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 | 666,335,510 | 13,081 | alexey-milovidov | Don't wait for performance test. | 2020-07-30T12:30:50 | 2020-07-30T12:30:50 | {} | MEMBER |
ClickHouse | ClickHouse | 665,330,644 | 12,960 | alexey-milovidov | Let it live for now (@blinkov has fixed it), but documentation is started to be unmaintainable, we need to cut some garbage out of it. What kind of garbage we can throw away:
- machine translated files (it's a nonsense added by mistake);
- split by versions (it also does not make sense, it does not work properly and ... | 2020-07-28T22:40:20 | 2020-07-28T22:40:20 | {} | MEMBER |
ClickHouse | ClickHouse | 665,487,845 | 12,509 | toby080508050805 | I found some clues. There has a query which stock at clickhouse. The query is sent from remote server via distribute table. It seems like wait over 15 mins so our procedures try to cancel(kill) queries but it still stock at system. After a while, the 9000 port cannot be connected.
```
2020.07.28 10:58:16.060644 [ ... | 2020-07-29T07:30:35 | 2020-07-29T07:30:35 | {} | NONE |
ClickHouse | ClickHouse | 665,488,190 | 13,057 | qoega | Do you have access_management enabled?
https://clickhouse.tech/docs/en/operations/access-rights/#enabling-access-control
It is done by setting `access_management` to 1 in `users.xml` | 2020-07-29T07:31:17 | 2020-07-29T07:31:17 | {
"+1": 15,
"hooray": 4
} | MEMBER |
ClickHouse | ClickHouse | 665,492,759 | 13,059 | artbeglaryan | Checked on 20.3.11.97 version too. The same situation | 2020-07-29T07:40:58 | 2020-07-29T07:40:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,554,565 | 13,060 | filimonov | Reproduces:
```sql
CREATE TABLE IF NOT EXISTS mytable_local (
created DateTime,
eventday Date,
eventhour DateTime,
eventminute DateTime,
user_id UInt32,
toaddress String,
integrationkey String
)
ENGINE = MergeTree()
PARTITION BY toYY... | 2020-07-29T09:34:14 | 2020-07-29T09:34:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,713,154 | 11,735 | ildus | Thanks! | 2020-07-29T14:53:56 | 2020-07-29T14:53:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,615,819 | 13,056 | alexey-milovidov | @akuzm An unintentional test for careful review: 1ae6e47 | 2020-07-29T11:43:45 | 2020-07-29T16:11:30 | {
"eyes": 1
} | MEMBER |
ClickHouse | ClickHouse | 665,841,801 | 13,086 | azat | >(Bash and all readline-based tools that I know don't do that and).
By default, but can be enabled in bash:
```bash
shopt -s histappend
# or proper append ...
PROMPT_COMMAND='history -a'
```
>it's a huge drawback
Agree, bash is completely unusable w/o this
>This is significant change for changelog, p... | 2020-07-29T18:55:46 | 2020-07-29T18:55:46 | {} | MEMBER |
ClickHouse | ClickHouse | 665,912,211 | 13,074 | zergon321 | Oh, ok, It seems that the official **Docker** image for **Zookeeper 3.6.1** is pretty buggy so I switched to `bitnami/zookeeper:3.6.1` and everything started working at last. Well, nevermind, the issue isn't related to **ClickHouse** so I'll just close it. | 2020-07-29T20:34:38 | 2020-07-29T20:34:38 | {} | NONE |
ClickHouse | ClickHouse | 666,009,955 | 13,055 | alexey-milovidov | No changes according to the existing perf tests. | 2020-07-30T00:55:08 | 2020-07-30T00:55:08 | {} | MEMBER |
ClickHouse | ClickHouse | 666,334,343 | 13,060 | den-crane | @veqryn BTW, if you add this condition `toYYYYMM(eventday) = 202007` in order to do a partition elimination because the table partitioned by YYYYMM(eventday) you actually don't need to do it. CH does partition pruning using `eventday >=` condition. | 2020-07-30T12:28:13 | 2020-07-30T12:28:13 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 666,372,915 | 12,748 | dsigitov | @den-crane, yes. I prepared an excerpt from the original table. The issue is reproducable there.
[clickhouse-sorting-bug-reference.tar.gz](https://github.com/ClickHouse/ClickHouse/files/5001530/clickhouse-sorting-bug-reference.tar.gz)
[log.txt](https://github.com/ClickHouse/ClickHouse/files/5001551/log.txt)
| 2020-07-30T13:43:38 | 2020-07-30T13:43:38 | {
"rocket": 1
} | NONE |
ClickHouse | ClickHouse | 664,035,390 | 12,749 | RonanMorgan | Hello,
Thank you for your answers,
After lots of tests I think I understand better the "way of thinking of clickhouse". I think my use case just not fit very well to the current features. (I need a transitional / precomputed table but views are more a precomputed request
My goals was to create a visitor data mod... | 2020-07-26T20:20:07 | 2020-07-29T09:30:29 | {} | NONE |
ClickHouse | ClickHouse | 665,589,812 | 12,755 | akuzm | I noticed there is another message just before the one you change:
```
<Debug> default.hits (SelectExecutor): Index `test_index` has dropped 1037 / 1084 granules.
<Debug> default.hits (SelectExecutor): Selected 1 parts by date, 1 parts by key, 1084 marks by primary key, 47 marks to read from 47 ranges
```
Is i... | 2020-07-29T10:42:57 | 2020-07-29T10:42:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,670,166 | 13,057 | filimonov | > Exception message is correct but it can be more detailed.
I think if we can't create user because of `access_management` is disabled, we should say smth like `Can't create user because access_management is disabled` the message `Not enough privileges` is very misleading.
| 2020-07-29T13:38:53 | 2020-07-29T13:39:19 | {
"+1": 19
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,819,169 | 13,086 | alexey-milovidov | This is significant change for changelog, please edit the description accordingly. | 2020-07-29T18:12:03 | 2020-07-29T18:12:03 | {} | MEMBER |
ClickHouse | ClickHouse | 665,250,336 | 12,548 | filimonov | Does is resolve #11576 and #10128 p.2 ? | 2020-07-28T20:02:23 | 2020-07-28T20:02:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,388,255 | 12,327 | ghost | @nikitamikhaylov
I learned a lot from this PR. Thank you very much. | 2020-07-29T02:07:43 | 2020-07-29T02:07:43 | {} | NONE |
ClickHouse | ClickHouse | 665,611,852 | 12,550 | Akazz | 0f5b089 emits a completely unrelated test failure in 00933_test_fix_extra_seek_on_compressed_cache | 2020-07-29T11:33:51 | 2020-07-29T11:33:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,891,775 | 13,081 | filimonov | Most probably also fixes https://github.com/ClickHouse/ClickHouse/issues/12406 | 2020-07-29T20:06:18 | 2020-07-29T20:06:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 666,052,646 | 13,099 | alexey-milovidov | This optimization is good on AMD Ryzen but bad on Intel Xeon E5-2650v2 (that is 7 years old) due to small cache size. | 2020-07-30T02:41:53 | 2020-07-30T02:41:53 | {} | MEMBER |
ClickHouse | ClickHouse | 665,250,957 | 10,128 | filimonov | https://github.com/ClickHouse/ClickHouse/issues/11576 | 2020-07-28T20:03:40 | 2020-07-28T20:03:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,382,903 | 9,278 | CurtizJ | We decided to set alias `polygon` to most performant (according to [paper](https://github.com/ClickHouse/clickhouse-presentations/blob/master/hse_2020/3rd/PolygonDict_full.pdf)) dictionary layout `polygon_index_cell` to avoid possible misconfigurations. Some users might choose the shortest and the simplest layout name.... | 2020-07-29T01:47:30 | 2020-07-29T01:47:30 | {} | MEMBER |
ClickHouse | ClickHouse | 665,495,168 | 13,057 | filimonov | I think the bug is 'misleading exception' | 2020-07-29T07:45:28 | 2020-07-29T07:45:28 | {
"+1": 11
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,630,382 | 13,057 | alexey-milovidov | Exception message is correct but it can be more detailed. | 2020-07-29T12:19:16 | 2020-07-29T12:19:16 | {} | MEMBER |
ClickHouse | ClickHouse | 665,331,592 | 12,769 | alexey-milovidov | @blinkov Could you please help with reading the output of Docs check? | 2020-07-28T22:42:51 | 2020-07-28T22:42:51 | {} | MEMBER |
ClickHouse | ClickHouse | 665,368,904 | 13,053 | alexey-milovidov | Docs check Ok. | 2020-07-29T00:55:37 | 2020-07-29T00:55:37 | {} | MEMBER |
ClickHouse | ClickHouse | 665,532,531 | 12,559 | winter7 | @den-crane Sorry friend, I have edited my comment. | 2020-07-29T08:54:48 | 2020-07-29T08:54:48 | {} | NONE |
ClickHouse | ClickHouse | 665,675,052 | 13,068 | alexey-milovidov | LGTM | 2020-07-29T13:48:02 | 2020-07-29T13:48:02 | {} | MEMBER |
ClickHouse | ClickHouse | 665,808,032 | 12,755 | bobrik | @akuzm it's similar in spirit, but somewhat different in the data it shows.
Index level log messages are cumulative per part, meaning that if you skip all granules in the first one, the second one has no work to do. Index level messages also operate on index sized granules rather than on marks. It's super handy if y... | 2020-07-29T17:49:14 | 2020-07-29T17:49:14 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,331,993 | 12,731 | alexey-milovidov | @akuzm @blinkov We need this :) | 2020-07-28T22:44:04 | 2020-07-28T22:44:04 | {} | MEMBER |
ClickHouse | ClickHouse | 665,397,724 | 11,913 | alexey-milovidov | @CurtizJ Do you want to take over this PR? | 2020-07-29T02:43:05 | 2020-07-29T02:43:22 | {} | MEMBER |
ClickHouse | ClickHouse | 665,497,844 | 10,464 | ryanmickler | For reference, you'll get this error if you exclude the `defaults`. see **https://github.com/ClickHouse/ClickHouse/pull/11628**
```
<http_handlers>
<defaults/> <!-- enable default HTTP handlers - otherwise prometheus exporter will not work -->
<rule>
<url>/custom_query</url>
... | 2020-07-29T07:50:33 | 2020-07-29T07:50:33 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 665,558,253 | 13,057 | Slach | hmm, ok, I understood,
I think exception with clarified message when we try to CREATE USER \ ROLE etc. about enabling access control will be good solution | 2020-07-29T09:41:28 | 2020-07-29T09:41:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,589,070 | 12,580 | CurtizJ | I found a bug, but it can appear only if query failed to connect to Redis and it throws an exception `Lost connection to Redis server`. After that all queries to dictionary fail with `DB::Exception: Assertion violation: _output in file`. Are you sure, that there was no failed queries beetween your runs? | 2020-07-29T10:41:25 | 2020-07-29T10:42:33 | {} | MEMBER |
ClickHouse | ClickHouse | 665,770,733 | 12,327 | nikitamikhaylov | > @nikitamikhaylov
> I learned a lot from this PR. Thank you very much.
@achimbab Look [here](https://github.com/ClickHouse/ClickHouse/pull/12992/files) how to write cmake files properly for future PR's. There are some fixes to current cmake lists, because we have lots of build types. | 2020-07-29T16:34:45 | 2020-07-29T16:34:45 | {} | MEMBER |
ClickHouse | ClickHouse | 665,916,733 | 12,642 | vitlibar | It seems the checks are failed because you're expecting a wrong exit code. The correct error code in that case is `493`, see [ErrorCodes.cpp](https://github.com/ClickHouse/ClickHouse/blob/cf505a92efc0a2a73d6e7c7a456f2b559b7a96bb/src/Common/ErrorCodes.cpp#L461). | 2020-07-29T20:44:03 | 2020-07-29T20:44:03 | {} | MEMBER |
ClickHouse | ClickHouse | 666,157,543 | 13,066 | ghost | @den-crane
Thank you for your answer. | 2020-07-30T06:37:25 | 2020-07-30T06:37:25 | {} | NONE |
ClickHouse | ClickHouse | 666,346,832 | 13,043 | filimonov | Superseded by #13105 | 2020-07-30T12:55:03 | 2020-07-30T12:55:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,360,545 | 10,851 | spihiker | 大家好。看到这个功能，真的很高兴。但从实际场景，我想提点建议，能否开发成table级别的引擎？因为大多生产实践，不会把整个库同步过来。只是一部分表。 | 2020-07-29T00:23:02 | 2020-07-29T00:23:02 | {} | NONE |
ClickHouse | ClickHouse | 665,503,388 | 13,059 | artbeglaryan | Can I manually edit metafiles before restart? | 2020-07-29T08:01:01 | 2020-07-29T08:01:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,817,919 | 13,086 | alexey-milovidov | The code looks Ok. | 2020-07-29T18:09:29 | 2020-07-29T18:09:29 | {} | MEMBER |
ClickHouse | ClickHouse | 665,847,178 | 13,086 | alexey-milovidov | I usually have 10s of tabs with bash opened and I would prefer shared history on every server.
So, when I click on random tab and press "Up", I will see full history.
You said that according to your experience, it's not convenient for clickhouse-client... Ok but I don't understand why. | 2020-07-29T19:06:50 | 2020-07-29T19:06:50 | {} | MEMBER |
ClickHouse | ClickHouse | 666,067,164 | 12,539 | spihiker | 已经升级到了20.3.13.127.今天再次出现由于select语句，出现CH服务闪退，服务挂掉的情况。请关注!
2020.07.30 10:56:24.089512 [ 4322 ] {} <Fatal> BaseDaemon: ########################################
2020.07.30 10:56:24.089599 [ 4323 ] {} <Fatal> BaseDaemon: ########################################
2020.07.30 10:56:24.089654 [ 4322 ] {} <Fatal> BaseDaemon: (... | 2020-07-30T03:16:54 | 2020-07-30T03:16:54 | {} | NONE |
ClickHouse | ClickHouse | 666,386,749 | 13,109 | alexey-milovidov | MergeTree tables are represented by a set of immutable data parts. Data parts are never modified, only created and deleted. INSERTs create new data parts. SELECTs take snapshot of required data parts and hold the snapshot during the query.
Snapshots are based on reference counting and snapshot acquire operation is ver... | 2020-07-30T14:07:40 | 2020-07-30T14:11:55 | {} | MEMBER |
ClickHouse | ClickHouse | 665,555,854 | 13,060 | veqryn | I don't have access to the individual clickhouse hosts, so I'll just run it once against local and it will give me a random one.
For the sake of brevity, I've edited the dates, as it gives the same problem.
Query with first of month:
```sql
select eventday, count(*) from mytable
where toYYYYMM(eventday) = 202007... | 2020-07-29T09:36:42 | 2020-07-29T09:36:42 | {} | NONE |
ClickHouse | ClickHouse | 665,558,493 | 13,060 | veqryn | thank you! | 2020-07-29T09:41:57 | 2020-07-29T09:41:57 | {} | NONE |
ClickHouse | ClickHouse | 665,937,977 | 12,177 | agelwarg | We just hit this today and it brought down 2 production nodes. We have a particular application path triggered by our users which creates temporary tables that, in aggregate, lead to an OOM on our Clickhouse nodes. We recognize that this is bad application code and we are investigating how to address it. However, th... | 2020-07-29T21:27:59 | 2020-07-29T21:27:59 | {} | NONE |
ClickHouse | ClickHouse | 666,254,563 | 12,508 | KochetovNicolai | 00960_live_view_watch_events_live
https://clickhouse-test-reports.s3.yandex.net/13075/d3bcf89ae427dcb2b9cd7def3ac7296de3def723/functional_stateless_tests_(address).html#fail1
```
2020-07-29 19:49:57 00960_live_view_watch_events_live: [ FAIL ] 120.86 sec. - return code 1
2020-07... | 2020-07-30T09:23:21 | 2020-07-30T09:23:21 | {} | MEMBER |
ClickHouse | ClickHouse | 666,389,706 | 13,109 | zergon321 | Thank you very much for the answer! | 2020-07-30T14:12:17 | 2020-07-30T14:12:17 | {} | NONE |
ClickHouse | ClickHouse | 665,550,697 | 13,060 | abyss7 | Thank you for your report!
Can you please provide the full results of both queries run over `mytable_local` - separately for 5-10 hosts of your cluster. | 2020-07-29T09:27:22 | 2020-07-29T09:27:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,979,344 | 12,642 | vzakaznikov | > It seems the checks are failed because you're expecting a wrong exit code. The correct error code in that case is `493`, see [ErrorCodes.cpp](https://github.com/ClickHouse/ClickHouse/blob/cf505a92efc0a2a73d6e7c7a456f2b559b7a96bb/src/Common/ErrorCodes.cpp#L461).
Thanks! I will take a look and fix failing tests. | 2020-07-29T23:20:17 | 2020-07-29T23:20:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 666,052,451 | 13,066 | den-crane | It's renamed after each CH upgrade which changes table's schema | 2020-07-30T02:41:23 | 2020-07-30T02:41:35 | {
"+1": 4
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 666,197,769 | 13,086 | azat | Looks like `pr-not-for-changelog` label should be removed? | 2020-07-30T07:45:45 | 2020-07-30T07:45:45 | {} | MEMBER |
ClickHouse | ClickHouse | 665,331,270 | 12,932 | alexey-milovidov | Not important. | 2020-07-28T22:41:58 | 2020-07-28T22:41:58 | {} | MEMBER |
ClickHouse | ClickHouse | 665,404,027 | 4,910 | ShonT | We too have a use case where we need null values in a dictionary and is stopping us from migrating our data to CH. Any update on when we could use this? | 2020-07-29T03:06:36 | 2020-07-29T03:06:36 | {} | NONE |
ClickHouse | ClickHouse | 665,462,356 | 12,769 | blinkov | @tetafro could you try to push any commit (empty is ok) to this branch to trigger the CI again? | 2020-07-29T06:32:42 | 2020-07-29T06:32:42 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,495,844 | 13,059 | filimonov | https://github.com/ClickHouse/ClickHouse/issues/10006
Workaround: use dictGetString / dictGetUInt etc | 2020-07-29T07:46:44 | 2020-07-29T07:51:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,665,315 | 13,067 | alexey-milovidov | The behaviour is 100% correct and better implementation is impossible even theoretically. | 2020-07-29T13:30:01 | 2020-07-29T13:30:01 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 665,741,641 | 13,017 | alesapin | Tests OK. | 2020-07-29T15:42:41 | 2020-07-29T15:42:41 | {} | MEMBER |
ClickHouse | ClickHouse | 665,817,515 | 13,086 | alexey-milovidov | But when we save history from different clients, all histories should be saved.
(Bash and all readline-based tools that I know don't do that and it's a huge drawback). | 2020-07-29T18:08:37 | 2020-07-29T18:08:37 | {} | MEMBER |
ClickHouse | ClickHouse | 665,986,523 | 12,327 | ghost | @nikitamikhaylov
Thank you. I will refer to the link you mentioned. | 2020-07-29T23:45:56 | 2020-07-29T23:45:56 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 665,259,021 | 12,548 | alexey-milovidov | @filimonov Yes, you can close / check the checkbox. | 2020-07-28T20:18:24 | 2020-07-28T20:18:24 | {} | MEMBER |
ClickHouse | ClickHouse | 665,741,961 | 7,785 | alesapin | Done here https://github.com/ClickHouse/ClickHouse/pull/13017 | 2020-07-29T15:43:16 | 2020-07-29T15:43:16 | {} | MEMBER |
ClickHouse | ClickHouse | 665,823,915 | 13,017 | nvartolomei | Happy this made to master!
Why return relative path instead of absolute? If multiple disks/volumes are used the path is ambiguous if I understand correctly. | 2020-07-29T18:19:32 | 2020-07-29T18:19:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 666,024,616 | 13,098 | CurtizJ | I found issue with similar stacktrace in https://github.com/ClickHouse/ClickHouse/pull/6124#issuecomment-514415258. | 2020-07-30T01:33:51 | 2020-07-30T01:34:10 | {} | MEMBER |
ClickHouse | ClickHouse | 666,155,807 | 7,862 | LeeMouRen | groupBitmapAnd has the same problem, how to solve this？groupBitmapOr -> groupBitmapMerge , groupBitmapAnd -> ??? thx | 2020-07-30T06:34:10 | 2020-07-30T06:34:10 | {} | NONE |
ClickHouse | ClickHouse | 666,287,494 | 13,073 | nikitamikhaylov | I've noticed, that compilation time increased. I think it is because we use `ccache`, and in this PR we have to recompile many files, but there maybe another reason.
I removed only headers, that are not directly used in files. | 2020-07-30T10:35:05 | 2020-07-30T10:35:05 | {} | MEMBER |
ClickHouse | ClickHouse | 666,313,892 | 13,101 | filimonov | Does it happens with queries like in #12539 ?
```
SELECT c1,c2....FROM t1 LEFT JOIN mysql('host,','username','dbname'.....) t2 ON ....
``` | 2020-07-30T11:38:36 | 2020-07-30T11:38:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,489,871 | 10,076 | toby080508050805 | I found other simple query cannot be killed. The query is sent from remote server. It stock at clickhouse whole day until I restart server.
```
2020.07.28 10:58:16.060644 [ 390 ] {1271c3c8-8f46-4b01-85a7-4eee6d79a94b} <Debug> executeQuery: (from [::ffff:192.168.177.4]:39208, initial_query_id: da045f5e-07b6-404f-b69f-... | 2020-07-29T07:34:53 | 2020-07-29T07:34:53 | {} | NONE |
ClickHouse | ClickHouse | 665,558,028 | 13,060 | filimonov | That seems very similar to https://github.com/ClickHouse/ClickHouse/issues/12406
Setting higher priority to raise attention, as that reproduces, affects core functionality, and leads to silent & wrong query results. | 2020-07-29T09:41:00 | 2020-07-29T09:41:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,667,708 | 12,485 | alesapin | Closed because this functionality implemented in other PR (I don't know which one). | 2020-07-29T13:34:20 | 2020-07-29T13:34:20 | {} | MEMBER |
ClickHouse | ClickHouse | 665,851,328 | 13,086 | azat | >I usually have 10s of tabs with bash opened and I would prefer shared history on every server.
>So, when I click on random tab and press "Up", I will see full history.
You mean inter-server history?
>You said that according to your experience, it's not convenient for clickhouse-client... Ok but I don't understa... | 2020-07-29T19:15:36 | 2020-07-29T19:15:36 | {} | MEMBER |
ClickHouse | ClickHouse | 665,899,806 | 11,576 | filimonov | Fixed in #12548 | 2020-07-29T20:10:42 | 2020-07-29T20:10:42 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 666,023,465 | 13,098 | CurtizJ | The second signal was a `SIGSEGV`, because if we print value of `info` from `signalHandler (sig=<optimized out>, info=0x7f2e983cb5f0, context=<optimized out>) ` we will got
```
(gdb) x 0x7f2e983cb5f0
0x7f2e983cb5f0: 0x0000000b
```
`siginfo_t` stores number of signal as its firts member.
| 2020-07-30T01:31:53 | 2020-07-30T01:31:53 | {} | MEMBER |
ClickHouse | ClickHouse | 666,187,890 | 7,862 | LeeMouRen | > groupBitmapAnd has the same problem, how to solve this？groupBitmapOr -> groupBitmapMerge , groupBitmapAnd -> ??? thx
The subquery can solve this Q,but it's imperfect. Looking forward to a reply
SELECT groupBitmapOr(column) FROM (select column from table_distribution where *)； | 2020-07-30T07:28:21 | 2020-07-30T07:28:21 | {} | NONE |
ClickHouse | ClickHouse | 666,392,333 | 13,105 | alexey-milovidov | What kind of build?, it's unclear. | 2020-07-30T14:16:30 | 2020-07-30T14:16:30 | {} | MEMBER |
ClickHouse | ClickHouse | 665,275,855 | 12,642 | vzakaznikov | @vitlibar, I've added the Markdown source for the [requirements](https://github.com/ClickHouse/ClickHouse/blob/a03d1d138126548883f5acc636abf2a21e17d94f/tests/testflows/rbac/requirements/requirements.md) and I've updated the [README.md](https://github.com/ClickHouse/ClickHouse/blob/a03d1d138126548883f5acc636abf2a21e17d9... | 2020-07-28T20:50:38 | 2020-07-28T20:50:38 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,427,596 | 9,637 | zhanglistar | > For now, you can't avoid that.
>
> It is a temporary workaround for a more serious problem, which exists in ClickHouse currently and can't be fixed easily without few refactorizations (in progress). It will be fixed in further versions
>
> In depends on those tasks:
>
> 1. Non-blocking ALTERs in MergeTree #8... | 2020-07-29T04:44:38 | 2020-07-29T04:44:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,566,788 | 13,057 | Slach | @vitlibar @abyss7 @filimonov
Also, misleading error messages for default user I get when run the following SQL queries:
```sql
SHOW USERS;
SHOW ROLES;
SELECT * FROM system.quota_usage;
``` | 2020-07-29T09:58:00 | 2020-07-29T09:58:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 665,570,271 | 12,508 | qoega | 01305_replica_create_drop_zookeeper https://clickhouse-test-reports.s3.yandex.net/0/795c09fdbbbe2c1f156d8b0c795cd9992297cb20/functional_stateless_tests_(unbundled).html#fail1
```
2020-07-28 00:50:45 [80aefae76e85] 2020.07.28 00:50:44.694754 [ 428 ] {255db9d1-b454-490b-a8f2-545af840b22d} <Error> executeQuery: Code: 99... | 2020-07-29T10:04:52 | 2020-07-29T10:04:52 | {} | MEMBER |
ClickHouse | ClickHouse | 665,759,281 | 11,422 | vkingnew | syntax highligting this functions how use？ | 2020-07-29T16:13:43 | 2020-07-29T16:13:43 | {} | NONE |
ClickHouse | ClickHouse | 665,996,394 | 12,589 | bobrik | @KochetovNicolai are you querying across multiple parts? I can see the log lines when I only hit one part:
```
MinMax index condition: (negate(toUnixTimestamp(column 0)) in [-1594515600, -1594515600]), unknown, unknown, and, and, (negate(toUnixTimestamp(column 0)) in [-1594515600, -1594515600]), and
Used optimized... | 2020-07-30T00:16:30 | 2020-07-30T00:16:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 666,020,299 | 12,762 | excitoon | @Akazz у меня тут похоже получилась циклическая зависимость, как бы её лучше разрулить? | 2020-07-30T01:23:05 | 2020-07-30T01:23:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 666,068,200 | 8,889 | spihiker | 已经升级到了20.3.13.127.今天再次出现由于select语句，出现CH服务闪退，服务挂掉的情况。请关注!
2020.07.30 10:56:24.089512 [ 4322 ] {} BaseDaemon: ########################################
2020.07.30 10:56:24.089599 [ 4323 ] {} BaseDaemon: ########################################
2020.07.30 10:56:24.089654 [ 4322 ] {} BaseDaemon: (version 20.3.13.127 (o... | 2020-07-30T03:19:01 | 2020-07-30T03:19:01 | {} | NONE |
ClickHouse | ClickHouse | 666,255,388 | 13,016 | KochetovNicolai | Performance — 6 faster, 1 slower | 2020-07-30T09:24:57 | 2020-07-30T09:24:57 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 665,391,349 | 13,054 | alexey-milovidov | Testflows Ok. | 2020-07-29T02:19:14 | 2020-07-29T02:19:14 | {} | MEMBER |
ClickHouse | ClickHouse | 665,978,619 | 12,177 | myrrc | > We just hit this today and it brought down 2 production nodes. We have a particular application path triggered by our users which creates temporary tables that, in aggregate, lead to an OOM on our Clickhouse nodes. We recognize that this is bad application code and we are investigating how to address it. However, thi... | 2020-07-29T23:17:43 | 2020-07-29T23:19:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 666,335,440 | 12,748 | den-crane | @dsigitov do you have reproducible example? Or maybe you can share you table data? backup copy of dataparts | 2020-07-30T12:30:40 | 2020-07-30T12:30:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 666,340,577 | 13,100 | alexey-milovidov | Don't wait for performance test. | 2020-07-30T12:41:52 | 2020-07-30T12:41:52 | {} | MEMBER |
ClickHouse | ClickHouse | 665,256,432 | 13,043 | alexey-milovidov | @Enmk You can create PR as a draft, then we will not merge it even if it's all green and the code is perfect. | 2020-07-28T20:15:18 | 2020-07-28T20:15:18 | {} | MEMBER |
ClickHouse | ClickHouse | 665,525,955 | 13,059 | alesapin | > Can I manually edit metafiles before restart?
Yes, it's absolutely OK if your table is not Replicated. In the case of replicated table, you have to fix metadata in zookeeper as well. | 2020-07-29T08:42:51 | 2020-07-29T08:42:51 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 665,553,653 | 12,749 | RonanMorgan | For those interested in : I discovered that using
OPTIMIZE TABLE all_visitor final
+
[finalizeAggregation](https://clickhouse.tech/docs/en/sql-reference/functions/other-functions/#function-finalizeaggregation)
instead of groupArrayMerge could do the trick : you don't need to made a second GROUP BY when select... | 2020-07-29T09:32:32 | 2020-07-29T12:26:25 | {} | NONE |
ClickHouse | ClickHouse | 665,664,687 | 13,067 | alexey-milovidov | Float32 contains 24 mantissa bits and cannot represent the number 79083738 exactly.
https://en.wikipedia.org/wiki/Single-precision_floating-point_format | 2020-07-29T13:28:53 | 2020-07-29T13:28:53 | {} | MEMBER |
ClickHouse | ClickHouse | 665,770,251 | 11,422 | alexey-milovidov | @vkingnew
Do I understand your question correctly?
Q: How to enable syntax highlighting in ClickHouse?
A: It is available and enabled since ClickHouse version 20.5.
When you install new ClickHouse version, `clickhouse-client` will have syntax highlighting. | 2020-07-29T16:33:49 | 2020-07-29T16:35:50 | {} | MEMBER |
ClickHouse | ClickHouse | 665,851,819 | 13,086 | alexey-milovidov | > Looks like you don't need/want this, then how about add an option for this?
No, option is too complex, no one will use it.
Let's proceed with your variant. | 2020-07-29T19:16:41 | 2020-07-29T19:16:41 | {} | MEMBER |
ClickHouse | ClickHouse | 666,311,490 | 13,101 | filimonov | 1) Does the problem reproduce every time with that query?
2) Can you share the query and schema samples?
-- machine translation to chineese: --
1）问题是否每次都随该查询重现？
2）您可以共享查询和架构示例吗？ | 2020-07-30T11:32:25 | 2020-07-30T11:32:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 689,947,313 | 14,525 | alexey-milovidov | Integration test:
`test_adaptive_granularity/test.py::test_version_update_two_nodes`
- flaky test @alesapin | 2020-09-10T03:04:39 | 2020-09-10T03:04:39 | {} | MEMBER |
ClickHouse | ClickHouse | 690,353,012 | 14,687 | den-crane | >
> Does this mean I cannot use clickhouse-copier to migrate data between two clusters with different time zone?
You can use clickhouse-copier. You need to configure the client to see results in desired timezone.
example :
**TZ=Asia/Shanghai** clickhouse-client -q 'select toDateTime(1599746203)' **--use_clien... | 2020-09-10T15:06:31 | 2020-09-10T15:06:54 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.