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 | 718,989,267 | 2,437 | alexey-milovidov | You can find the grammar in ANTLR format here:
https://github.com/ClickHouse/ClickHouse/blob/abcb3bcbe2c6f4c28caf6840b7482d83477e6bf4/src/Parsers/New/ClickHouseParser.g4
It's not directly used by ClickHouse but *roughly* corresponds to its syntax. | 2020-10-29T20:03:01 | 2020-10-29T20:03:01 | {} | MEMBER |
ClickHouse | ClickHouse | 719,004,516 | 1,576 | alexey-milovidov | @KochetovNicolai the issue is not fixed.
Scalar subqueries are still not integrated into query pipeline.
They cannot be cancelled and the progress is not shown. | 2020-10-29T20:32:13 | 2020-10-29T20:32:40 | {} | MEMBER |
ClickHouse | ClickHouse | 718,976,519 | 16,484 | alexey-milovidov | These functions are also deficient for the following reason: #2479 | 2020-10-29T19:37:40 | 2020-10-29T19:37:40 | {} | MEMBER |
ClickHouse | ClickHouse | 719,061,024 | 16,529 | azat | @alexey-milovidov BTW is there any way to ignore specific test for the flacky checks? Since newly added test will fail if it will run multiple times (obviously, due overlap of entries in the query_log) | 2020-10-29T22:31:07 | 2020-10-29T22:31:07 | {} | MEMBER |
ClickHouse | ClickHouse | 719,113,065 | 16,517 | ahutsunshine | @den-crane Can you give me the release logs for this?
1. What's the reason to support such case which changed the behavior and why didnt support before?
2. We can use the `Nullable` feature in `20.6.8.5`. What do you mean `19.16.14.65, 20.6.8.5 both are out of support`? | 2020-10-30T01:20:39 | 2020-10-30T01:20:39 | {} | NONE |
ClickHouse | ClickHouse | 719,115,378 | 16,517 | den-crane | https://github.com/ClickHouse/ClickHouse/blob/master/CHANGELOG.md#clickhouse-release-v20527-stable-2020-07-02
It was implemented here :
_Return non-Nullable result from COUNT(DISTINCT), and uniq aggregate functions family. If all passed values are NULL, return zero instead. This improves SQL compatibility. #11661 (al... | 2020-10-30T01:28:55 | 2020-10-30T01:29:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 719,412,949 | 16,537 | azat | > i start my process 2 hours,only commit not query。
Not sure what "only commit" means, but:
>2020.10.29 23:45:16.320714 [ 10473 ] {171eac9e-a25d-45b0-bd59-7ee8f6533105} executeQuery: Code: 241, e.displayText() = DB::Exception: Memory limit (total) exceeded: would use 24.72 GiB (attempt to allocate chunk of 507557... | 2020-10-30T08:32:23 | 2020-10-30T08:35:21 | {} | MEMBER |
ClickHouse | ClickHouse | 718,298,537 | 16,484 | alexey-milovidov | > Users will just enable it back and will continue complaining.
That's ok, at least they will pay attention to the name of the setting. | 2020-10-29T01:17:36 | 2020-10-29T01:17:36 | {} | MEMBER |
ClickHouse | ClickHouse | 718,967,557 | 9,349 | alexey-milovidov | Yes, ClickHouse supports WITH FILL modifier to ORDER BY.
https://clickhouse.tech/docs/en/sql-reference/statements/select/order-by/#orderby-with-fill | 2020-10-29T19:19:35 | 2020-10-29T19:19:35 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 718,968,827 | 9,402 | alexey-milovidov | Yes, it is left from unfinished delete. This folder is not used by the server anymore. | 2020-10-29T19:22:02 | 2020-10-29T19:22:02 | {} | MEMBER |
ClickHouse | ClickHouse | 718,980,926 | 2,446 | alexey-milovidov | We already use it but the mapping is rough.
We use 400, 401, 404, 411, 413, 500, 501, 503.
| 2020-10-29T19:46:16 | 2020-10-29T19:46:16 | {} | MEMBER |
ClickHouse | ClickHouse | 719,026,006 | 16,510 | alexey-milovidov | @ildus The behaviour comes from `sumMap` function that was specifically made this way. But it does not make sense for `minMap`, `maxMap`. | 2020-10-29T21:13:41 | 2020-10-29T21:13:41 | {} | MEMBER |
ClickHouse | ClickHouse | 719,462,984 | 16,541 | ucasfl | test `01534_lambda_array_join` Memory limit (for user) exceeded. | 2020-10-30T10:09:22 | 2020-10-30T10:09:22 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 718,297,914 | 15,938 | alexey-milovidov | @Avogar It because performance test runs two servers: old and new.
This error is Ok for old server (no further actions required). | 2020-10-29T01:15:28 | 2020-10-29T01:15:57 | {} | MEMBER |
ClickHouse | ClickHouse | 718,766,681 | 16,395 | macEar | @filimonov
I have additional questions:
What do you mean by "check if the server / client clocks are in sync."? Epoch sync or timezone? Server and client must have NTP sync both.
What possible reasons of those errors in log? Only idea is time sync? | 2020-10-29T13:51:49 | 2020-10-29T14:27:03 | {} | NONE |
ClickHouse | ClickHouse | 718,874,332 | 16,498 | alexey-milovidov | No relevant changes. | 2020-10-29T16:35:59 | 2020-10-29T16:35:59 | {} | MEMBER |
ClickHouse | ClickHouse | 719,001,857 | 1,802 | alexey-milovidov | Duplicate of #1647 | 2020-10-29T20:27:02 | 2020-10-29T20:27:02 | {} | MEMBER |
ClickHouse | ClickHouse | 718,274,119 | 16,421 | l1t1 | how to solve the problem?
```
$ cat /etc/init.d/clickhouse-server |sed -n '381,390p'
(
if $FLOCK -n 9; then
main "$@"
else
echo "Init script is already running" && exit 1
fi
) 9> $LOCKFILE
``` | 2020-10-28T23:55:03 | 2020-10-28T23:55:03 | {} | NONE |
ClickHouse | ClickHouse | 718,866,003 | 16,423 | KochetovNicolai | `Yandex third-party checks` is broken in master
`Integration tests (asan) ` is a ci issue
| 2020-10-29T16:22:39 | 2020-10-29T16:22:39 | {} | MEMBER |
ClickHouse | ClickHouse | 718,871,018 | 16,485 | socketpair | @alexey-milovidov "totally impractical" ?
https://fedoraproject.org/wiki/LTOByDefault
I would not say so. I would forcibly enable it. If chosen compiler does not support it, the build should fail, as I think. | 2020-10-29T16:30:35 | 2020-10-29T16:30:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,983,872 | 2,329 | alexey-milovidov | How to fix:
1. Install ClickHouse version 20.10 or newer.
2. Set up `max_bytes_before_external_group_by` to the appropriate value. | 2020-10-29T19:52:30 | 2020-10-29T19:52:30 | {} | MEMBER |
ClickHouse | ClickHouse | 718,994,693 | 1,540 | alexey-milovidov | FYI @amosbird it is similar to the "trivial count optimization". | 2020-10-29T20:12:58 | 2020-10-29T20:12:58 | {} | MEMBER |
ClickHouse | ClickHouse | 719,018,243 | 3,133 | alexey-milovidov | Fixed in https://github.com/ClickHouse/ClickHouse/pull/7907. | 2020-10-29T20:58:42 | 2020-10-29T20:58:42 | {} | MEMBER |
ClickHouse | ClickHouse | 719,075,462 | 11,298 | abyss7 | Status for now (0_stateless): `443 failed, 1533 passed in 698.17s (0:11:38)` | 2020-10-29T23:12:49 | 2020-10-29T23:12:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,195,385 | 16,438 | azat | >Also you can generate lookup table, see ProfileEvents, CurrentMetrics.
Remember one more point against, error codes values are not sequential (548, 999), (1002, 2001) ranges are not in use, so lookup table will be suboptimal | 2020-10-28T20:39:37 | 2020-10-28T20:39:37 | {} | MEMBER |
ClickHouse | ClickHouse | 718,706,367 | 16,252 | filimonov | openssl has needed tools to create a mock server (and it's installed in out docker image)
Check that out
```bash
# bit better/safer to have those hosts predefined...
grep -qE '\slocalhost($|\s|\n|\r)' /etc/hosts || echo '127.0.0.1 localhost' >> /etc/hosts
# actually we can use just localhost instead (simpler... | 2020-10-29T11:59:03 | 2020-10-29T13:33:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 719,007,113 | 5,046 | alexey-milovidov | I agree, lambda argument name should shadow captured column names. | 2020-10-29T20:37:21 | 2020-10-29T20:37:21 | {} | MEMBER |
ClickHouse | ClickHouse | 719,315,339 | 16,411 | excitoon | @alexey-milovidov can you please merge it? #16235 was actually unfinished and contained redundant commits which will be reverted by this PR. | 2020-10-30T06:58:17 | 2020-10-30T06:58:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 719,437,636 | 15,179 | chang0391 | @zhang2014
have you already began to fix the issue? | 2020-10-30T09:14:45 | 2020-10-30T09:14:45 | {} | NONE |
ClickHouse | ClickHouse | 719,447,404 | 16,497 | fastio | @alexey-milovidov thanks for your reply. It doesn't reproduce in version 20.8.4.11. | 2020-10-30T09:34:54 | 2020-10-30T09:34:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,208,718 | 8,828 | fibersel | @alexey-milovidov Oh, I got it.Which library would be the best for LZMA2?
I have spotted https://tukaani.org/xz/ and https://github.com/frida/xz | 2020-10-28T21:06:08 | 2020-10-28T21:06:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,585,066 | 15,977 | TCeason | Now [it](https://clickhouse-test-reports.s3.yandex.net/15977/34926c9a65b07cf4a54b03453b3afc9866619c03/integration_tests_flaky_check_(asan)/integration_run.0.txt.out.log) seems like run the grants sql do not take effect at `test_materialize_database_err_sync_user_privs_5_7`'s test but it return success when run `test_ma... | 2020-10-29T10:10:38 | 2020-10-29T10:10:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,959,134 | 9,349 | ivishnevs | Sounds good for me too | 2020-10-29T19:03:31 | 2020-10-29T19:03:31 | {} | NONE |
ClickHouse | ClickHouse | 718,977,798 | 16,512 | alexey-milovidov | It is not possible to kill DDL (CREATE, DROP, RENAME...) queries.
The issue with DROP query is related to a deadlock in Kafka engine. I assign @filimonov as Kafka expert.
Workaround: restart clickhouse-server. | 2020-10-29T19:39:58 | 2020-10-29T19:40:25 | {} | MEMBER |
ClickHouse | ClickHouse | 718,985,066 | 2,175 | alexey-milovidov | We are going to treat Nested data type in the same way as Array of named Tuples.
The work is in progress in #14196 | 2020-10-29T19:55:03 | 2020-10-29T19:55:03 | {} | MEMBER |
ClickHouse | ClickHouse | 719,003,111 | 2,515 | alexey-milovidov | The feature was implemented in the following PR: https://github.com/ClickHouse/ClickHouse/pull/8326 | 2020-10-29T20:29:28 | 2020-10-29T20:29:28 | {} | MEMBER |
ClickHouse | ClickHouse | 719,405,139 | 16,532 | alesapin | > probably the table is corrupted #9019
Can you show your `CREATE TABLE` query? Index corruption during mutation is an extremely rare bug which was quickly fixed. But if it happened to you, not the whole table is currupted, but only the index. You can insert select all data to another table and try to delete data ag... | 2020-10-30T08:24:04 | 2020-10-30T08:24:04 | {} | MEMBER |
ClickHouse | ClickHouse | 719,473,243 | 15,179 | tavplubix | > data inconsistency between MySQL and CH
Looks like duplicate of #14829. There is no actual data inconsistency, just `select count()` returns wrong result. You can check it by explicitly executing `select count() from materialize_mysql_db.table final where _sign=1` instead of simple `select count() from materialize... | 2020-10-30T10:30:05 | 2020-10-30T10:30:05 | {} | MEMBER |
ClickHouse | ClickHouse | 719,475,933 | 15,211 | tavplubix | #15409 | 2020-10-30T10:35:37 | 2020-10-30T10:35:37 | {} | MEMBER |
ClickHouse | ClickHouse | 718,272,352 | 16,438 | alexey-milovidov | > add a function instead
Better to do both :)
> error codes values are not sequential
Does not matter, it's just about 1000 elements. | 2020-10-28T23:49:01 | 2020-10-28T23:49:01 | {} | MEMBER |
ClickHouse | ClickHouse | 719,403,763 | 11,133 | azat | >[ 23165 ] {5a41e916-458e-4372-9f94-e327b725daa1} <Error> executeQuery: Poco::Exception. Code: 1000, e.code() = 36, e.displayText() = Bad path syntax: /var/lib/clickhouse/data/database/test.....
>because path name was too long, then i try to change password and replication user name to shorter, but it was not enough. ... | 2020-10-30T08:22:12 | 2020-10-30T08:22:12 | {} | MEMBER |
ClickHouse | ClickHouse | 719,469,781 | 16,532 | alex-karo | One of corrupted tables
```
CREATE TABLE kline
(
`date` Date,
`time` DateTime,
`exchangeId` UInt16,
`tickerId` UInt32,
`base` String,
`quote` String,
`high` Nullable(Decimal(38, 12)),
`low` Nullable(Decimal(38, 12)),
`open` Nullable(Decimal(38, 12)),
`close` Nullable... | 2020-10-30T10:22:56 | 2020-10-30T10:22:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,969,882 | 5,182 | alexey-milovidov | Basic support for OpenTelemetry has been merged and available in version 20.11 testing.
What is left to implement is convenient export of tracing data.
@akuzm is going to provide documentation / examples. | 2020-10-29T19:24:09 | 2020-10-29T19:24:09 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 718,993,843 | 5,553 | alexey-milovidov | https://clickhouse.tech/docs/en/engines/table-engines/integrations/kafka/#table_engine-kafka-creating-a-table | 2020-10-29T20:11:13 | 2020-10-29T20:11:13 | {} | MEMBER |
ClickHouse | ClickHouse | 718,994,804 | 5,182 | sasah | > Basic support for OpenTelemetry has been merged and available in version 20.11 testing.
> What is left to implement is convenient export of tracing data.
>
> @akuzm is going to provide documentation / examples.
https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/master/receiver | 2020-10-29T20:13:13 | 2020-10-29T20:13:13 | {} | NONE |
ClickHouse | ClickHouse | 718,874,365 | 16,498 | 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-... | 2020-10-29T16:36:01 | 2020-10-29T16:36:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,968,398 | 9,402 | alexey-milovidov | Simply ignore or delete it manually. | 2020-10-29T19:21:16 | 2020-10-29T19:21:16 | {} | MEMBER |
ClickHouse | ClickHouse | 719,021,094 | 2,652 | alexey-milovidov | > We would like to add an option to use header for reordering columns in the future.
The options has been added:
`input_format_with_names_use_header`
`input_format_skip_unknown_fields` | 2020-10-29T21:04:13 | 2020-10-29T21:04:13 | {} | MEMBER |
ClickHouse | ClickHouse | 719,411,014 | 16,521 | alesapin | `00840_long_concurrent_select_and_drop_deadlock` known flaky test | 2020-10-30T08:31:02 | 2020-10-30T08:31:02 | {} | MEMBER |
ClickHouse | ClickHouse | 719,429,507 | 16,537 | xiashaohua | > > i start my process 2 hours,only commit not query。
>
> Not sure what "only commit" means, but:
>
> > 2020.10.29 23:45:16.320714 [ 10473 ] {171eac9e-a25d-45b0-bd59-7ee8f6533105} executeQuery: Code: 241, e.displayText() = DB::Exception: Memory limit (total) exceeded: would use 24.72 GiB (attempt to allocate chun... | 2020-10-30T08:56:58 | 2020-10-30T08:56:58 | {} | NONE |
ClickHouse | ClickHouse | 718,273,566 | 8,828 | alexey-milovidov | It is the same code, just different copies. Let's use this one: https://github.com/xz-mirror/xz
If we will need to make any changes, we will create a fork in https://github.com/ClickHouse-Extras/ | 2020-10-28T23:53:18 | 2020-10-28T23:53:18 | {} | MEMBER |
ClickHouse | ClickHouse | 718,420,352 | 16,372 | TCeason | Try to build the newest version? | 2020-10-29T07:27:35 | 2020-10-29T07:27:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 719,002,187 | 1,802 | alexey-milovidov | We are going to simplify ALIAS processing that will make indices to work. | 2020-10-29T20:27:38 | 2020-10-29T20:27:38 | {} | MEMBER |
ClickHouse | ClickHouse | 719,017,057 | 2,220 | alexey-milovidov | The difference is that internal replication will repair data automatically after failures. It keeps data 100% consistent and byte-identical.
While the "replication" through Distributed table just duplicates INSERTs and does not guarantee consistency and repair.
We can call it "poor" replication as opposed to the fu... | 2020-10-29T20:56:17 | 2020-10-29T20:56:17 | {} | MEMBER |
ClickHouse | ClickHouse | 719,030,698 | 15,301 | alexey-milovidov | Duplicate of #13098. | 2020-10-29T21:23:27 | 2020-10-29T21:23:27 | {} | MEMBER |
ClickHouse | ClickHouse | 719,041,170 | 16,517 | den-crane | CH started to support Nullable aggregateFunction since 20.5.2.7
19.16.14.65 , 20.6.8.5 both are out of support!
Supported versions 20.3.20.6 (LTS) , 20.8.5.45 , 20.9.4.76, 20.10.3.30
Probably you should upgrade and use 20.10.3.30 | 2020-10-29T21:42:17 | 2020-10-29T21:42:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 719,042,278 | 16,510 | den-crane | duplicate of https://github.com/ClickHouse/ClickHouse/issues/16087 | 2020-10-29T21:44:52 | 2020-10-29T21:45:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 719,043,621 | 16,532 | den-crane | probably the table is corrupted https://github.com/ClickHouse/ClickHouse/issues/9019 | 2020-10-29T21:47:59 | 2020-10-29T21:47:59 | {
"confused": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 719,242,211 | 16,438 | azat | `Yandex synchronization check (only for Yandex employees)` fails, any clue? | 2020-10-30T06:02:56 | 2020-10-30T06:02:56 | {} | MEMBER |
ClickHouse | ClickHouse | 719,453,981 | 15,211 | chang0391 | @zhang2014
Have the function of support MaterializeMySQL table engine already began to do it? | 2020-10-30T09:49:32 | 2020-10-30T09:49:32 | {} | NONE |
ClickHouse | ClickHouse | 718,188,511 | 16,438 | azat | >errorCodeToName might be more descriptive...
>Because the argument is already an "error code".
Ok (I though about this, but decided to make it less descriptive, since it returns some "mnemonic" name)
>"X macro" that generates switch/case looks appropriate.
>Compiler is using "jump table" for int-based switches... | 2020-10-28T20:26:33 | 2020-10-28T20:26:33 | {} | MEMBER |
ClickHouse | ClickHouse | 718,968,129 | 16,528 | CurtizJ | Sleep was needed, when `OPTIMIZE` didn't wait for currently running merges, now it's useless. | 2020-10-29T19:20:46 | 2020-10-29T19:20:46 | {} | MEMBER |
ClickHouse | ClickHouse | 718,998,871 | 2,547 | alexey-milovidov | @filimonov AFAIK it should be not an issue in recent versions. | 2020-10-29T20:21:14 | 2020-10-29T20:21:14 | {} | MEMBER |
ClickHouse | ClickHouse | 719,009,275 | 2,305 | alexey-milovidov | > Will I have faster requests if I make 3 shards?
Heavy queries will be faster and light queries will be slightly slower.
> Can I have on each server 1 shard replicated to other 2 servers?
Yes, it is named "cross-replication" (search for "ClickHouse cross replication"). I don't recommend this configuration bec... | 2020-10-29T20:41:38 | 2020-10-29T20:41:38 | {} | MEMBER |
ClickHouse | ClickHouse | 719,020,428 | 2,074 | alexey-milovidov | This issue has been rendered irrelevant. | 2020-10-29T21:02:52 | 2020-10-29T21:02:52 | {} | MEMBER |
ClickHouse | ClickHouse | 719,156,692 | 16,529 | amosbird | Perhaps we can have another `long_query_log` table for this purpose? And `system.query_log` will still have `QueryStart` of those small queries. | 2020-10-30T03:57:25 | 2020-10-30T03:57:38 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 719,172,875 | 16,536 | alexey-milovidov | ClickHouse-over-KikiMR build should be adjusted. | 2020-10-30T04:59:50 | 2020-10-30T04:59:50 | {} | MEMBER |
ClickHouse | ClickHouse | 719,425,662 | 13,862 | DoubleWei-go | > > Probably need better example.
>
> It means optimizing the JOIN order is non-trivial :).
Are there any optimization rules similar to reorder join of the computing engine? | 2020-10-30T08:48:02 | 2020-10-30T08:48:02 | {} | NONE |
ClickHouse | ClickHouse | 718,283,272 | 15,977 | TCeason | "Integration tests flaky check (asan)" failed but "Integration tests (.*)" is succeeded. Which one should I believe？ @tavplubix | 2020-10-29T00:24:49 | 2020-10-29T00:24:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,605,626 | 16,478 | tavplubix | Functional stateless tests - 01527_dist_sharding_key_dictGet_reload and 01293_show_clusters - broken in release branches
Integration tests (release) - test_https_replication/test.py::test_replication_after_partition - flacky | 2020-10-29T10:24:54 | 2020-10-29T10:24:54 | {} | MEMBER |
ClickHouse | ClickHouse | 719,018,764 | 2,927 | alexey-milovidov | Это сообщение никак не связано со схлопыванием данных. То есть, появление таких ошибок не мешает схлопыванию. | 2020-10-29T20:59:38 | 2020-10-29T20:59:38 | {} | MEMBER |
ClickHouse | ClickHouse | 719,025,459 | 16,497 | alexey-milovidov | No, I did not see this error before.
Does it reproduce in version 20.10? | 2020-10-29T21:12:31 | 2020-10-29T21:12:31 | {} | MEMBER |
ClickHouse | ClickHouse | 719,222,461 | 16,529 | azat | >Perhaps we can have another long_query_log table for this purpose? And system.query_log will still have QueryStart of those small queries.
@amosbird I don't see why the queries with QueryStart should be in the query_log in this case, do you have any use case in mind?
And I don't like the idea of adding one more ta... | 2020-10-30T05:47:42 | 2020-10-30T05:47:42 | {} | MEMBER |
ClickHouse | ClickHouse | 719,299,070 | 16,529 | alexey-milovidov | > And I don't like the idea of adding one more table just for this
+1 | 2020-10-30T06:46:10 | 2020-10-30T06:46:10 | {} | MEMBER |
ClickHouse | ClickHouse | 719,308,560 | 16,529 | amosbird | > And I don't like the idea of adding one more table just for this
Yeah, It doesn't make much sense.
> I don't see why the queries with QueryStart should be in the query_log
If a query finishes in less duration than log_queries_min_query_duration_ms, will we have
`QueryStart` logged (since we don't known the... | 2020-10-30T06:53:19 | 2020-10-30T06:53:40 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 719,347,084 | 16,529 | amosbird | > it should not, logging of EXCEPTION_BEFORE_START/QUERY_START are ignored explicitly if log_queries_min_query_duration_ms is set.
Cool. Does it also affect `query_thread_log`? | 2020-10-30T07:28:09 | 2020-10-30T07:28:09 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 718,982,778 | 2,297 | alexey-milovidov | We need a UI for cluster monitoring with the following levels:
1. Clusters overview: the list of clusters with their nodes and some health info of the nodes.
2. Server metrics: graphs for available metrics of single server or a group of servers.
3. Logs navigation and search: display query_log, text_log with filteri... | 2020-10-29T19:50:18 | 2020-10-29T19:50:40 | {} | MEMBER |
ClickHouse | ClickHouse | 719,013,193 | 2,379 | alexey-milovidov | > Is there an issue with optimize final in relation to replicated tables?
It works in the same way as for non-Replicated tables. You should consider the cost of data processing.
> why is there a difference?
Long time ago there was a limitation that empty parts cannot exist in a table. That's why we have to lea... | 2020-10-29T20:49:38 | 2020-10-29T20:49:38 | {} | MEMBER |
ClickHouse | ClickHouse | 719,021,647 | 3,401 | alexey-milovidov | Need to check if it's relevant for version 20.10. | 2020-10-29T21:05:23 | 2020-10-29T21:05:23 | {} | MEMBER |
ClickHouse | ClickHouse | 719,247,558 | 16,438 | azat | >Yandex synchronization check (only for Yandex employees) fails, any clue?
I guess it can be a test, since errorCodeToName was disabled for the arcadia build, but the test not
But with current implementation errorCodeToName can be build for arcadia too, adjusted, will see how it goes | 2020-10-30T06:06:58 | 2020-10-30T06:06:58 | {} | MEMBER |
ClickHouse | ClickHouse | 719,274,117 | 16,395 | macEar | I restarted Clickhouse-server and it can't start, it keeps on restarting always.
In error logs I see:
```
2020.10.30 06:19:03.378210 [ 48 ] {} <Warning> AWSClient: If the signature check failed. This could be because of a time skew. Attempting to adjust the signer.
2020.10.30 06:19:03.378238 [ 48 ] {} <Warning>... | 2020-10-30T06:27:05 | 2020-10-30T06:31:44 | {} | NONE |
ClickHouse | ClickHouse | 719,335,219 | 6,472 | stephanieee777 | HI, I also have this problem, too. My cluster have 2 shards, each shard have 2 replicas. Version of server is v20.9.2.20. (already latest stable version)
2020.10.30 14:52:04.927336 [ 103 ] {} <Error>xxxxxxxx.`.inner.xxxxxxxx`: Code: 40, e.displayText() = DB::Exception: Checksums of parts don't match: hash of uncomp... | 2020-10-30T07:16:04 | 2020-10-30T07:16:04 | {} | NONE |
ClickHouse | ClickHouse | 719,460,260 | 16,532 | alex-karo | Seems to I have the same bug with index corruption.
Prerequisites:
1) Table, that ALTER'ed before (on tables, than never changed, everything is fine).
2) Updated clickhouse version from 19.8.3.8 to 20.9.3.45
3) Moved data to replicated table via official guide (move/attach partitions)
I checked, only version up... | 2020-10-30T10:03:10 | 2020-10-30T10:22:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,175,739 | 16,460 | tavplubix | Related: #11035 | 2020-10-28T20:01:18 | 2020-10-28T20:01:18 | {} | MEMBER |
ClickHouse | ClickHouse | 718,593,671 | 16,511 | alexey-milovidov | It's available in the latest testing ("nighly" build), version 20.11. | 2020-10-29T10:16:42 | 2020-10-29T10:16:42 | {} | MEMBER |
ClickHouse | ClickHouse | 718,986,669 | 2,203 | alexey-milovidov | No.
The reason is that the "left" subquery is not buffered in memory. It is processed in a streaming fashion with lookups in the right table. That's why we have to first make a lookup-optimized data structure for the "right" subquery and then proceed with streaming reading and joining of the "left" subquery. It allo... | 2020-10-29T19:58:10 | 2020-10-29T19:58:43 | {} | MEMBER |
ClickHouse | ClickHouse | 718,987,558 | 2,152 | alexey-milovidov | It means that the client disconnected. | 2020-10-29T19:59:57 | 2020-10-29T19:59:57 | {} | MEMBER |
ClickHouse | ClickHouse | 719,006,236 | 6,093 | alexey-milovidov | It is totally possible to show a progress for OPTIMIZE query. At least when it is executed on the same server. | 2020-10-29T20:35:37 | 2020-10-29T20:35:37 | {} | MEMBER |
ClickHouse | ClickHouse | 719,329,771 | 16,529 | azat | >If a query finishes in less duration than log_queries_min_query_duration_ms, will we have
QueryStart logged (since we don't known the duration yet when query starts)? I feel that it'll always be logged and might confuse users.
@amosbird it should not, logging of [EXCEPTION_BEFORE_START](https://github.com/ClickHou... | 2020-10-30T07:10:31 | 2020-10-30T07:10:31 | {} | MEMBER |
ClickHouse | ClickHouse | 718,309,150 | 15,087 | fastio | @Akazz Hi, could you review this pr? Thanks | 2020-10-29T01:55:12 | 2020-10-29T01:55:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,568,251 | 15,236 | alesapin | 
| 2020-10-29T09:56:40 | 2020-10-29T09:56:40 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 718,803,867 | 15,236 | alesapin | `00993_system_parts_race_condition_drop_zookeeper` -- flaky in master | 2020-10-29T14:49:36 | 2020-10-29T14:49:36 | {} | MEMBER |
ClickHouse | ClickHouse | 718,984,240 | 2,301 | alexey-milovidov | @KochetovNicolai Please check how it works with the new pipeline. | 2020-10-29T19:53:22 | 2020-10-29T19:53:22 | {} | MEMBER |
ClickHouse | ClickHouse | 719,000,518 | 2,538 | alexey-milovidov | If you specify FORMAT JSON in a query, it will return the result in specified format:
```
CREATE TABLE default.table_name ON CLUSTER 'test_shard_localhost' (x UInt32) ENGINE = Memory FORMAT JSON
``` | 2020-10-29T20:24:23 | 2020-10-29T20:24:23 | {} | MEMBER |
ClickHouse | ClickHouse | 719,012,904 | 16,513 | den-crane | duplicate https://github.com/ClickHouse/ClickHouse/issues/11755 | 2020-10-29T20:49:04 | 2020-10-29T20:49:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 719,145,434 | 16,184 | hczhcz | @Avogar
The actual purpose of this function is mentioned in the docs.
> This function can be used as a faster uniqExact on sorted data.
I made some changes so that the result is consistent with `groupArray`.
```sql
SELECT
arrayReduce('countEqualRanges', groupArray(x)),
countEqualRanges(x)
FROM ... | 2020-10-30T03:15:22 | 2020-10-30T03:15:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 719,371,789 | 16,529 | azat | >Cool. Does it also affect query_thread_log?
Good question, not till c7ba2894f5a01cbee1de08bdf4b46a46f594171b (just added) | 2020-10-30T07:53:29 | 2020-10-30T07:53:29 | {} | MEMBER |
ClickHouse | ClickHouse | 720,537,579 | 16,536 | abyss7 | > @abyss7 It's strange but KikiMR tests are broken in this PR.
Will fix ASAP | 2020-11-02T15:21:43 | 2020-11-02T15:21:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 721,035,432 | 16,598 | alexey-milovidov | > 00840_long_concurrent_select_and_drop_deadlock
- flaky.
> 01493_alter_remove_properties_zookeeper
| 2020-11-03T10:36:35 | 2020-11-03T10:36:35 | {} | MEMBER |
ClickHouse | ClickHouse | 720,417,352 | 5,576 | alexey-milovidov | Still relevant in 20.11. | 2020-11-02T11:34:10 | 2020-11-02T11:34:10 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.