,949 | 25,508 | tavplubix | Looks like a particular case of #23424 | 2021-06-19T20:49:10 | 2021-06-19T20:49:10 | {} | MEMBER |
ClickHouse | ClickHouse | 863,639,105 | 11,193 | alexey-milovidov | Yes, background pool is shared for all tables. | 2021-06-18T00:11:50 | 2021-06-18T00:11:50 | {} | MEMBER |
ClickHouse | ClickHouse | 863,863,637 | 23,052 | vdimir | Resolved with https://github.com/ClickHouse/ClickHouse/pull/24061 | 2021-06-18T08:33:43 | 2021-06-18T08:33:43 | {} | MEMBER |
ClickHouse | ClickHouse | 863,979,447 | 25,404 | alexey-milovidov | Tests can be added in `tests/queries/0_stateless` | 2021-06-18T11:47:10 | 2021-06-18T11:47:10 | {} | MEMBER |
ClickHouse | ClickHouse | 864,381,961 | 4,489 | Mall0c | For anyone who comes across this: using clickhouse-client, issuing a query with a large result set and having output_format_pretty_max_rows set to a large number, like the default 10k, can obviously slow down the performance of the query as such and tell you a wrong number of rows processed per second, total time and s... | 2021-06-19T09:31:05 | 2021-06-19T09:31:48 | {} | NONE |
ClickHouse | ClickHouse | 863,956,005 | 17,281 | alexey-milovidov | Done #16618 | 2021-06-18T11:04:06 | 2021-06-18T11:04:06 | {} | MEMBER |
ClickHouse | ClickHouse | 863,956,850 | 21,338 | alex-zaitsev | ```
select event, formatReadableSize(value)
from system.events where event in ('InsertedBytes','ZooKeeperBytesSent','ZooKeeperBytesReceived');
┌─event──────────────────┬─formatReadableSize(value)─┐
│ InsertedBytes │ 89.17 GiB │
│ ZooKeeperBytesSent │ 111.04 GiB │
│ Zo... | 2021-06-18T11:05:28 | 2021-06-18T11:06:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 863,957,729 | 25,404 | TszKitLo40 | OK, I will try to solve this issue. After finishing the parser, can I filed the PR first and let you have a look? I am not sure where should the tests be added.
| 2021-06-18T11:06:51 | 2021-06-18T11:07:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 864,048,576 | 25,457 | morozovsk | >just because the new data cannot be added to the view during import?
yes
>Besides, are there any other questions?
no
>Can I use this parameter for offline scheduled data import?
yes
>For loading stock data to new materialized view, is there any suggested solution;
What do you mean? | 2021-06-18T13:42:10 | 2021-06-18T13:42:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 864,219,092 | 25,459 | alesapin | Ok, now we have all run logs. | 2021-06-18T18:53:47 | 2021-06-18T18:53:47 | {} | MEMBER |
ClickHouse | ClickHouse | 864,257,981 | 25,421 | alesapin | cannot merge anything. Integration tests are completely broken. | 2021-06-18T20:27:05 | 2021-06-18T20:27:12 | {} | MEMBER |
ClickHouse | ClickHouse | 864,308,942 | 25,416 | shaon-chowdhury | @UnamedRus thank you! Added these at the kafka table level | 2021-06-18T22:54:36 | 2021-06-18T22:54:36 | {} | NONE |
ClickHouse | ClickHouse | 864,438,899 | 25,444 | tavplubix | Functional stateless tests (release, DatabaseReplicated) - 01914_exchange_dictionaries - need to update `ASTRenameQuery::formatQueryImpl(...)` | 2021-06-19T17:21:49 | 2021-06-19T17:21:49 | {} | MEMBER |
ClickHouse | ClickHouse | 864,496,421 | 25,404 | TszKitLo40 | I think it should be parsed like this `if (s_distinct_on.ignore(pos, expected)) {
if (!exp_list.parse(pos, limit_by_expression_list, expected))
return false;
}` but how can we parse `limit_by_lenth`, it seems that there is no keyword about `limit by ` in `SELECT a, b, c... LIMIT 1... | 2021-06-20T04:06:04 | 2021-06-20T04:06:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 863,643,738 | 20,048 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/issues/25449 | 2021-06-18T00:25:16 | 2021-06-18T00:25:16 | {
"heart": 1
} | MEMBER |
ClickHouse | ClickHouse | 863,659,455 | 21,582 | shy1721615055 | > 有关主机的信息位于带有`<Trace>`级别的先前日志消息中。
> 您的集群配置错误：副本无法相互连接。
> You have misconfigured your cluster: replicas cannot connect to each other.
Excuse me, is there any problem with the cluster configuration | 2021-06-18T01:11:06 | 2021-06-18T01:11:06 | {} | NONE |
ClickHouse | ClickHouse | 863,908,405 | 25,404 | TszKitLo40 | > It can be solved completely within parser (see `ParserSelectQuery.h/cpp`).
>
> If `DISTINCT ON (...)` is parsed in a query, create an AST similarly to `LIMIT BY`.
> If both `DISTINCT ON (...)` and `LIMIT BY` are present, the query should not be parsed.
Should `DISTINCT ON (...)` be parsed like https://github.c... | 2021-06-18T09:43:51 | 2021-06-18T09:43:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 864,091,764 | 23,656 | kssenii | `Yandex synchronization check ` - `fatal error: 'clickhouse/base/ext/shared_ptr_helper.h' file not found` - not related | 2021-06-18T14:46:46 | 2021-06-18T14:46:46 | {} | MEMBER |
ClickHouse | ClickHouse | 863,947,504 | 23,188 | vdimir | Related to https://github.com/ClickHouse/ClickHouse/pull/23881, but is not solved | 2021-06-18T10:47:22 | 2021-06-18T10:47:22 | {} | MEMBER |
ClickHouse | ClickHouse | 864,342,380 | 25,498 | vzakaznikov | @alexey-milovidov, the fail (https://clickhouse-test-reports.s3.yandex.net/25498/8615e31c88bf0085392d239f53be9a13d56817f3/testflows_check.html#fail1) in RBAC test looks like a bug with mutations. Here is gist that shows commands that lead to the error https://gist.github.com/vzakaznikov/a3cfb7dbd98b430aa582a04803a82da0... | 2021-06-19T02:21:11 | 2021-06-19T02:23:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 864,424,372 | 25,494 | azat | >[AST fuzzer (MSan) — Exception: Maximum parse depth (1000) exceeded. Consider rising max_parser_depth parameter.](https://clickhouse-test-reports.s3.yandex.net/25494/c6339b4c08ad8f50bf34831367de9a1ea90b1e09/fuzzer_msan/report.html#fail1)
This fails even w/o this patch.
BTW does hitting this limit should fail the f... | 2021-06-19T16:00:14 | 2021-06-19T16:00:14 | {} | MEMBER |
ClickHouse | ClickHouse | 864,444,206 | 25,494 | alexey-milovidov | It shouldn't. We need to correct fuzzer to not fail on this exception. | 2021-06-19T18:08:50 | 2021-06-19T18:08:50 | {} | MEMBER |
ClickHouse | ClickHouse | 864,537,784 | 25,404 | alexey-milovidov | DISTINCT ON ... is equivalent to LIMIT 1 BY ...
So, the limit_length will be simply set to 1. | 2021-06-20T11:14:16 | 2021-06-20T11:14:16 | {} | MEMBER |
ClickHouse | ClickHouse | 863,654,351 | 25,294 | gyuton | @kitaisreal Hi! Can you review this, please? | 2021-06-18T00:56:30 | 2021-06-18T00:56:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 863,790,133 | 21,521 | amosbird | It's useless indeed. | 2021-06-18T06:25:05 | 2021-06-18T06:25:05 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 864,147,203 | 21,390 | UnamedRus | Working in 21.7
```
CREATE OR REPLACE VIEW x10 AS
WITH d AS
(
SELECT toIntervalHour(1) AS delay
)
SELECT
now(),
now() - d.delay
FROM d
Query id: 2f4c470f-7cc9-46b4-a0ae-a4f5f1a8d24c
Ok.
0 rows in set. Elapsed: 0.010 sec.
localdomain :) SELECT * FROM x10;
SELECT *
FROM... | 2021-06-18T16:19:03 | 2021-06-18T16:19:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 864,344,925 | 25,485 | amosbird | Cannot reproduce in master. Do you happen to disable `enable_global_with_statement` ? | 2021-06-19T02:48:20 | 2021-06-19T02:48:20 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 864,447,449 | 25,494 | azat | >It shouldn't. We need to correct fuzzer to not fail on this exception.
#25510 | 2021-06-19T18:37:04 | 2021-06-19T18:37:04 | {} | MEMBER |
ClickHouse | ClickHouse | 864,537,017 | 25,513 | alexey-milovidov | Sorry, I don't understand, where 16 comes from and how `shardN_replicaM` does not support infinite number of nodes, e.g. `shard9999_replica9999` | 2021-06-20T11:07:33 | 2021-06-20T11:07:33 | {} | MEMBER |
ClickHouse | ClickHouse | 863,639,547 | 11,193 | alexey-milovidov | The default value of the size of background pool is mostly optimal, there is little sense tweaking it. | 2021-06-18T00:13:04 | 2021-06-18T00:13:04 | {} | MEMBER |
ClickHouse | ClickHouse | 863,644,070 | 25,445 | alexey-milovidov | Maybe #22381 and #25165 will help.
| 2021-06-18T00:26:13 | 2021-06-18T00:26:13 | {} | MEMBER |
ClickHouse | ClickHouse | 863,907,052 | 25,421 | alesapin | Ok, it should be more clear with https://github.com/ClickHouse/ClickHouse/pull/25459 | 2021-06-18T09:41:30 | 2021-06-18T09:41:30 | {} | MEMBER |
ClickHouse | ClickHouse | 863,950,733 | 25,422 | alexey-milovidov | See `FunctionsCoding.h/cpp`, `FunctionHex`, `FunctionUnhex` | 2021-06-18T10:54:00 | 2021-06-18T10:54:00 | {} | MEMBER |
ClickHouse | ClickHouse | 863,951,808 | 25,268 | alexey-milovidov | @MaxWk
> If we implement this feature，is it necessary to keep it?
Yes, it will act as "default default". So, if `default_database` is specified for user in `users.xml` or with `CREATE USER`, it will be used; if not - the `default_database` from `config.xml` will be used. | 2021-06-18T10:56:07 | 2021-06-18T10:56:07 | {} | MEMBER |
ClickHouse | ClickHouse | 864,350,644 | 25,268 | MaxWk | > @MaxWk
>
> > If we implement this feature，is it necessary to keep it?
>
> Yes, it will act as "default default". So, if `default_database` is specified for user in `users.xml` or with `CREATE USER`, it will be used; if not - the `default_database` from `config.xml` will be used.
ok,please assign it to me | 2021-06-19T03:54:47 | 2021-06-19T03:54:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 864,387,743 | 4,489 | alexey-milovidov | Or `FORMAT Null` to avoid sending data to client at all. | 2021-06-19T10:29:34 | 2021-06-19T10:29:34 | {} | MEMBER |
ClickHouse | ClickHouse | 864,481,575 | 25,508 | TCeason | > Looks like a particular case of #23424
Yes, I see it. But I think if not config in xml, the password should also be mask. | 2021-06-20T00:38:48 | 2021-06-20T00:38:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 863,855,024 | 25,319 | amosbird | I was wrong. This is correlated subqueries and it's not supported yet. https://github.com/ClickHouse/ClickHouse/issues/6697 | 2021-06-18T08:20:03 | 2021-06-18T08:20:03 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 863,894,350 | 25,404 | alexey-milovidov | It can be solved completely within parser (see `ParserSelectQuery.h/cpp`).
If `DISTINCT ON (...)` is parsed in a query, create an AST similarly to `LIMIT BY`.
If both `DISTINCT ON (...)` and `LIMIT BY` are present, the query should not be parsed. | 2021-06-18T09:20:36 | 2021-06-18T09:20:36 | {} | MEMBER |
ClickHouse | ClickHouse | 863,976,632 | 25,311 | alexey-milovidov | Ok. | 2021-06-18T11:41:20 | 2021-06-18T11:41:20 | {} | MEMBER |
ClickHouse | ClickHouse | 863,982,361 | 24,009 | vlstv | I changed this configuration to `async_socket_for_remote=0` but it didn't solve the issue | 2021-06-18T11:52:59 | 2021-06-18T11:52:59 | {} | NONE |
ClickHouse | ClickHouse | 864,190,092 | 23,416 | Tasselmi | > #17956
how do you solve this problem? I got the same error. | 2021-06-18T17:48:56 | 2021-06-18T17:48:56 | {} | NONE |
ClickHouse | ClickHouse | 863,912,621 | 25,404 | alexey-milovidov | I can be parsed similarly, the contents in parens should be parsed as expression list:
```
if (!exp_list.parse(pos, limit_by_expression_list, expected))
return false;
```
| 2021-06-18T09:50:49 | 2021-06-18T09:50:49 | {} | MEMBER |
ClickHouse | ClickHouse | 863,939,882 | 24,885 | Jokser | @vdimir Tests passed. I think at last PR can be merged. | 2021-06-18T10:33:29 | 2021-06-18T10:33:29 | {
"hooray": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 864,418,909 | 25,509 | alexey-milovidov | Thank you! I was wondering why it does not always work... | 2021-06-19T15:12:08 | 2021-06-19T15:12:08 | {} | MEMBER |
ClickHouse | ClickHouse | 863,636,873 | 17,862 | alexey-milovidov | I remember similar use case that we faced in Yandex.Metrica when we were storing aggregated data.
E.g. keep only top 1000 URLs for every website for every batch of inserted data.
But it was before we started using ClickHouse.
After migrated to ClickHouse, we just decided to not aggregate data at all :rofl:
Nev... | 2021-06-18T00:06:21 | 2021-06-18T00:06:45 | {} | MEMBER |
ClickHouse | ClickHouse | 863,841,657 | 25,410 | vdimir | > Stress test (thread) — Fatal message in clickhouse-server.log
Only one record with Fatal
```
2021.06.17 18:02:59.844616 [ 428 ] {} <Fatal> Application: Child process was terminated by signal 11.
``` | 2021-06-18T08:00:06 | 2021-06-18T08:00:06 | {} | MEMBER |
ClickHouse | ClickHouse | 863,910,582 | 25,445 | siradjev | > Maybe #22381 and #25165 will help.
Yes, they would reduce amount of problem, this FR is to simply provide better troubleshooting means for admins. | 2021-06-18T09:47:29 | 2021-06-18T09:47:29 | {} | NONE |
ClickHouse | ClickHouse | 863,937,634 | 25,461 | alexey-milovidov | > I encountered the following error
`Killed [1240/9765]`
You don't have enough memory to perform multithreaded build.
Lower the number of threads by running `ninja -j1` | 2021-06-18T10:29:30 | 2021-06-18T10:29:30 | {} | MEMBER |
ClickHouse | ClickHouse | 863,942,822 | 24,885 | vdimir | > @vdimir Tests passed. I think at last PR can be merged.
Great! That you for this fix | 2021-06-18T10:38:52 | 2021-06-18T10:38:52 | {} | MEMBER |
ClickHouse | ClickHouse | 863,978,389 | 13,527 | alexey-milovidov | Yes, this will work.
Also `arrayReduce('corr', arr1, arr2)` will calculate the cosine distance. | 2021-06-18T11:44:57 | 2021-06-18T11:44:57 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 864,226,180 | 25,416 | UnamedRus | https://kb.altinity.com/altinity-kb-integrations/altinity-kb-kafka/altinity-kb-kafka-main-parsing-loop
You need to change: kafka_flush_interval_ms setting
You can do it on Kafka table level. | 2021-06-18T19:09:56 | 2021-06-18T19:09:56 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 864,301,684 | 25,471 | alexey-milovidov | > Unable to find image 'yandex/clickhouse-deb-builder:25471-ae28df41ac14d338dafdd669d5330c01bce73d87' locally
docker: Error response from daemon: toomanyrequests: You have reached your pull rate limit. You may increase the limit by authenticating and upgrading: | 2021-06-18T22:24:43 | 2021-06-18T22:24:43 | {} | MEMBER |
ClickHouse | ClickHouse | 864,375,604 | 25,493 | DipalPrajapati | Hi, Thank you for replying but it is still giving the same.
I already have MV on kafka table.
What could be the issue?
And if it fetches partition clickhouse side then at kafka side , produce consumer are not working with below errors.
`
[2021-06-19 08:57:40,380] WARN [Consumer clientId=consumer-console-consu... | 2021-06-19T08:28:19 | 2021-06-19T09:02:50 | {} | NONE |
ClickHouse | ClickHouse | 864,429,097 | 25,485 | shawel | yes it is enabled on both nodes.
```
SELECT *
FROM settings
WHERE name = 'enable_global_with_statement'
Query id: 42af2cd9-0d79-4955-9609-be50b9294375
┌─name─────────────────────────┬─value─┬─changed─┬─description───────────────────────────────────────────────────┐
│ enable_global_with_statement │ 1 │ ... | 2021-06-19T16:39:42 | 2021-06-19T16:39:42 | {} | NONE |
ClickHouse | ClickHouse | 864,521,628 | 24,870 | azat | >[Functional stateless tests (ANTLR debug) Timeout, fail: 3, passed: 38, skipped: 15](https://clickhouse-test-reports.s3.yandex.net/24870/26c59961ae74a2879c3425c9f361002b10b927a9/functional_stateless_tests_(antlr_debug).html#fail1)
NULL dereference with ANTLR, fixed in #25358 | 2021-06-20T08:55:26 | 2021-06-20T08:55:34 | {} | MEMBER |
ClickHouse | ClickHouse | 863,796,436 | 13,527 | lochv | @alexey-milovidov maybe
With [1.1,1.2] as a
select uid, (arraySum((a_i, feature_i) -> (a_i * feature_i), a, feature) / sqrt(arraySum(x -> (x * x), a) * arraySum(x -> (x * x), feature))) as distance from features
| 2021-06-18T06:38:57 | 2021-06-18T06:38:57 | {} | NONE |
ClickHouse | ClickHouse | 863,997,125 | 24,341 | Avogar | @Mergifyio update | 2021-06-18T12:18:46 | 2021-06-18T12:18:46 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 864,073,655 | 25,451 | morozovsk | mysqli is deprecated in php 8. try to use pdo instead. | 2021-06-18T14:18:53 | 2021-06-18T14:18:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 864,134,752 | 21,740 | DipalPrajapati | @filimonov I am facing this issue again somehow clickhouse is not able to assign the partition.
CREATE TABLE my2 (
message String
)
ENGINE = Kafka
SETTINGS kafka_broker_list = 'host:9093',
kafka_topic_list = 'my1',
kafka_group_name = 'my2223',
kafka_num_consumers = '1',
kafka_for... | 2021-06-18T15:57:47 | 2021-06-18T16:40:51 | {} | NONE |
ClickHouse | ClickHouse | 864,186,394 | 17,710 | Tasselmi | @alexey-milovidov WE NEED HELP. SOS.
```
SELECT *
FROM id_val
ANY LEFT JOIN id_val_join USING (id)
SETTINGS join_use_nulls = 0
Query id: 5edb59e7-8efe-4f80-b1fd-d7c4eb8067d6
0 rows in set. Elapsed: 0.001 sec.
Received exception from server (version 21.3.4):
Code: 264. DB::Exception: Received from ... | 2021-06-18T17:40:17 | 2021-06-18T17:40:41 | {} | NONE |
ClickHouse | ClickHouse | 864,246,176 | 25,496 | kssenii | No such issue in 21.6. | 2021-06-18T19:59:06 | 2021-06-18T19:59:06 | {} | MEMBER |
ClickHouse | ClickHouse | 864,538,576 | 25,513 | azat | >Sorry, I don't understand, where 16 comes from and how shardN_replicaM does not support infinite number of nodes, e.g. shard9999_replica9999
Since for async INSERT (`insert_distributed_sync=0`) and `internal_replication=true` it creates directory that includes all replicas in the name, i.e. `/var/lib/clickhouse/dat... | 2021-06-20T11:21:35 | 2021-06-20T11:24:08 | {} | MEMBER |
ClickHouse | ClickHouse | 863,635,507 | 20,192 | alexey-milovidov | @filimonov No update, let's assume it was fixed. | 2021-06-18T00:03:00 | 2021-06-18T00:03:00 | {} | MEMBER |
ClickHouse | ClickHouse | 863,641,387 | 10,759 | alexey-milovidov | @azat
> Can you try turn it off and see if you can reproduce the spikes?
We have profiler enabled ("always on profiling") by default on >1000 servers in >60 production clusters without any troubles.
In fact, you can enable it with even higher frequency for some subset of queries and actually figure out the ca... | 2021-06-18T00:18:17 | 2021-06-18T00:18:17 | {} | MEMBER |
ClickHouse | ClickHouse | 864,148,747 | 21,390 | UnamedRus | Fixed with https://github.com/ClickHouse/ClickHouse/pull/22657 in clickhouse version 21.5 | 2021-06-18T16:21:49 | 2021-06-18T16:21:49 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 864,535,646 | 23,648 | Felixoid | 21.7 is not available on docker hub, unfortunately
But it's still there for 21.6
```
for str in 'string\nstring\n' 'string\nstring' 'string\n' 'string'; do echo test $str:; printf $str curl -F 'metrics_list=@-;' 'http://localhost:8123/?metrics_list_format=TSV&metrics_list_structure=Path+String&query=SELECT+*+FRO... | 2021-06-20T10:57:01 | 2021-06-20T10:57:01 | {} | MEMBER |
ClickHouse | ClickHouse | 863,654,061 | 25,065 | gyuton | @CurtizJ Hi! Can you review this, please? | 2021-06-18T00:55:29 | 2021-06-18T00:55:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 863,762,076 | 25,451 | 1995hzczjut | Which ClickHouse server version to use
v21.6.3.14-stable | 2021-06-18T05:15:25 | 2021-06-18T05:15:25 | {} | NONE |
ClickHouse | ClickHouse | 863,864,202 | 25,404 | TszKitLo40 | Can you give me some advice for solving this issue? | 2021-06-18T08:34:37 | 2021-06-18T08:34:37 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 863,867,569 | 25,422 | alexey-milovidov | Similarly to `hex`, they also accept String, FixedString, Float32/64...
`bin` from String and FixedString returns bits starting from the most significant bit of the first byte to the least significant bit of the last byte. In contrast to integers, leading zero bits are included. Floats are treated as their byte repr... | 2021-06-18T08:40:02 | 2021-06-18T08:40:12 | {} | MEMBER |
ClickHouse | ClickHouse | 863,971,983 | 7,659 | alexey-milovidov | @filimonov In this case we will have to always keep in mind that system.columns is not only for columns. It will break 3rd party tools.
Better to introduce a separate table. | 2021-06-18T11:32:26 | 2021-06-18T11:32:26 | {} | MEMBER |
ClickHouse | ClickHouse | 863,907,645 | 25,400 | alexey-milovidov | clang-tidy:
```
2021-06-17 20:23:59 /ClickHouse/contrib/murmurhash/include/MurmurHash2.h:6:9: error: macro name is a reserved identifier [-Werror,-Wreserved-id-macro]
2021-06-17 20:23:59 #define _MURMURHASH2_H_
``` | 2021-06-18T09:42:32 | 2021-06-18T09:42:32 | {} | MEMBER |
ClickHouse | ClickHouse | 863,979,096 | 25,404 | alexey-milovidov | Sure, let's do it!
| 2021-06-18T11:46:24 | 2021-06-18T11:46:24 | {} | MEMBER |
ClickHouse | ClickHouse | 864,365,679 | 25,493 | RO-29 | > I am using a new group each time.
```
<kafka>
..
<auto_offset_reset>earliest</auto_offset_reset>
..
</kafka>
```
This must be applied to kafka settings but be careful about it, it could lead to duplicate if you will re-start with same consumer group
>Can't get assignment. It can be caused by... | 2021-06-19T06:49:29 | 2021-06-19T06:49:29 | {} | NONE |
ClickHouse | ClickHouse | 863,863,209 | 25,422 | TszKitLo40 | Should `bin` and `unbin` only accepts `int` or `long`? | 2021-06-18T08:32:59 | 2021-06-18T08:32:59 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 863,877,890 | 25,319 | alexey-milovidov | @JZuming So, it's not a bug but a missing feature.
We found your fuzzer very useful, because it also can find real issues in code like #25411
So, if you can filter out this type of issues (unsupported correlated subqueries) and continue searching others, it will be very appreciated. | 2021-06-18T08:55:55 | 2021-06-18T08:55:55 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 863,986,043 | 25,450 | nikitamikhaylov | This issue is already fixed and will be available in next patch release (21.3.13) | 2021-06-18T11:59:58 | 2021-06-18T11:59:58 | {} | MEMBER |
ClickHouse | ClickHouse | 864,178,573 | 17,710 | Tasselmi | I got the same error , make me confused. | 2021-06-18T17:23:02 | 2021-06-18T17:23:02 | {} | NONE |
ClickHouse | ClickHouse | 864,357,690 | 23,864 | azat | >[AST fuzzer (debug) — Logical error: 'Unexpected dictionary type for DataTypeLowCardinality: IntervalHour'.](https://clickhouse-test-reports.s3.yandex.net/23864/51e81ae4c276c9c63edc7b737a0f7b7f30626c84/fuzzer_debug/report.html#fail1)
Fixed in #25410
>[AST fuzzer (UBSan) — ../base/common/../common/DecomposedFloa... | 2021-06-19T05:15:49 | 2021-06-19T18:49:09 | {} | MEMBER |
ClickHouse | ClickHouse | 873,308,416 | 21,558 | alexey-milovidov | @kitaisreal FYI it is related to type inference in KeyCondition.
KeyCondition may think that `data` has Array type while in fact it is unnested by ARRAY JOIN. | 2021-07-02T23:56:46 | 2021-07-02T23:56:46 | {} | MEMBER |
ClickHouse | ClickHouse | 873,335,875 | 25,523 | filimonov | > @filimonov
> Thanks, should I use the kafka_thread_per_consumer = 1 ?
That should help to avoid the problem
| 2021-07-03T03:17:47 | 2021-07-03T03:17:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 872,422,033 | 25,064 | kssenii | It is better to use `std::map<String, std::set<String>>;` instead of using `std::unordered_map<String, std::unordered_set<String>>;` then the test will be fixed and it was discussed that ordered is better.
btw, no idea why ` Needs 'can be tested' label` appeared... | 2021-07-01T17:23:03 | 2021-07-01T17:23:03 | {} | MEMBER |
ClickHouse | ClickHouse | 872,450,778 | 25,897 | Jack012a | This is not true. I didn't run clickhouse-local from a directory where it has config.xml. | 2021-07-01T18:11:54 | 2021-07-01T18:11:54 | {} | NONE |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.