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 | 843,635,864 | 24,264 | alexey-milovidov | No idea why do we need jQuery. | 2021-05-18T23:36:16 | 2021-05-18T23:36:16 | {} | MEMBER |
ClickHouse | ClickHouse | 843,635,530 | 24,264 | alexey-milovidov | https://popper.js.org/
\- the website looks fishy - too shiny, like it throws syrup on your face... Most likely another pile of garbage. | 2021-05-18T23:35:26 | 2021-05-18T23:38:21 | {} | MEMBER |
ClickHouse | ClickHouse | 843,935,536 | 24,115 | awesomeleo | Thanks a lot. | 2021-05-19T09:45:06 | 2021-05-19T09:45:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,953,733 | 24,157 | alesapin | @Mergifyio update | 2021-05-19T10:08:29 | 2021-05-19T10:08:29 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 844,010,309 | 24,243 | alesapin | `Child process was terminated by signal 9 (KILL).` | 2021-05-19T11:26:06 | 2021-05-19T11:26:06 | {} | MEMBER |
ClickHouse | ClickHouse | 844,013,526 | 24,171 | alesapin | ```
2021.05.16 21:55:40.890568 [ 2830 ] {d39d5b0a-e07b-43cc-afd0-65aba3c1be84} <Debug> executeQuery: (from [::1]:35980, using production parser) SELECT * FROM ( SELECT number AS x FROM numbers(10) ) AS t1 ALL FULL JOIN storage_join_race USING (x) FORMAT Null;
2021.05.16 21:55:40.891531 [ 2830 ] {d39d5b0a-e07b-43cc-af... | 2021-05-19T11:30:23 | 2021-05-19T11:31:46 | {} | MEMBER |
ClickHouse | ClickHouse | 844,150,180 | 21,893 | rlazimi-dev | Hi @traceon, I am able to authenticate with my LDAP user, with anonymous binding forbidden, using the configuration tips you provided - thanks! | 2021-05-19T14:18:06 | 2021-05-19T14:18:06 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 844,173,853 | 24,059 | tonyhb | > I'm curious why there are so many likes?) We still have to implement data import from ZooKeeper. But it should be OK for fresh installations if you don't use zookeeper auth (quite weird thing).
Not sure about everyone else, but I've been watching your progress closely since https://github.com/ClickHouse/ClickHouse... | 2021-05-19T14:43:29 | 2021-05-19T14:43:29 | {} | NONE |
ClickHouse | ClickHouse | 844,206,119 | 24,294 | siviae | I am sorry for "unfinished code" label, "feature request" is more suitable, I guess. | 2021-05-19T15:17:54 | 2021-05-19T15:17:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,207,829 | 24,102 | larry-cdn77 | After some more investigation I realised that my buffer table settings were leading to well over 1 insert a second (min/max criteria are per layer rather than whole table)
I have adjusted settings to the point where the StorageBufferFlush event changes no master than 1-2 a second, will see if more aggregating tables... | 2021-05-19T15:19:49 | 2021-05-19T15:19:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,931,845 | 24,102 | larry-cdn77 | Good news is that making buffer table flush less frequently seemed to reduce the replication pressure and I was able to add aggregating tables until I had 14 (all triggered from one MergeTree table as shown earlier)
Bad news is that after some hours I found a few nodes with inflated replication queues (and two of th... | 2021-05-19T09:40:23 | 2021-05-19T15:47:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,631,343 | 24,264 | alexey-milovidov | The bug was introduced in #11728. | 2021-05-18T23:24:27 | 2021-05-18T23:24:27 | {} | MEMBER |
ClickHouse | ClickHouse | 843,643,530 | 23,504 | alexey-milovidov | It's better to avoid implementation of this feature. | 2021-05-18T23:58:50 | 2021-05-18T23:58:50 | {} | MEMBER |
ClickHouse | ClickHouse | 843,705,064 | 6,613 | MaxWk | i want try this | 2021-05-19T02:55:50 | 2021-05-19T02:55:50 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,848,742 | 24,205 | alesapin | 
:( | 2021-05-19T08:00:13 | 2021-05-19T08:00:13 | {} | MEMBER |
ClickHouse | ClickHouse | 843,948,380 | 19,649 | vdimir | I removed estimation of number of objects and `max_download_threads` always used for every file not depending on `num_streams`.
Performance reevaluated with same script, almost same, case with many files and many threads per file suffers more, but it's expectable.
<details>
<summary>Details</summary>
```
>>>... | 2021-05-19T10:01:40 | 2021-05-19T10:01:40 | {} | MEMBER |
ClickHouse | ClickHouse | 844,018,280 | 22,286 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
_Hey, I reacted but my real name is @Mergifyio_ | 2021-05-19T11:37:33 | 2021-05-19T11:37:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,111,903 | 24,275 | ucasfl | > @ucasFL Let's tune settings like in MergeTree https://github.com/ClickHouse/ClickHouse/blob/master/src/Storages/MergeTree/MergeTreeDataMergerMutator.cpp#L1184-L1186
OK. | 2021-05-19T13:34:28 | 2021-05-19T13:34:28 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 841,207,451 | 24,102 | larry-cdn77 | The same setup described above is now running version 21.3.10
At a glance the replication pressure remained high and I continue to get 'too many parts'
The Altinity query shows that replication queue mostly consists of actions with the following characteristics:
- fetches
- postponed up to 100-200 times
- po... | 2021-05-14T12:17:35 | 2021-05-19T15:16:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,383,517 | 24,277 | filimonov | Build of packages for platforms other than amd64 is not supported. | 2021-05-19T18:59:30 | 2021-05-19T18:59:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,593,255 | 24,194 | alexey-milovidov | We can add functional tests instead of integration tests - it's easy, because sqlite database is a single file. | 2021-05-18T21:58:03 | 2021-05-18T21:58:17 | {} | MEMBER |
ClickHouse | ClickHouse | 843,756,929 | 23,900 | chickony | @alexey-milovidov
Why have you marked it as Invalid?
I have just verified again now on latest docker and the following flow still fails
1-
`CREATE TABLE hits_100m_obfuscated (`WatchID` UInt64, `JavaEnable` UInt8, `Title` String, `GoodEvent` Int16, `EventTime` DateTime, `EventDate` Date, `CounterID` UInt32, `Clie... | 2021-05-19T05:27:05 | 2021-05-19T05:27:05 | {} | NONE |
ClickHouse | ClickHouse | 844,019,072 | 17,871 | alesapin | :( | 2021-05-19T11:38:34 | 2021-05-19T11:38:34 | {} | MEMBER |
ClickHouse | ClickHouse | 841,226,747 | 24,102 | larry-cdn77 | I thought I would add draw total replication queue size to demonstrate how readily I observe the symptoms. 12:20 in the chart below is when I go from 5 aggregations to 3.
 | 2021-05-14T12:56:54 | 2021-05-19T15:16:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,457,006 | 24,249 | alesapin |
??? | 2021-05-18T19:04:29 | 2021-05-18T19:04:29 | {} | MEMBER |
ClickHouse | ClickHouse | 843,690,068 | 24,218 | zhbdesign | > Please, show the exact entry in your config.xml file.
| 2021-05-19T02:09:53 | 2021-05-19T02:09:53 | {} | NONE |
ClickHouse | ClickHouse | 843,747,722 | 23,504 | filimonov | Well the problem is that for users which made a 'canonical setup' of Ubuntu / debian clickhouse can't work properly. And debugging the problem is not straight-forward.
We should either support that default configuration, either warn user (for example during server start) that we don't support that. Second option is... | 2021-05-19T05:02:02 | 2021-05-19T05:02:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,844,741 | 24,247 | alesapin | ttl_replicated also have to be fixed. | 2021-05-19T07:55:03 | 2021-05-19T07:55:03 | {} | MEMBER |
ClickHouse | ClickHouse | 843,929,645 | 24,059 | alesapin | `test_rename_column/test.py::test_rename_with_parallel_merges` yaaarrrr | 2021-05-19T09:37:22 | 2021-05-19T09:37:22 | {} | MEMBER |
ClickHouse | ClickHouse | 844,030,667 | 24,149 | nshah14285 | @kssenii
Any update on this?
Thanks in Advance. | 2021-05-19T11:55:44 | 2021-05-19T11:55:44 | {} | NONE |
ClickHouse | ClickHouse | 844,077,747 | 24,249 | alesapin | `01622_defaults_for_url_engine` -- absolutely random test. I'll merge. | 2021-05-19T12:54:33 | 2021-05-19T12:54:33 | {} | MEMBER |
ClickHouse | ClickHouse | 844,079,823 | 24,249 | alesapin | DecomposeFloat -- reported https://github.com/ClickHouse/ClickHouse/issues/24283 | 2021-05-19T12:57:04 | 2021-05-19T12:57:04 | {} | MEMBER |
ClickHouse | ClickHouse | 843,208,935 | 24,023 | alesapin | I'll retry the task. | 2021-05-18T14:14:45 | 2021-05-18T14:14:45 | {} | MEMBER |
ClickHouse | ClickHouse | 843,476,868 | 24,237 | UnamedRus | > Now I'm confused what's the expected behavior.
Hmm, expected behavior for me is:
for HAVING and ORDER BY if there is aggregate function around name use column, if there is no aggregate function use alias.
| 2021-05-18T19:29:37 | 2021-05-18T19:29:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,016,669 | 24,249 | alesapin | Only `Functional stateless tests (release, DatabaseReplicated)` is important | 2021-05-19T11:35:01 | 2021-05-19T11:35:01 | {} | MEMBER |
ClickHouse | ClickHouse | 843,287,666 | 24,246 | joschi | It looks like GCS doesn't support the [DeleteObjects](https://docs.aws.amazon.com/AmazonS3/latest/API/API_DeleteObjects.html ) request and possibly other batch APIs which is being used by ClickHouse:
[Google Cloud Storage Issue Tracker: Support deleteObjects API](https://issuetracker.google.com/issues/162653700)
... | 2021-05-18T15:51:23 | 2021-05-18T15:51:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,501,491 | 24,225 | UnamedRus | Seems reasonable, but it's a bit pity that you would need to shuffle values in array in order to do Cartesian product. | 2021-05-18T19:45:12 | 2021-05-18T19:45:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,635,768 | 24,264 | alexey-milovidov | Our documentation is using Bootstrap by mistake.
Usage of Bootstrap leads to unmaintainable CSS. | 2021-05-18T23:36:01 | 2021-05-18T23:36:01 | {} | MEMBER |
ClickHouse | ClickHouse | 843,644,958 | 24,145 | alexey-milovidov | `toRelativeSecondNum` is returning the number of seconds from the fixed but unspecified point in time.
To obtain the number of seconds from specified point in time, simply subtract two DateTime values:
`SELECT now() - toDateTime('2000-01-01 00:00:00')`
https://gh-api.clickhouse.tech/play?user=play#U0VMRUNUIG5vdy... | 2021-05-19T00:02:29 | 2021-05-19T00:02:29 | {} | MEMBER |
ClickHouse | ClickHouse | 843,652,163 | 6,613 | alexey-milovidov | This feature can be easily implemented, but I don't have it in plans for my team. | 2021-05-19T00:21:04 | 2021-05-19T00:21:04 | {} | MEMBER |
ClickHouse | ClickHouse | 844,133,207 | 24,275 | alesapin | @ucasFL What about `max_streams_to_max_threads_ratio`? | 2021-05-19T13:59:05 | 2021-05-19T13:59:05 | {} | MEMBER |
ClickHouse | ClickHouse | 844,199,199 | 24,092 | larry-cdn77 | I have ran into the same situation again and this time `STOP|START TTL MERGES` does not seem to help
Sifting through replicated log execution in MergeTree it would appear that `merges_with_ttl_counter` (as implemented in 5622e6daa6bf27e651a46e4482ffc9decd924ede) got out of sync
Is this at all likely, and is ther... | 2021-05-19T15:10:08 | 2021-05-19T15:48:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,221,541 | 24,175 | IgorLelyakov | Владимир огромное спасибо!
Установил SET join_algorithm = 'partial_merge';
Все прошло успешно!
С уважением,
Игорь | 2021-05-18T14:29:01 | 2021-05-18T14:29:01 | {} | NONE |
ClickHouse | ClickHouse | 843,427,947 | 24,194 | kssenii | Please, remove submodules update, which was added by accident -- looks like all of them were added in the last commit, then just revert it. And `sqlite` library should be added as submodule - you can delete `contrib/sqlite` and then add it again by executing in contrib dir `git submodule add ...`- that's all, then you'... | 2021-05-18T18:33:47 | 2021-05-18T18:36:36 | {} | MEMBER |
ClickHouse | ClickHouse | 843,523,908 | 24,201 | filimonov | > That's nice of you to help.
> I have the following pipe: `set -o pipefail ; tar -c metadata/default/iampartitioned.sql metadata/default.sql shadow/*/data/default/iampartitioned | /usr/bin/zstd -T5 | ...` and it fails with rc == 1 with hardlink changes. `--warning=no-file-changed` is no use here -- it only masks the ... | 2021-05-18T20:14:37 | 2021-05-18T20:22:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,642,688 | 23,311 | alexey-milovidov | Perf test has shown that uniqTheta is too slow on some use cases. It's interesting to figure out, why. | 2021-05-18T23:56:20 | 2021-05-18T23:56:20 | {} | MEMBER |
ClickHouse | ClickHouse | 843,953,900 | 24,157 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-05-19T10:08:42 | 2021-05-19T10:08:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,098,554 | 24,086 | den-crane | > Anyway, we are a bit disappointed with the clickhouse and the way it handles database parts.
> These servers were under have load caused by huge number of selects.
> At one moment when I summed threads from system.processes I come to the 4.5k for 100 queries.
> If database relays on merging, than merging should be... | 2021-05-19T13:18:29 | 2021-05-19T13:18:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,345,487 | 24,252 | KochetovNicolai | @azat I think I have fixed your cases. Do you have more? | 2021-05-19T18:08:01 | 2021-05-19T18:08:01 | {} | MEMBER |
ClickHouse | ClickHouse | 844,354,385 | 24,149 | kssenii | @nshah14285 I'll do in the next few days, it just requires more changes than it seems. | 2021-05-19T18:19:44 | 2021-05-19T18:19:44 | {} | MEMBER |
ClickHouse | ClickHouse | 843,651,676 | 24,263 | alexey-milovidov | The implementation can be similar to #6613. | 2021-05-19T00:19:42 | 2021-05-19T00:19:42 | {} | MEMBER |
ClickHouse | ClickHouse | 843,653,179 | 14,730 | alexey-milovidov | This feature is implemented and scheduled for release 21.6 in June. | 2021-05-19T00:23:49 | 2021-05-19T00:23:49 | {} | MEMBER |
ClickHouse | ClickHouse | 843,823,082 | 24,262 | UnamedRus | I checked and it does looks like it produces wrong result. sadly. | 2021-05-19T07:26:20 | 2021-05-19T07:26:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,017,120 | 24,171 | alesapin | unrelated failures, but looks dangerous. cc @vdimir | 2021-05-19T11:35:51 | 2021-05-19T11:36:00 | {} | MEMBER |
ClickHouse | ClickHouse | 843,643,848 | 23,504 | alexey-milovidov | Unless you will show that there is popular scenario where this feature will make significant benefits. | 2021-05-18T23:59:49 | 2021-05-18T23:59:49 | {} | MEMBER |
ClickHouse | ClickHouse | 843,691,623 | 24,218 | zhbdesign | > Please, show the exact entry in your config.xml file.
Is the change made in 21.4.1? Is the previous version the default < remote_ url_ allow_ hosts></remote_ url_ allow_ Hosts > means all allowed | 2021-05-19T02:14:37 | 2021-05-19T02:14:37 | {} | NONE |
ClickHouse | ClickHouse | 843,851,793 | 24,250 | vdimir | It calculates sum of length of all ranges (segments on numeric axis) without counting intersection twice, am I right?
Could you provide some use-cases? | 2021-05-19T08:04:13 | 2021-05-19T08:04:13 | {} | MEMBER |
ClickHouse | ClickHouse | 843,891,532 | 24,272 | rogerlzp | after another try, using the exact value ('202104') instead of expression ( toString(toYYYYMM(today()-30)) ), then the data is good. | 2021-05-19T08:49:58 | 2021-05-19T08:49:58 | {} | NONE |
ClickHouse | ClickHouse | 844,036,427 | 24,269 | alesapin | First of all, this version is out of support, so I recommend updating to the latest stable or LTS version. If you want to check your version for memory leaks you can build it with address sanitizer. Use CMake option `-DSANITIZE=address`. | 2021-05-19T12:03:03 | 2021-05-19T12:03:03 | {} | MEMBER |
ClickHouse | ClickHouse | 844,134,483 | 17,871 | alesapin | For some reason the first part was renamed to all_1_1_0
```
2021.05.19 14:02:20.868172 [ 53 ] {} <Trace> default.test_move (b8d35bd2-45fa-4e50-b8d3-5bd245fa2e50): Renaming temporary part detached/tmp_fetch_all_0_0_0 to all_1_1_0.
```
So
```Part all_0_0_0 not found locally (version 21.6.1.6879) (from 172.18.0.1:408... | 2021-05-19T14:00:29 | 2021-05-19T14:00:29 | {} | MEMBER |
ClickHouse | ClickHouse | 844,170,619 | 23,911 | ucasfl | > @ucasFL
> Thank you for catching this.
>
> Am I right that here we have the following 2 issues?
>
> 1. broken `getPosition()` method in ReadBufferFromS3
> 2. non-idempotent `next()` call for `SeekAvoidingReadBuffer -> ReadIndirectBufferFromS3 -> ReadBufferFromS3` chain?
>
> What if for 2nd issue we'll use ... | 2021-05-19T14:40:15 | 2021-05-19T14:40:15 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 843,980,228 | 322 | filimonov | GROUPING SETS: https://github.com/ClickHouse/ClickHouse/pull/24172 | 2021-05-19T10:44:43 | 2021-05-19T10:44:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,073,318 | 24,268 | MaxWk | looks like duplicate with https://github.com/ClickHouse/ClickHouse/issues/23901 | 2021-05-19T12:49:18 | 2021-05-19T12:49:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,214,000 | 24,201 | mallniya | > But do it really fail? Or it's only a matter or bad result code, while the archive is correct?
Yes, you're right, I can verify it only now on live experiments (as oppose to cron logs) without pipe.
>Exit code 1 for create command literally means "some files was changed", and it's ok to ignore it in that case
W... | 2021-05-19T15:26:43 | 2021-05-19T15:26:43 | {} | NONE |
ClickHouse | ClickHouse | 844,365,703 | 23,992 | sevirov | Internal documentation ticket: DOCSUP-9597. | 2021-05-19T18:34:52 | 2021-05-19T18:34:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,263,437 | 23,783 | KochetovNicolai | @azat I have made a different implementation in #24252 | 2021-05-18T15:19:23 | 2021-05-18T15:19:23 | {} | MEMBER |
ClickHouse | ClickHouse | 843,633,652 | 24,264 | alexey-milovidov | TODO: remove as many JavaScript libraries as possible. | 2021-05-18T23:30:03 | 2021-05-18T23:30:03 | {} | MEMBER |
ClickHouse | ClickHouse | 843,847,777 | 24,257 | lubo | Nope, the page doesn't contain any Atom/RSS link. | 2021-05-19T07:58:55 | 2021-05-19T07:58:55 | {} | NONE |
ClickHouse | ClickHouse | 844,086,013 | 24,275 | alesapin | @ucasFL Let's tune settings like in MergeTree https://github.com/ClickHouse/ClickHouse/blob/master/src/Storages/MergeTree/MergeTreeDataMergerMutator.cpp#L1184-L1186 | 2021-05-19T13:04:08 | 2021-05-19T13:04:08 | {} | MEMBER |
ClickHouse | ClickHouse | 844,365,878 | 24,124 | sevirov | Internal documentation ticket: DOCSUP-9597. | 2021-05-19T18:35:07 | 2021-05-19T18:35:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,202,915 | 13,195 | fzhedu | `groupUniqArray ` and `grougArray ` just accept only only expression, but `group_concat(x,x,x...)` could accept more expressions and order by. So how Clickhouse compeletly support `group_concat(x,x,x...)` of mysql? | 2021-05-18T14:06:58 | 2021-05-18T14:06:58 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 843,458,520 | 24,146 | azat | >Can we test, that settings are actually used and change behavior?
The test will not be simple (or will require some additional metrics), but it is possible.
Note, that I tested it manually and it works. | 2021-05-18T19:06:55 | 2021-05-18T19:06:55 | {} | MEMBER |
ClickHouse | ClickHouse | 843,539,084 | 24,201 | filimonov | Also you can try to use PIPESTATUS instead of pipefail
https://unix.stackexchange.com/a/14276
Or try to pass attributes the command line (not sure if it will work) or pass the files to the tar through stdin - see some ideas in
https://unix.stackexchange.com/a/391502 | 2021-05-18T20:32:37 | 2021-05-18T20:33:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,975,585 | 24,205 | KochetovNicolai | From `Integration tests (thread)`
```
2021-05-19 02:15:41,631 Totally finished tests 1328/1329
2021-05-19 02:15:41,631 Finally all tests done, going to compress test dir
```
Timeout was at 02:23:14 | 2021-05-19T10:38:20 | 2021-05-19T10:38:20 | {} | MEMBER |
ClickHouse | ClickHouse | 844,071,719 | 24,260 | alesapin | Yes it looks like bug and we even have a comment in the code: https://github.com/ClickHouse/ClickHouse/blob/master/src/Storages/MergeTree/MergeTreePartsMover.cpp#L255. So seems like your server was powered off exactly on this line. Maybe we can do atomic rename of two directories... | 2021-05-19T12:47:19 | 2021-05-19T12:47:19 | {} | MEMBER |
ClickHouse | ClickHouse | 843,273,972 | 24,157 | kssenii | @Mergifyio update | 2021-05-18T15:32:31 | 2021-05-18T15:32:31 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 843,933,747 | 24,249 | alesapin | Let's try one more time. `Connection loss`/`Session expired` errors look fixed, but tests in DatabaseReplicated fail for other reasons. | 2021-05-19T09:42:44 | 2021-05-19T09:42:44 | {} | MEMBER |
ClickHouse | ClickHouse | 844,016,945 | 24,224 | abyss7 | Probably, you may find an ALIAS column type more helpful for your scenario. | 2021-05-19T11:35:30 | 2021-05-19T11:35:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,033,650 | 24,272 | alesapin | Yes, it's known unexpected behavior. Delete mutations don't affect deduplication blocks. Maybe we have to implement such logic under the setting. | 2021-05-19T11:59:41 | 2021-05-19T11:59:41 | {} | MEMBER |
ClickHouse | ClickHouse | 843,453,949 | 24,237 | amosbird | I did it at first https://github.com/ClickHouse/ClickHouse/pull/22044/files#diff-858ec3896af7064c6c5ef2c2807a7da65d5f2e793d8fb00763b0fe54982f5ed9R141
Then I reverted them https://github.com/ClickHouse/ClickHouse/pull/24022
Now I'm confused what's the expected behavior. | 2021-05-18T19:00:15 | 2021-05-18T19:00:15 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 843,473,586 | 24,225 | den-crane | It's the correct / expected behavior. It looks unexpected only with a constant
In case of
```
select arrayJoin(col1),
dictGet(.... attr1, arrayJoin(col1)),
dictGet(.... attr2, arrayJoin(col1))
```
it allows to get expected results
because it's the same as
```
select arrayJoin(col1) as k,
... | 2021-05-18T19:27:26 | 2021-05-18T19:27:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,597,680 | 24,257 | alexey-milovidov | Maybe CVE list has subscription capabilities?
https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=ClickHouse | 2021-05-18T22:07:33 | 2021-05-18T22:07:33 | {} | MEMBER |
ClickHouse | ClickHouse | 843,634,480 | 24,264 | alexey-milovidov | Some of these libraries are trash: https://github.com/ClickHouse/ClickHouse/tree/master/website/js | 2021-05-18T23:32:22 | 2021-05-18T23:32:22 | {} | MEMBER |
ClickHouse | ClickHouse | 843,649,681 | 24,236 | alexey-milovidov | The current behaviour is Ok.
Throwing exception is also Ok.
> If Enums non intersecting, clickhouse can merge them safely.
We shall not do that, the Enums can be totally unrelated. E.g.
```
dbms Enum8('MySQL' = 1, 'Postgres' = 2),
beverage Enum8('Vodka' = 10, 'Samogon' = 11)
``` | 2021-05-19T00:14:28 | 2021-05-19T00:15:06 | {} | MEMBER |
ClickHouse | ClickHouse | 844,014,377 | 23,911 | Jokser | @ucasFL
Thank you for catching this.
Am I right that here we have the following 2 issues?
1) broken `getPosition()` method in ReadBufferFromS3
2) non-idempotent `next()` call for `SeekAvoidingReadBuffer -> ReadIndirectBufferFromS3 -> ReadBufferFromS3` chain?
What if for 2nd issue we'll use `nextIfAtEnd()` met... | 2021-05-19T11:31:35 | 2021-05-19T11:31:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,414,133 | 24,151 | azat | >[Functional stateless tests flaky check (address) — Timeout, fail: 67, passed: 2682](https://clickhouse-test-reports.s3.yandex.net/24151/147ae1749fb0d548a1283ffcd32428ec4e9fd542/functional_stateless_tests_flaky_check_(address).html#fail1)
`01414_push_predicate_when_contains_with_clause` - #24259 | 2021-05-18T18:12:01 | 2021-05-18T18:12:01 | {} | MEMBER |
ClickHouse | ClickHouse | 843,630,710 | 24,262 | den-crane | ```
CREATE TABLE xxxxxx_avg ( `key` UInt32, `x` AggregateFunction(avgWeighted, Float64, UInt64) ) ENGINE = AggregatingMergeTree ORDER BY key;
echo "MSwi3KLWEwAAAACsvh4iCg==" | base64 -d | clickhouse-client -mn --query "INSERT INTO xxxxxx_avg FORMAT CSV";
```
Need to move these parts / columns to anoth... | 2021-05-18T23:22:52 | 2021-05-18T23:43:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,980,720 | 20,274 | filimonov | I think it can be closed (superseded by https://github.com/ClickHouse/ClickHouse/pull/24172 ) | 2021-05-19T10:45:32 | 2021-05-19T10:45:32 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,018,107 | 22,286 | alesapin | @Mergify update | 2021-05-19T11:37:17 | 2021-05-19T11:37:17 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 843,634,174 | 24,264 | alexey-milovidov | FYI, the library "embedd" not only has a clunky name,
it is also not being maintained in the past three years. | 2021-05-18T23:31:28 | 2021-05-18T23:31:28 | {} | MEMBER |
ClickHouse | ClickHouse | 843,922,041 | 24,250 | ucasfl | > It calculates sum of length of all ranges (segments on numeric axis) without counting intersection twice, am I right?
Yes, you are right.
> Could you provide some use-cases?
The function is come from an our online requirement. Make a simple example, there are a lot of video play records, one record includes... | 2021-05-19T09:27:30 | 2021-05-19T09:27:30 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 843,981,498 | 19,426 | filimonov | wip pr https://github.com/ClickHouse/ClickHouse/pull/24172 | 2021-05-19T10:46:45 | 2021-05-19T10:46:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 844,007,996 | 24,023 | alesapin | `column_column_comparison | xml.etree.ElementTree.ParseError: no element found: line 39, column 0` | 2021-05-19T11:22:54 | 2021-05-19T11:22:54 | {} | MEMBER |
ClickHouse | ClickHouse | 844,076,459 | 24,059 | alesapin | I'm curious why there are so many likes?) We still have to implement data import from ZooKeeper. But it should be OK for fresh installations if you don't use zookeeper auth (quite weird thing). | 2021-05-19T12:53:08 | 2021-05-19T12:53:08 | {} | MEMBER |
ClickHouse | ClickHouse | 844,306,956 | 24,251 | amosbird | You can set `prefer_column_name_to_alias = 1`. | 2021-05-19T17:17:45 | 2021-05-19T17:17:45 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 844,359,380 | 16,947 | azat | Superseded by #24252 | 2021-05-19T18:26:37 | 2021-05-19T18:26:37 | {} | MEMBER |
ClickHouse | ClickHouse | 837,046,661 | 23,726 | akuzm | @mergifyio merge | 2021-05-10T17:53:35 | 2021-05-10T17:53:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 837,085,788 | 17,769 | alexey-milovidov | @Mergifyio update | 2021-05-10T18:18:05 | 2021-05-10T18:18:05 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 838,832,391 | 23,746 | bnaecker | Also, I'm not sure anything needs to be updated in `.gitmodules`, but I'd like to check. Of the patches that I've put up, only the modules referring to gRPC and Poco point to a specific branch. The rest pull from HEAD, if I'm not mistaken. The branches recorded in `.gitmodules` for those are `1.33.2` and `clickhouse`, ... | 2021-05-11T17:00:41 | 2021-05-11T17:00:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 838,833,607 | 24,018 | alexey-milovidov | Fixed in https://github.com/ClickHouse/ClickHouse/pull/23631. | 2021-05-11T17:01:35 | 2021-05-11T17:01:35 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.