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 | 591,970,113 | 2,013 | alexey-milovidov | It is the documented behaviour of `visitParam` functions. These functions are intended for specific use cases and you cannot say that these functions are any good for JSON (they are fast though).
For JSON processing, we have `JSONExtractRaw`, `JSONExtractString` - those are true JSON processing functions :) | 2020-02-27T13:27:50 | 2020-02-27T13:27:50 | {} | MEMBER |
ClickHouse | ClickHouse | 592,203,948 | 9,421 | alexey-milovidov | It is true that `data.table` was created on old ClickHouse version?
Then the reason of this exception is that the new table is created with new index granularity settings.
You can create a new table with the same index granularity as the old table.
@alesapin will help with example. | 2020-02-27T22:13:12 | 2020-02-27T22:13:12 | {} | MEMBER |
ClickHouse | ClickHouse | 592,215,156 | 9,079 | yurriy | It seems that the failures in the last check are not caused by this patch | 2020-02-27T22:41:48 | 2020-02-27T22:41:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 592,497,865 | 9,421 | alesapin | > Got it. Will try recreate dst table with new format.
>
> But why I have error "Inconsistent granularity" when both tables have same granularity?
> `MergeTree(field1, (field1, field2, field3), 8192)`
Bugfix for previous segfault bug https://github.com/ClickHouse/ClickHouse/pull/9435. | 2020-02-28T12:46:32 | 2020-02-28T12:46:32 | {} | MEMBER |
ClickHouse | ClickHouse | 592,512,569 | 5,131 | Deninc | This should be fixed, or at least give some warnings! | 2020-02-28T13:28:48 | 2020-02-28T13:28:48 | {} | NONE |
ClickHouse | ClickHouse | 591,366,856 | 9,378 | filimonov | Perfect report, thanks.
I will try to reproduce the problem, and check what's going on.
We may expect duplicates, but not holes in that usecase. | 2020-02-26T10:57:55 | 2020-02-26T10:57:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,411,953 | 9,378 | bgranvea | The test case here is simplified because on my real platform, I have several MV behind the same Kafka Engine and each MV inserts in a distributed table on top of a replicated table. I don't know if it makes a difference on how to detect that insertion failed or was partial (only some of the MV). | 2020-02-26T12:55:59 | 2020-02-26T12:55:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,830,019 | 9,337 | MikeZaharov | It's ok for us to use 2 arrays if this approach guarantees the order.
I mean we'll be able to get header's value somehow like this:
`arrayElement(header_values, indexOf(header_names, 'header_name')) as header_name` | 2020-02-27T07:46:26 | 2020-02-27T07:46:26 | {} | NONE |
ClickHouse | ClickHouse | 591,869,435 | 9,367 | ghost | @alexey-milovidov is it normal that the CI is not completely green? Do tests fail randomly? | 2020-02-27T09:26:31 | 2020-02-27T09:26:31 | {} | NONE |
ClickHouse | ClickHouse | 591,880,308 | 9,409 | clemrodriguez | @alexey-milovidov
Hi,
I opened this PR as a clone of #7274, which was polluted. Nevertheless, this one seems to still have commits that come from my fork master branch (merges&rebased). I hope it won't prevent its analysis.
Best regards. | 2020-02-27T09:50:45 | 2020-02-27T09:50:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 592,201,952 | 9,422 | alexey-milovidov | The test (along with the feature) is not in release.
1. Open the file on GitHub: https://github.com/ClickHouse/ClickHouse/blob/master/dbms/tests/queries/0_stateless/00975_move_partition_merge_tree.sql
2. Click on commit sha or commit message: https://github.com/ClickHouse/ClickHouse/commit/acaa9cdbaddbf3b1afd18c75a... | 2020-02-27T22:08:40 | 2020-02-27T22:08:55 | {} | MEMBER |
ClickHouse | ClickHouse | 592,255,538 | 9,379 | alexey-milovidov | One build has failed due to internal CI issue.
> Functional stateless tests (debug)
- no space left on device. | 2020-02-28T00:57:07 | 2020-02-28T00:57:07 | {} | MEMBER |
ClickHouse | ClickHouse | 591,415,215 | 9,379 | nikitamikhaylov | Test will be added. | 2020-02-26T13:04:03 | 2020-02-26T13:04:03 | {} | MEMBER |
ClickHouse | ClickHouse | 592,045,516 | 9,405 | nikitamikhaylov | Are you sure that your queries are really concurrent? If they are rather quick, they will be executed successively. Otherwise the exception will be thrown
https://github.com/ClickHouse/ClickHouse/blob/6e37d765e678edcb7e1813dc463dc2b690ee0854/dbms/src/Interpreters/ProcessList.cpp#L116 | 2020-02-27T16:18:37 | 2020-02-27T16:18:37 | {} | MEMBER |
ClickHouse | ClickHouse | 592,380,113 | 9,412 | alexey-milovidov | > Integration tests (release) Pending — Started
> Stress test (address) Pending — Started
Did not finished in about 10 hours. | 2020-02-28T07:11:39 | 2020-02-28T07:11:39 | {} | MEMBER |
ClickHouse | ClickHouse | 592,428,318 | 9,421 | alesapin | Yes, just use setting in storage definition:
```
CREATE TABLE ...
...
ENGINE = MergeTree() ... SETTINGS index_granularity_bytes = 0.
```
Or you can enable mixed granularity for table:
```
CREATE TABLE ...
...
ENGINE = MergeTree() ... SETTINGS enable_mixed_granularity_parts = 1.
```
Also you can setup last s... | 2020-02-28T09:29:50 | 2020-02-28T09:29:50 | {} | MEMBER |
ClickHouse | ClickHouse | 592,492,496 | 9,421 | alesapin | Oh, this is another bug... you cannot modify settings for tables with old format. You should create new table with new format. Old-formatted tables doesn't support settings at all. | 2020-02-28T12:29:43 | 2020-02-28T12:29:43 | {} | MEMBER |
ClickHouse | ClickHouse | 591,943,619 | 2,013 | alexeydubinin | Hi. I think you should reopen issue, because isValidJSON function return 1(valid) for both json's (with spaced and without). | 2020-02-27T12:24:19 | 2020-02-27T12:24:19 | {} | NONE |
ClickHouse | ClickHouse | 591,944,099 | 9,413 | nikitamikhaylov | Test will be added soon. | 2020-02-27T12:25:34 | 2020-02-27T12:25:34 | {} | MEMBER |
ClickHouse | ClickHouse | 592,031,067 | 8,636 | songjiao | we meet the same problem
> 2020.02.27 23:04:36.749619 [ 1338 ] {} <Fatal> BaseDaemon:
2020.02.27 23:04:36.749734 [ 1338 ] {} <Fatal> BaseDaemon: (version 19.17.6.36 (official build)) (from thread 491) Received signal Segmentation fault (11).
2020.02.27 23:04:36.749846 [ 1338 ] {} <Fatal> BaseDaemon: Address: 0x10... | 2020-02-27T15:51:44 | 2020-02-27T15:51:44 | {} | NONE |
ClickHouse | ClickHouse | 592,244,577 | 8,782 | alexey-milovidov | Let's wait for performance tests. | 2020-02-28T00:17:30 | 2020-02-28T00:17:30 | {} | MEMBER |
ClickHouse | ClickHouse | 592,456,936 | 8,282 | ifoo | Not the same query, but also a failing alter with the same error. Attached is the log+stack trace:
```
2020.02.28 11:27:57.271312 [ 43 ] {} <Information> siem.normalized_local: Became leader
2020.02.28 11:28:18.652428 [ 59 ] {0c9d9acf-d543-4700-9b80-2981492695dc} <Information> executeQuery: Read 32 rows, 4.08 KiB ... | 2020-02-28T10:43:10 | 2020-02-28T10:43:10 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 592,468,175 | 9,433 | alexey-milovidov | > which gets picked up by librdkafka
- rdkafka is build without flag that enables C11 threads;
- it implements its own methods for C11 threads with identical names;
- but if `-lpthread` is before `-lrdkafka`, linker is leaving the dependency on C11 symbols from pthread (?) | 2020-02-28T11:14:01 | 2020-02-28T11:14:01 | {} | MEMBER |
ClickHouse | ClickHouse | 591,463,021 | 9,379 | nikitamikhaylov | No backport needed, because in previous versions we don't have "async update" feature. | 2020-02-26T14:47:04 | 2020-02-26T14:47:04 | {} | MEMBER |
ClickHouse | ClickHouse | 591,744,494 | 3,617 | gubinjie | @chenxing-xc Hello, when we use clickhouse to read hdfs files, the following exception often occurs，Can you tell me what problems may cause this exception, and what methods can be used to resolve this exception，thank you！
-----------------------
executeQuery: Code: 210, e.displayText() = DB::Exception: Fail to read ... | 2020-02-27T02:21:27 | 2020-02-27T02:21:27 | {} | NONE |
ClickHouse | ClickHouse | 592,223,453 | 9,079 | alexey-milovidov | > test_polymorphic_parts/test.py::test_polymorphic_parts_diff_versions | ERROR
Yes, it's unrelated.
> test_storage_mysql/test.py::test_table_function | FAILED
Need to dig into it...
| 2020-02-27T23:06:56 | 2020-02-27T23:06:56 | {} | MEMBER |
ClickHouse | ClickHouse | 591,487,304 | 9,378 | filimonov | 1) When several MV are attached then inserts are not atomic. One table may get data, the other one - failed.
2) The expected behavior in that case: then ANY of MV was failed, the whole block should be inserted once again (so if 9 of 10 MV was successful the first time, they will have that block duplicated, and the ten... | 2020-02-26T15:33:26 | 2020-02-26T15:33:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,506,949 | 9,256 | emyr666 | also one that returns the filename part.
| 2020-02-26T16:09:01 | 2020-02-26T16:09:01 | {} | NONE |
ClickHouse | ClickHouse | 592,037,642 | 2,096 | vigneshrajarr | @filimonov
> You have chaged minArray to just first element of result array for step_2 time. That can give wrong result, because clickhouse can find some event which happened after step3, so step3 will not be foubd
I am very happy you mentioned this. We are using Postgres for analytics and analysing CH now.
In Po... | 2020-02-27T16:03:32 | 2020-02-27T16:03:32 | {} | NONE |
ClickHouse | ClickHouse | 592,138,520 | 8,636 | amosbird | There should be a query string in the trace log before the crash. | 2020-02-27T19:34:14 | 2020-02-27T19:34:14 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 592,440,200 | 9,412 | alesapin | I'll cherry-pick it to 19.17 and 19.16 manually. | 2020-02-28T10:00:32 | 2020-02-28T10:00:32 | {} | MEMBER |
ClickHouse | ClickHouse | 592,483,941 | 8,282 | johnatannvmd | Look like this is expected behavior for DDL operations. From changelogs "Clickhouse Release 1.1.54276, 2017-08-16":
> Multiple improvements and fixes for clustered DDL queries. Of particular interest is the new setting distributed_ddl_task_timeout, which limits the time to wait for a response from the servers in the... | 2020-02-28T12:02:04 | 2020-02-28T12:03:06 | {} | NONE |
ClickHouse | ClickHouse | 591,657,343 | 7,795 | alexey-milovidov | @blinkov This PR is significant. | 2020-02-26T21:38:18 | 2020-02-26T21:38:18 | {} | MEMBER |
ClickHouse | ClickHouse | 591,992,417 | 9,382 | CurtizJ | Test added in #9412. | 2020-02-27T14:13:52 | 2020-02-27T14:13:52 | {} | MEMBER |
ClickHouse | ClickHouse | 592,223,933 | 9,079 | alexey-milovidov | Yes, it's known issue that happened to appear on master. | 2020-02-27T23:08:28 | 2020-02-27T23:08:28 | {} | MEMBER |
ClickHouse | ClickHouse | 592,243,657 | 8,782 | alexey-milovidov | Maybe because of awful results of the performance test: https://clickhouse-test-reports.s3.yandex.net/8782/50be423165854c621b92f0b00099adbdb014fe02/performance_test/comparison_min_time_gcc_9.html
(that are for unknown reason) | 2020-02-28T00:15:21 | 2020-02-28T00:15:21 | {} | MEMBER |
ClickHouse | ClickHouse | 592,255,649 | 9,379 | alexey-milovidov | > Integration tests (asan)
- known issue. | 2020-02-28T00:57:32 | 2020-02-28T00:57:32 | {} | MEMBER |
ClickHouse | ClickHouse | 592,289,833 | 9,405 | 42997455 | Thanks @nikitamikhaylov , I may not quit understand the really concurrent, I guess it mean how many queries run in same time?
what I have set in config.xml is:
`
<yandex>
...
<max_concurrent_queries>100</max_concurrent_queries>
<max_concurrent_queries_for_user>1</max_concurrent_queries_for_user>
...
</yandex>... | 2020-02-28T03:11:21 | 2020-02-28T03:11:21 | {} | NONE |
ClickHouse | ClickHouse | 592,464,713 | 9,433 | amosbird | > symbols like cnd_broadcast
which gets picked up by librdkafka | 2020-02-28T11:04:04 | 2020-02-28T11:04:04 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 591,449,002 | 9,374 | mark-angelo | @den-crane Denis, this issue is blocking a major release and we would be grateful if it can be given priority. Thank you so much! | 2020-02-26T14:19:07 | 2020-02-26T14:19:07 | {} | NONE |
ClickHouse | ClickHouse | 591,557,523 | 9,384 | alexey-milovidov | > Functional stateless tests (debug) — fail: 59, passed: 221
- No space left on device. | 2020-02-26T17:53:59 | 2020-02-26T17:53:59 | {} | MEMBER |
ClickHouse | ClickHouse | 591,627,377 | 9,383 | excitoon | > It's unclear what bug was fixed. Looks like an assertion for the internal logic in code.
I've updated the comment. | 2020-02-26T20:29:21 | 2020-02-26T20:29:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 592,473,315 | 8,282 | johnatannvmd | Have the same issue. CH hangs for timeout and then return error with time out. After the client closes a connection, it turns out that the field successfully created on the cluster.
```
2020.02.28 11:15:33.223406 [ 57 ] {8fdbd9d4-99ef-454e-9922-c8ec99bd251f} <Error> executeQuery: Code: 48, e.displayText() = DB::Exc... | 2020-02-28T11:28:48 | 2020-02-28T11:39:12 | {} | NONE |
ClickHouse | ClickHouse | 592,490,464 | 9,427 | alesapin | `test_polymorphic_parts/test.py::test_polymorphic_parts_diff_versions` failure is known. | 2020-02-28T12:23:10 | 2020-02-28T12:23:10 | {} | MEMBER |
ClickHouse | ClickHouse | 591,937,541 | 8,956 | alexey-milovidov | The issue is fixed. No errors after
```
for i in {1..1000000}; do echo -n "$i " && curl -k https://localhost:8443; done
``` | 2020-02-27T12:07:28 | 2020-02-27T12:07:28 | {} | MEMBER |
ClickHouse | ClickHouse | 591,941,850 | 8,956 | alexey-milovidov | Cannot provide an automated test case.
It does not quickly reproduce even if I run
```
pv /dev/random > /dev/null
```
in parallel. | 2020-02-27T12:19:43 | 2020-02-27T12:22:40 | {} | MEMBER |
ClickHouse | ClickHouse | 592,084,797 | 9,276 | Tri0L | Logs during query `select now(), sum(parts_to_do) from system.mutations;`
https://pastebin.com/8dqQ5Y1N
Query results:
```
:) select now(), sum(parts_to_do) from system.mutations;
SELECT
now(),
sum(parts_to_do)
FROM system.mutations
Timeout exceeded while receiving data from server. Waited for ... | 2020-02-27T17:37:12 | 2020-02-27T17:37:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 592,223,979 | 9,079 | alexey-milovidov | Thank you! | 2020-02-27T23:08:36 | 2020-02-27T23:08:36 | {} | MEMBER |
ClickHouse | ClickHouse | 592,242,785 | 9,420 | alexey-milovidov | Let's check if they finish before morning. | 2020-02-28T00:12:43 | 2020-02-28T00:12:43 | {} | MEMBER |
ClickHouse | ClickHouse | 592,478,269 | 9,421 | Tri0L | Yes, this table created with old format.
But for now in `show create table` I see that index_granularity are the same
`MergeTree(field1, (field1, field2, field3), 8192)`. Do I need to change it anyway?
I tried ALTER, but have error:
```
:) ALTER TABLE tmp.test1 MODIFY SETTING enable_mixed_granularity_par... | 2020-02-28T11:44:20 | 2020-02-28T11:44:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,445,042 | 9,367 | alexey-milovidov | I have also faced this issue. It cannot happen in production, because this constructor is unused (it only happens in separate test programs). I tried to fix it some time ago but faced some difficulties. Need to review the fix. | 2020-02-26T14:10:51 | 2020-02-26T14:11:09 | {} | MEMBER |
ClickHouse | ClickHouse | 591,659,698 | 9,383 | alexey-milovidov | @excitoon Do you mean that it's not a LOGICAL_ERROR but a user related error? | 2020-02-26T21:43:46 | 2020-02-26T21:43:46 | {} | MEMBER |
ClickHouse | ClickHouse | 591,937,774 | 8,886 | alexey-milovidov | The issue is fixed in https://github.com/ClickHouse/ClickHouse/pull/8956 | 2020-02-27T12:08:10 | 2020-02-27T12:08:10 | {} | MEMBER |
ClickHouse | ClickHouse | 592,125,353 | 9,413 | nikitamikhaylov | I looked through our test and didn't find any test for CurrentMetrics. :tired_face: | 2020-02-27T19:06:27 | 2020-02-27T19:06:27 | {} | MEMBER |
ClickHouse | ClickHouse | 592,155,911 | 9,413 | alexey-milovidov | > I looked through our test and didn't find any test for CurrentMetrics.
Global values of CurrentMetrics were difficult to test, because we have to check it at the right time.
Now it's possible to use `system.metric_log`.
(I'm not sure the test will worth adding) | 2020-02-27T20:14:27 | 2020-02-27T20:14:27 | {} | MEMBER |
ClickHouse | ClickHouse | 591,347,105 | 8,505 | KochetovNicolai | It works now in master and was accidentally fixed after #7796. So I have just backported #7796 back to 19.17 and 20.1. And added a test for `MergeTree`. | 2020-02-26T10:12:46 | 2020-02-26T10:12:46 | {} | MEMBER |
ClickHouse | ClickHouse | 591,513,400 | 9,374 | den-crane | Storing tuple just making all worse. It's still two implicit columns and you just make CH to do
`tupleElement(tuple(x,y),1)` instead of `x`
```
create table X (x Int64, A Tuple(String,String)) Engine=MergeTree order by x;
insert into X values (1, tuple('a','b'));
ls -1 /var/lib/clickhouse/data/default/X/all_1_... | 2020-02-26T16:21:26 | 2020-02-26T16:22:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,791,490 | 9,263 | filimonov | /cc @chenxing-xc | 2020-02-27T05:36:48 | 2020-02-27T05:36:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,999,729 | 8,594 | alexey-milovidov | boost::shared_ptr - looks strange.
Is the assignment operator atomic by default (it's unlikely)?
Why don't use MultiVersion as discussed? | 2020-02-27T14:26:57 | 2020-02-27T14:26:57 | {} | MEMBER |
ClickHouse | ClickHouse | 592,493,442 | 9,421 | Tri0L | Got it. Will try recreate dst table with new format.
But why I have error "Inconsistent granularity" when both tables have same granularity?
`MergeTree(field1, (field1, field2, field3), 8192)`
| 2020-02-28T12:32:32 | 2020-02-28T12:32:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 592,494,838 | 2,013 | alexeydubinin | Thanks! | 2020-02-28T12:36:51 | 2020-02-28T12:36:51 | {} | NONE |
ClickHouse | ClickHouse | 591,405,934 | 9,374 | pgruderman | Hi Denis,
Thanks for your reply. Reproducing this problem is difficult and can't be done with a small number of inserts (as far as I know). For my repro, I inserted 3000 rows (one at a time over the native protocol). (You may need more rows than this.)
> ? avgIfMerge ?
> ?cpu_utilization.1 ?
> What is this?
... | 2020-02-26T12:40:36 | 2020-02-26T12:45:51 | {} | NONE |
ClickHouse | ClickHouse | 591,500,419 | 9,367 | ghost | Thanks for your feedback, @alexey-milovidov. Feel free to comment on the code. The fix is done together with @LeoErcolanelli who contributed to some features previsouly. | 2020-02-26T15:57:13 | 2020-02-26T15:57:13 | {} | NONE |
ClickHouse | ClickHouse | 591,509,195 | 9,374 | den-crane | @mark-angelo @pgruderman I have zero relation to CH development. I am just stupid DBA from Canada who are trying to reproduce your issue.
1. I doubt that this issue exists.
2. I doubt that this issue will be fixed this year if it exists.
You can try https://www.altinity.com/ if you need a real support | 2020-02-26T16:13:12 | 2020-02-26T16:15:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,527,204 | 9,374 | den-crane | Floating points numbers are tricky and the result depends of numbers order. Try to check Int64 / IntDiv | 2020-02-26T16:49:01 | 2020-02-26T16:49:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,658,670 | 8,073 | alexey-milovidov | @blinkov Your bot has made chaotic changes. | 2020-02-26T21:41:22 | 2020-02-26T21:41:22 | {} | MEMBER |
ClickHouse | ClickHouse | 591,942,345 | 8,956 | alexey-milovidov | Now I will try to replace it with version 1.1.1. | 2020-02-27T12:21:02 | 2020-02-27T12:21:02 | {} | MEMBER |
ClickHouse | ClickHouse | 592,145,399 | 8,577 | millin | Nested json objects can be declared as named tuples, if anyone implements idea #7787 | 2020-02-27T19:49:14 | 2020-02-27T19:49:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 592,472,225 | 9,408 | alexey-milovidov | @proller Please review the patch: https://github.com/ClickHouse/ClickHouse/commit/8d51824ddcb604b6f179a0216f0d32ba5612bd2e | 2020-02-28T11:25:46 | 2020-02-28T11:25:46 | {} | MEMBER |
ClickHouse | ClickHouse | 591,569,004 | 9,375 | alexey-milovidov | Yes, it will be very easy to add support for custom substitution for NULL. | 2020-02-26T18:19:48 | 2020-02-26T18:19:48 | {} | MEMBER |
ClickHouse | ClickHouse | 591,599,026 | 9,383 | alexey-milovidov | It's unclear what bug was fixed. Looks like an assertion for the internal logic in code. | 2020-02-26T19:24:00 | 2020-02-26T19:24:00 | {} | MEMBER |
ClickHouse | ClickHouse | 591,791,228 | 3,617 | filimonov | @gubinjie #9263 | 2020-02-27T05:35:52 | 2020-02-27T05:35:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 592,243,198 | 8,782 | alexey-milovidov | Why don't merge? | 2020-02-28T00:14:02 | 2020-02-28T00:14:02 | {} | MEMBER |
ClickHouse | ClickHouse | 592,257,335 | 9,374 | den-crane | Yes, AggregatingMergeTree corrupts data during merge of Tuple(AggregateFunction, AggregateFunction).
All CH versions 18.14.19 -- 20.2
[aggrMTTupleCorruption.zip](https://github.com/ClickHouse/ClickHouse/files/4264779/aggrMTTupleCorruption.zip) ./run.sh
AggregatingMT has 4 columns
```
fa Aggre... | 2020-02-28T01:04:42 | 2020-02-28T01:05:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 592,333,660 | 3,617 | gubinjie | @chenxing-xc Hello, we want to know what you said using the clickhouse process's user method. If you use this method, will it solve the previous exception? Thank you | 2020-02-28T06:21:30 | 2020-02-28T06:21:30 | {} | NONE |
ClickHouse | ClickHouse | 592,438,412 | 9,412 | CurtizJ | Looks like compatability is fixed for replication from new version to old, but not the opposite case. Is it possible to fix fetching parts from old to new version replica? | 2020-02-28T09:56:06 | 2020-02-28T09:56:21 | {} | MEMBER |
ClickHouse | ClickHouse | 591,451,880 | 9,380 | alexey-milovidov | Changelog entry is not descriptive enough. I will edit. | 2020-02-26T14:24:50 | 2020-02-26T14:24:50 | {} | MEMBER |
ClickHouse | ClickHouse | 591,768,567 | 9,361 | den-crane | >>ORDER BY tuple()
>what does it do?
A table will be without any order key. (Order by 1).
>>ORDER BY (id0) -- the worst idea as it can be -- NEVER EVER DO THAT.
>why?
Because the idea behind OLAP columnar db is to place a big data into some sort order which allows to compress data better and the same time u... | 2020-02-27T04:02:54 | 2020-02-27T04:02:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,788,565 | 9,383 | filimonov | TableA on disk a, tableB on disk B.
`Alter table tableB attach partition from tableB`
It hardlinks the part from TableA to tableB folder on **disk a** and add a part to list of parts of tableB (maybe some guard /logical error should be fired when adding a part from bad disk).
I think it rates to code preservi... | 2020-02-27T05:26:06 | 2020-02-27T05:26:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,884,846 | 9,367 | alexey-milovidov | It isn't. Fails are not random. | 2020-02-27T10:00:43 | 2020-02-27T10:00:43 | {} | MEMBER |
ClickHouse | ClickHouse | 592,037,075 | 2,013 | alexey-milovidov | Because it is nested inside "event".
Example:
```
JSONExtractString(json, 'event', 'login')
``` | 2020-02-27T16:02:28 | 2020-02-27T16:02:28 | {} | MEMBER |
ClickHouse | ClickHouse | 592,245,258 | 9,070 | alexey-milovidov | 01083_expressions_in_engine_arguments
- well known issue. | 2020-02-28T00:19:20 | 2020-02-28T00:19:20 | {} | MEMBER |
ClickHouse | ClickHouse | 592,463,193 | 9,433 | alexey-milovidov | For the reference: before this change, the linker command contained duplicate `-lpthread` in the middle of the list of libraries and we get dependency on C11 threads (symbols like `cnd_broadcast`) if we build with glibc 2.28+. | 2020-02-28T11:00:04 | 2020-02-28T11:00:37 | {} | MEMBER |
ClickHouse | ClickHouse | 592,498,448 | 9,421 | Tri0L | Another error with dst table in new format =\
```
:) ALTER TABLE tmp.test1 REPLACE PARTITION '202001' FROM data.table;
ALTER TABLE tmp.test1
REPLACE PARTITION '202001' FROM data.ta... | 2020-02-28T12:48:34 | 2020-02-28T12:48:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,443,381 | 9,367 | ghost | hello @KochetovNicolai, when running dbms/src/DataStreams/tests/**collapsing_sorted_stream**.cpp, i've an exit code 11 and when debugging with lldb, i found the stack is full of calls on `static MutablePtr create(std::initializer_list<T> && arg) { return create(std::forward<std::initializer_list<T>>(arg)); }`
It see... | 2020-02-26T14:07:23 | 2020-02-26T14:07:23 | {} | NONE |
ClickHouse | ClickHouse | 591,817,705 | 9,383 | excitoon | > @excitoon Do you mean that it's not a LOGICAL_ERROR but a user related error?
Changed to `BAD_ARGUMENTS` | 2020-02-27T07:09:09 | 2020-02-27T07:09:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 591,970,214 | 3,617 | chenxing-xc | this looks like clickhouse cannot access hdfs datanode. Currently we don't
use the "user" in the uri(user@ip...) to access hdfs, instead use
clickhouse process's user.
filimonov <notifications@github.com> 于2020年2月27日周四 下午1:35写道：
> @gubinjie <https://github.com/gubinjie> #9263
> <https://github.com/ClickHouse/ClickHo... | 2020-02-27T13:28:03 | 2020-02-27T13:28:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 592,015,355 | 2,013 | alexeydubinin | Thanks for reply. But why for this json JSONExtractString not working?
```sql
WITH '{"event": {"login": "djgbjjnwdivin@example.com"}}' as json
select isValidJSON(json), visitParamExtractString(json, 'login'), JSONExtractString(json, 'login');
```
The only working solution is visitParamExtractRaw :confused:
```s... | 2020-02-27T15:17:15 | 2020-02-27T15:30:00 | {} | NONE |
ClickHouse | ClickHouse | 592,109,341 | 9,403 | nikitamikhaylov | ClickHouse doesn't support MultiPolygon's. But we support polygons with holes. | 2020-02-27T18:31:22 | 2020-02-27T18:31:22 | {} | MEMBER |
ClickHouse | ClickHouse | 592,218,628 | 6,943 | alexey-milovidov | Compact data parts has been merged to master but not enabled by default (until the next release and including it). | 2020-02-27T22:51:58 | 2020-02-27T22:51:58 | {} | MEMBER |
ClickHouse | ClickHouse | 592,244,029 | 8,782 | alexey-milovidov | I think that the results of performance test are awful because it is compared to master, but master was optimized multiple times.
Please merge this PR with master. | 2020-02-28T00:16:13 | 2020-02-28T00:16:13 | {} | MEMBER |
ClickHouse | ClickHouse | 592,331,281 | 9,405 | 42997455 | Hi @nikitamikhaylov , the 'max_concurrent_queries_for_user' working for me now, I thought it is global parameter before, add in config.xml, under the max_concurrent_queries wrong place.
Now I put it in users.xml, and working for me now, thanks! | 2020-02-28T06:11:50 | 2020-02-28T06:11:50 | {} | NONE |
ClickHouse | ClickHouse | 592,480,353 | 8,282 | johnatannvmd | Logs from host1
```
2020.02.28 11:40:36.878534 [ 22 ] {} <Information> default.table_name (ReplicatedMergeTreeAlterThread): Version of metadata nodes in ZooKeeper changed. Waiting for structure write lock.
2020.02.28 11:41:37.929610 [ 32 ] {3bc70d72-7acf-493a-900d-ba7f98f444f2} <Error> executeQuery: Code: 48, e.disp... | 2020-02-28T11:50:47 | 2020-02-28T11:50:47 | {} | NONE |
ClickHouse | ClickHouse | 592,154,595 | 9,419 | alexey-milovidov | It looks like we cannot effectively cancel a query from system.mutations (there is no points in execution when the query can be cancelled). | 2020-02-27T20:11:13 | 2020-02-27T20:11:13 | {} | MEMBER |
ClickHouse | ClickHouse | 592,205,496 | 9,256 | alexey-milovidov | @emyr666 We already have:
```
SELECT basename('/home/milovidov/work/ClickHouse/dbms/src/Storages/MergeTree/MergeTreeIndexGranularity.h')
┌─basename('/home/milovidov/work/ClickHouse/dbms/src/Storages/MergeTree/MergeTreeIndexGranularity.h')─┐
│ MergeTreeIndexGranularity.h ... | 2020-02-27T22:16:34 | 2020-02-27T22:16:34 | {} | MEMBER |
ClickHouse | ClickHouse | 592,380,205 | 9,418 | alexey-milovidov | Does not work. | 2020-02-28T07:11:59 | 2020-02-28T07:11:59 | {} | MEMBER |
ClickHouse | ClickHouse | 592,464,330 | 9,432 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/issues/9434 | 2020-02-28T11:03:04 | 2020-02-28T11:03:04 | {} | MEMBER |
ClickHouse | ClickHouse | 592,462,687 | 9,408 | proller | ```
#elif defined(__FreeBSD__)
if (0 != pthread_getthreadid_np(...
``` | 2020-02-28T10:58:40 | 2020-02-28T11:15:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 592,469,219 | 9,408 | alexey-milovidov | @proller We need some code to get thread id on FreeBSD.
(it's not an option to simply return zero) | 2020-02-28T11:17:01 | 2020-02-28T11:17:01 | {} | MEMBER |
ClickHouse | ClickHouse | 574,326,175 | 6,078 | alexey-milovidov | It's harder than expected. | 2020-01-14T19:09:49 | 2020-01-14T19:09:49 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.