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 | 636,508,108 | 11,162 | azat | >I've edited the changelog entry to make it more comprehensible (so the user can understand the scope of this issue):
I always forget making changelog entry "user friendly", thanks! | 2020-05-31T18:19:28 | 2020-05-31T18:19:28 | {} | MEMBER |
ClickHouse | ClickHouse | 636,878,481 | 9,473 | alexey-milovidov | Looks also easy to fix :) | 2020-06-01T14:03:05 | 2020-06-01T14:03:05 | {} | MEMBER |
ClickHouse | ClickHouse | 637,303,271 | 11,346 | qoega | 2docs: we have to update information about OPTIMIZE FINAL behavior is this experiment will be successful. | 2020-06-02T06:17:43 | 2020-06-02T06:17:43 | {} | MEMBER |
ClickHouse | ClickHouse | 637,309,396 | 11,335 | qoega | fails in checks are not relevant | 2020-06-02T06:34:01 | 2020-06-02T06:34:01 | {} | MEMBER |
ClickHouse | ClickHouse | 637,294,142 | 11,331 | shubham721 | Hi @den-crane
Thanks for the reply.
I have checked queries which you have shared, But still, none of it satisfies the functionality I need. I would describe What I did
create table exp(
return_id String,
status String,
rto_count UInt32,
time DateTime('Asia/Kolkata')
)
Engine=MergeTree()
PARTITION BY tuple... | 2020-06-02T05:50:34 | 2020-06-02T08:11:31 | {} | NONE |
ClickHouse | ClickHouse | 637,373,367 | 11,349 | alesapin | All functional tests passed. | 2020-06-02T08:17:25 | 2020-06-02T08:17:25 | {} | MEMBER |
ClickHouse | ClickHouse | 637,385,153 | 1,576 | Siao-pin | > > Still facing same issue in 20.4.3.16
>
> Please post your query
Problem was with mutation. UPDATE was performed on table and later in code this table dropped. This DROP query become immortal. | 2020-06-02T08:37:29 | 2020-06-02T08:37:29 | {
"eyes": 1
} | NONE |
ClickHouse | ClickHouse | 637,421,903 | 7,512 | winter7 | Can we "alter" a database to `Atomic` engine? | 2020-06-02T09:38:46 | 2020-06-02T09:38:46 | {} | NONE |
ClickHouse | ClickHouse | 637,555,889 | 11,369 | akuzm | > > @Enmk FYI
>
> could you please add a link to the test run ?
It's not a test run, I just found it locally with some experimental fuzzer I'm making. The query from the description should reproduce the problem. | 2020-06-02T13:49:31 | 2020-06-02T13:49:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,503,690 | 9,113 | akuzm | > It's OK that performance tests have failed, master doesn't have my setting yet
By the way, if you want to run a performance comparison with master in CI, you can add just a setting that does nothing as a separate PR, and then compare to that. Or maybe I should just make it a warning... | 2020-05-31T17:41:51 | 2020-05-31T17:41:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,532,647 | 11,128 | blinkov | @alexey-milovidov yes, there's some room for improvement. Keywords, function names, and other stuff that's supposed to be part of a query or config need to be marked as such in English source (usually by backticks, but probably headers need some alternative). It's already been done in a significant part of the content,... | 2020-05-31T21:28:13 | 2020-05-31T21:28:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,835,092 | 11,222 | kulkarnikv | Please try 20.4.4. The same behavior can be seen using docker image - screenshot below.
 | 2020-06-01T12:34:35 | 2020-06-01T12:34:35 | {} | NONE |
ClickHouse | ClickHouse | 637,249,178 | 11,350 | alexey-milovidov | Note: it's possible to make a test with `expect`. | 2020-06-02T03:30:29 | 2020-06-02T03:30:29 | {} | MEMBER |
ClickHouse | ClickHouse | 637,353,341 | 78 | jackpgao | +1 | 2020-06-02T07:40:42 | 2020-06-02T07:40:42 | {} | NONE |
ClickHouse | ClickHouse | 637,547,421 | 11,348 | alexey-milovidov | Yes, I understand - avoid reloading of tables is the motivation of this design.
Also we can prepare a wrapper named StorageLazy. It will contain internal table and proxy all methods with special support for drop, detach and getName. | 2020-06-02T13:36:49 | 2020-06-02T13:36:49 | {} | MEMBER |
ClickHouse | ClickHouse | 636,475,278 | 11,316 | alexey-milovidov | > Stress test (thread) — Sanitizer assert (in stderr.log)
Actually it's OOM on Sandbox server (false message). | 2020-05-31T13:59:31 | 2020-05-31T13:59:31 | {} | MEMBER |
ClickHouse | ClickHouse | 636,475,999 | 8,840 | alexey-milovidov | Still failed, will revert.
https://clickhouse-test-reports.s3.yandex.net/11314/43c2a466723660b4ad0f87db194c20f5e7028ae9/integration_tests_(asan).html#fail1
https://clickhouse-test-reports.s3.yandex.net/10678/1b0c9fae6cfaa8e3bfdfc6ba494179f2f4fc14ee/integration_tests_(asan).html#fail1 | 2020-05-31T14:05:17 | 2020-05-31T14:06:21 | {} | MEMBER |
ClickHouse | ClickHouse | 636,486,911 | 11,128 | alexey-milovidov | @blinkov We have to manually revert translation of all function names
and also to fix machine translation script. | 2020-05-31T15:31:20 | 2020-05-31T15:31:20 | {} | MEMBER |
ClickHouse | ClickHouse | 636,773,936 | 5,298 | akuzm | Updated Poco to 1.9.3. | 2020-06-01T10:36:56 | 2020-06-01T10:36:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,236,976 | 11,282 | jencoldeng | thanks for your answer. I konw what to do.
I will make a cron task, calculate a **Daily Summary Table** data using data in **Minute Summary table**, every a few minute.
The Daily Summary Table will use **ReplacingMergeTree**, and use **insert_time** as its version.
Then, a view (not MV) is used to query data i... | 2020-06-02T02:47:20 | 2020-06-02T02:47:20 | {} | NONE |
ClickHouse | ClickHouse | 637,241,961 | 9,933 | alexey-milovidov | @abyss7 Did not respond in 21 days. | 2020-06-02T03:04:54 | 2020-06-02T03:04:54 | {} | MEMBER |
ClickHouse | ClickHouse | 637,358,618 | 6,843 | azat | >BTW, initial_user can be easily fabricated by client.
This can be fixed by overwriting it before sending distributed query
>Maybe better to always use initial_user to calculate per-user limits?
I was thinking about applying limits for both users, "default"+initiator (by using this syntax, i.e. user1+user2, li... | 2020-06-02T07:50:21 | 2020-06-02T07:50:21 | {} | MEMBER |
ClickHouse | ClickHouse | 637,473,328 | 11,237 | akuzm | Arcadia is under maintentance, and other tests have passed before. Will merge. | 2020-06-02T11:21:31 | 2020-06-02T11:21:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,539,213 | 11,370 | akuzm | ```
1 SELECT * FROM VALUES('a UInt64, s String', (100, ''), (257, 'three'), (1.000100016593933, 'three'))
```
Another one. | 2020-06-02T13:23:00 | 2020-06-02T13:23:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,561,793 | 11,355 | alexey-milovidov | And we have two more:
```
test_insert_distributed_load_balancing/test.py::test_prefer_localhost_replica_0_load_balancing_nearest_hostname
test_ttl_move/test.py::test_concurrent_alter_with_ttl_move[concurrently_altering_ttl_mt-MergeTree()]
```
| 2020-06-02T13:59:40 | 2020-06-02T13:59:40 | {} | MEMBER |
ClickHouse | ClickHouse | 636,530,590 | 11,320 | alexey-milovidov | 33 checks passed. | 2020-05-31T21:11:01 | 2020-05-31T21:11:01 | {} | MEMBER |
ClickHouse | ClickHouse | 637,354,287 | 11,028 | vitlibar | > @vitlibar can you help Potya to fix Arcadia build?
>
> ```
> ld: error: undefined symbol: DB::registerFunctionPort(DB::FunctionFactory&)
> >>> referenced by registerFunctionsURL.cpp:37 (/-S/clickhouse/src/Functions/URL/registerFunctionsURL.cpp:37)
> >>> registerFunctionsURL.cpp.o:(DB::registerFunc... | 2020-06-02T07:42:30 | 2020-06-02T07:42:30 | {} | MEMBER |
ClickHouse | ClickHouse | 637,527,000 | 11,369 | akuzm | @Enmk FYI | 2020-06-02T13:01:47 | 2020-06-02T13:01:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,490,859 | 11,162 | alexey-milovidov | @abyss7 No work done on review, unassigned. | 2020-05-31T16:03:01 | 2020-05-31T16:03:01 | {} | MEMBER |
ClickHouse | ClickHouse | 636,545,471 | 8,505 | den-crane | >actually some of my query will execute exceed the max_execution_time
All queries always execute longer than max_execution_time.
set max_execution_time = 1 ;
select sleepEachRow(3) from numbers(1000) settings max_block_size=20;
DB::Exception: Timeout exceeded: elapsed 59.999714133 seconds, maximum: 1
The fir... | 2020-05-31T23:16:23 | 2020-05-31T23:16:23 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,789,497 | 11,222 | lajiyuan | I can't reproduce your problem, because when doing SELECT, clickhouse-server will report an error, that is Received exception from server (version 20.3.5):
Code: 43. DB :: Exception: Received from clickhouse-server: 9000. DB :: Exception: Illegal types of arguments (DateTime64 (3), String) of function greaterOrEquals. | 2020-06-01T11:07:06 | 2020-06-01T11:07:06 | {} | NONE |
ClickHouse | ClickHouse | 636,878,347 | 9,473 | alexey-milovidov | Ok. It means that after "Cancelling query", clickhouse-client should also suppress output of subsequent results. | 2020-06-01T14:02:48 | 2020-06-01T14:02:48 | {} | MEMBER |
ClickHouse | ClickHouse | 637,335,742 | 11,028 | qoega | @vitlibar can you help Potya to fix Arcadia build?
```
ld: error: undefined symbol: DB::registerFunctionPort(DB::FunctionFactory&)
>>> referenced by registerFunctionsURL.cpp:37 (/-S/clickhouse/src/Functions/URL/registerFunctionsURL.cpp:37)
>>> registerFunctionsURL.cpp.o:(DB::registerFunctionsURL(DB::F... | 2020-06-02T07:06:02 | 2020-06-02T07:06:02 | {} | MEMBER |
ClickHouse | ClickHouse | 637,385,893 | 10,376 | qoega | After EtcdKeeper implementation I think we need to rename `common/ZooKeeper` to something more implementation agnostic(like `common/Consensus`). And remove ZooKeeper from exceptions and logs, when we use different backends. Otherwise it will be very confusing. | 2020-06-02T08:38:42 | 2020-06-02T08:38:42 | {} | MEMBER |
ClickHouse | ClickHouse | 637,458,752 | 10,591 | vzakaznikov | Also, we wanted to have something like `last_ttl` and `next_ttl`. Are we going to implement this? | 2020-06-02T10:55:19 | 2020-06-02T10:55:19 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,477,221 | 9,113 | dimarub2000 | It's OK that performance tests have failed, master doesn't have my setting yet | 2020-05-31T14:15:41 | 2020-05-31T14:15:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,484,652 | 11,134 | alexey-milovidov | Looks like we should rebuild (and publish) some Docker images? | 2020-05-31T15:15:14 | 2020-05-31T15:15:14 | {} | MEMBER |
ClickHouse | ClickHouse | 636,530,640 | 11,317 | alexey-milovidov | > Stress test (thread) — Sanitizer assert (in stderr.log)
False positive. | 2020-05-31T21:11:26 | 2020-05-31T21:11:26 | {} | MEMBER |
ClickHouse | ClickHouse | 636,935,353 | 11,282 | jencoldeng | This is my code:
```
--Detail table, nothing to change
CREATE TABLE Detail_local(
statdate UInt64,
stattime UInt64,
domain String,
flow UInt64,
request UInt64)
ENGINE = ReplicatedMergeTree('/clickhouse/tables/Detail/{shard}', '{replica}')
PARTITION BY statdate
ORDER BY (statdate,stattime,do... | 2020-06-01T15:52:06 | 2020-06-01T15:52:06 | {} | NONE |
ClickHouse | ClickHouse | 637,034,926 | 11,338 | alexey-milovidov | Initial issue was fixed but there are still some things to fix. | 2020-06-01T18:38:48 | 2020-06-01T18:39:02 | {} | MEMBER |
ClickHouse | ClickHouse | 636,481,524 | 11,318 | alexey-milovidov | Ok. All tests have passed.
> Functional stateless tests (memory) — Some queries hung
Actually it's Ok (false positive). Some queries continue to run in background for 10 seconds, and the fail happeded due to test order randomization. Will fix it in our infrastructure. | 2020-05-31T14:50:52 | 2020-05-31T14:50:52 | {} | MEMBER |
ClickHouse | ClickHouse | 636,481,641 | 11,318 | alexey-milovidov | > Integration tests (asan) — failed:1
It's flacky test, #11320
> Stress test (thread) — Sanitizer assert (in stderr.log)
Actually it's OOM on our test machine. | 2020-05-31T14:51:40 | 2020-05-31T14:51:40 | {} | MEMBER |
ClickHouse | ClickHouse | 636,489,396 | 11,244 | erwan-koffi | Indeed if the two rows were inserted with the same query, this would lead for correct result.
This behaviour is coherent with other Merge Tree engines, but not with states and AggregatedMergeTree that are supposed to support INSERT SELECT.
Maybe I overlooked it but is there a way to use the INSERT SELECT syntax f... | 2020-05-31T15:51:23 | 2020-05-31T15:51:23 | {} | NONE |
ClickHouse | ClickHouse | 636,487,959 | 11,230 | alexey-milovidov | Need to apply some style fixes:
Http -> HTTP
(I'll do it) | 2020-05-31T15:40:03 | 2020-05-31T22:22:23 | {} | MEMBER |
ClickHouse | ClickHouse | 637,177,280 | 11,328 | den-crane | What CH version do you use? | 2020-06-01T23:23:02 | 2020-06-01T23:23:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,179,424 | 11,282 | den-crane | I meant you can only calculate MAX from preadaggregated by minute MV.
**It's impossible** in CH to calculate `MAX ( SUM() group by minute) group by day` using MV.
I think it have no sense to built such MV, it's easier to use Detail_local directly. Or calculate `MAX ( SUM() group by minute) group by day` after day... | 2020-06-01T23:30:02 | 2020-06-01T23:30:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,522,935 | 11,348 | tavplubix | > Why we have so inconvenient interface of Lazy database?
We can make it more convenient by saving not null `current_storage` in `DatabaseLazyIterator::next()` and just returning it from `DatabaseLazyIterator::table()`. But it will cause reloading of all tables in Lazy database when executing `DROP DATABASE` and `SE... | 2020-06-02T12:54:03 | 2020-06-02T12:54:03 | {} | MEMBER |
ClickHouse | ClickHouse | 637,554,053 | 11,369 | Enmk | > @Enmk FYI
could you please add a link to the test run ? | 2020-06-02T13:46:18 | 2020-06-02T13:46:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,715,016 | 11,166 | alesapin | Yes, in `20.3` alters became completely asynchronous operation, and errors are not checked. I have created the feature request for such a check in https://github.com/ClickHouse/ClickHouse/issues/11329. | 2020-06-01T08:59:20 | 2020-06-01T08:59:20 | {} | MEMBER |
ClickHouse | ClickHouse | 636,860,621 | 11,307 | alexey-milovidov | Thank you! I was puzzled looking at this test. | 2020-06-01T13:27:34 | 2020-06-01T13:27:34 | {} | MEMBER |
ClickHouse | ClickHouse | 637,236,489 | 11,288 | alexey-milovidov | @filimonov Yes, 20.5 branch does not exist right now, it will be created after this checklist will be done. | 2020-06-02T02:45:43 | 2020-06-02T02:45:43 | {} | MEMBER |
ClickHouse | ClickHouse | 637,551,576 | 10,294 | mrigger | Thanks a lot for pointing out our work @Schaefer-Thomas! SQLancer is now freely available at https://github.com/sqlancer/sqlancer. I even started with a very preliminary ClickHouse implementation, but did not have time to continue working on it. @alexey-milovidov, feel free to reach out to me if you have any plans in u... | 2020-06-02T13:42:07 | 2020-06-02T13:42:07 | {} | NONE |
ClickHouse | ClickHouse | 637,562,468 | 11,345 | alexey-milovidov | One case is fixed, more remaining. CC @Enmk | 2020-06-02T14:00:49 | 2020-06-02T14:01:55 | {} | MEMBER |
ClickHouse | ClickHouse | 636,487,264 | 11,256 | alexey-milovidov | Logical error:
```
2020.05.29 02:58:44.483982 [ 207 ] {} <Trace> test_3bd5c8.alter_table2: Renaming temporary part tmp_mut_-2_0_0_1_9 to -2_0_0_1_9.
2020.05.29 02:58:44.485340 [ 207 ] {} <Error> : Logical error: 'Part -2_0_0_1_9 intersects next part -2_0_0_2 (state Committed). It is a bug.'.
```
In test `00992... | 2020-05-31T15:34:14 | 2020-05-31T15:34:14 | {} | MEMBER |
ClickHouse | ClickHouse | 636,625,206 | 9,473 | filimonov | > @filimonov
>
> > ability to stop printing the big output to console when Ctrl+C is pressed
>
> Usually the issue is in the terminal side.
> clickhouse-client already sent it's output to the terminal but it takes long time to display it.
@alexey-milovidov That is NOT true in most of the cases. Check my sampl... | 2020-06-01T05:44:26 | 2020-06-01T05:48:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,870,047 | 11,288 | filimonov | i'm also adding more settings for Kafka and adjusting some defaults (ETA ~ 1d), It would be great to pick those to 20.5 too. | 2020-06-01T13:46:15 | 2020-06-01T13:46:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,243,304 | 11,328 | shubham721 | @den-crane 20.3.8.53 | 2020-06-02T03:09:16 | 2020-06-02T08:13:06 | {} | NONE |
ClickHouse | ClickHouse | 636,513,120 | 9,113 | dimarub2000 | > > It's OK that performance tests have failed, master doesn't have my setting yet
>
> By the way, if you want to run a performance comparison with master in CI, you can add just a setting that does nothing as a separate PR, and then compare to that. Or maybe I should just make it a warning...
Ок, good to know, b... | 2020-05-31T18:57:27 | 2020-05-31T18:57:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,632,376 | 11,308 | filimonov | That should be supported by Kafka table itself, not by Materialized View.
Working on it right now. | 2020-06-01T06:08:19 | 2020-06-01T06:08:19 | {
"rocket": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,201,622 | 11,346 | alexey-milovidov | It should be ready for merge even without ReplicatedMergeTree. | 2020-06-02T00:44:33 | 2020-06-02T00:44:33 | {} | MEMBER |
ClickHouse | ClickHouse | 637,452,617 | 10,591 | vzakaznikov | There is `move_ttl_info.expression` but there is no `delete_ttl_info.expression`?
| 2020-06-02T10:41:06 | 2020-06-02T10:41:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,479,041 | 11,315 | alexey-milovidov | > Integration tests (asan) — failed:1, passed:647, error:0
See https://github.com/ClickHouse/ClickHouse/pull/11320 | 2020-05-31T14:30:54 | 2020-05-31T14:30:54 | {} | MEMBER |
ClickHouse | ClickHouse | 636,503,106 | 11,322 | alexey-milovidov | Yes, it works.
Before:
```
for i in {1..1000}; do clickhouse-client -n < 01292_optimize_final_force_.sql; done
10 0
45 0
45 0
10 0
10 0
45 0
45 0
10 0
55 0
55 0
55 0
45 0
10 0
45 0
10 0
45 0
10 0
10 0
```
... | 2020-05-31T17:36:51 | 2020-05-31T17:36:51 | {} | MEMBER |
ClickHouse | ClickHouse | 636,582,134 | 11,279 | zhlovy | yes
Parameter of 'max_memory_usage' changes should not affect the restart of the service
so this maybe a bug or
do we need to pay attentions to restart the server when storage policy is 'moving_from_ssd_to_hdd' ? | 2020-06-01T02:38:46 | 2020-06-01T02:38:46 | {} | NONE |
ClickHouse | ClickHouse | 636,681,521 | 11,166 | donghy-coredumped | in version 20.1.6 , I execute the same operation and return the information as follows:
localhost :) alter table tt modify column event UInt32;
ALTER TABLE tt
MODIFY COLUMN
`event` UInt32
Received exception from server (version 20.1.6):
Code: 6. DB::Exception: Received from localhost:9000. DB:... | 2020-06-01T08:03:05 | 2020-06-01T08:03:05 | {} | NONE |
ClickHouse | ClickHouse | 637,207,520 | 11,330 | alexey-milovidov | > Unit tests MSAN clang-10
Fixed in master. | 2020-06-02T01:06:36 | 2020-06-02T01:06:36 | {} | MEMBER |
ClickHouse | ClickHouse | 637,550,899 | 11,309 | alexey-milovidov | It's done for non-Replicated MergeTree but still to-do for ReplicatedMergeTree. | 2020-06-02T13:40:58 | 2020-06-02T13:40:58 | {} | MEMBER |
ClickHouse | ClickHouse | 637,052,400 | 9,897 | alexey-milovidov | We should write history file after every query.
This should also solve the issue when you run two clickhouse-clients on the same server and the history is saved only from one of them. | 2020-06-01T19:17:12 | 2020-06-01T19:17:23 | {} | MEMBER |
ClickHouse | ClickHouse | 637,176,586 | 11,331 | den-crane | Examples:
https://gist.github.com/den-crane/49ce2ae3a688651b9c2dd85ee592cb15
https://gist.github.com/den-crane/d03524eadbbce0bafa528101afa8f794
https://gist.github.com/den-crane/6eff375752a236a456e1b3dc2ca7db62 | 2020-06-01T23:20:45 | 2020-06-01T23:20:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,547,717 | 11,348 | alexey-milovidov | @nikvas0 What do you think? | 2020-06-02T13:37:20 | 2020-06-02T13:37:20 | {} | MEMBER |
ClickHouse | ClickHouse | 637,561,378 | 11,355 | alexey-milovidov | > failed:2
@alesapin Missing whitespace after colon. | 2020-06-02T13:59:00 | 2020-06-02T13:59:00 | {} | MEMBER |
ClickHouse | ClickHouse | 636,478,346 | 10,678 | alexey-milovidov | > Yandex synchronization check
Will fix after merge. | 2020-05-31T14:25:11 | 2020-05-31T14:25:11 | {} | MEMBER |
ClickHouse | ClickHouse | 636,488,081 | 11,073 | alexey-milovidov | Merged in #11252 | 2020-05-31T15:40:53 | 2020-05-31T15:40:53 | {} | MEMBER |
ClickHouse | ClickHouse | 636,490,624 | 9,577 | alexey-milovidov | @excitoon
We should do it in the same way as for `remote_servers` (cluster configuration).
See ConfigReloader, setClustersConfig...
Data should be available as shared_ptr under mutex (it is copied and assigned to the new version under mutex).
Storages can obtain a shared copy of this shared_ptr every time wh... | 2020-05-31T16:01:12 | 2020-05-31T16:01:12 | {} | MEMBER |
ClickHouse | ClickHouse | 637,174,205 | 11,332 | den-crane | -->ENGINE = AggregatingMergeTree() ORDER BY (m_date ,country)
Otherwise country column calculated by any() function by AggregatingMergeTree
```
CREATE MATERIALIZED VIEW default.Items_MV
ENGINE = AggregatingMergeTree() PARTITION BY m_date ORDER BY (m_date ,country)
AS SELECT
m_date,
country,
countStat... | 2020-06-01T23:13:41 | 2020-06-01T23:17:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,250,243 | 10,357 | YiuRULE | @fibersel Did you have time to work on it yet ? I have a bit of free time now, but I will let you the task if you are already on it :) | 2020-06-02T03:34:12 | 2020-06-02T03:34:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,564,085 | 11,259 | alexey-milovidov | Flaky test `test_ttl_move/test.py::test_concurrent_alter_with_ttl_move[concurrently_altering_ttl_mt-MergeTree()]` is discovered. | 2020-06-02T14:03:23 | 2020-06-02T14:03:23 | {} | MEMBER |
ClickHouse | ClickHouse | 636,475,181 | 11,314 | alexey-milovidov | > Stress test (thread) — Sanitizer assert (in stderr.log)
Actually it's OOM on CI machine (false message). | 2020-05-31T13:58:46 | 2020-05-31T13:58:46 | {} | MEMBER |
ClickHouse | ClickHouse | 636,475,412 | 11,313 | alexey-milovidov | > Performance — Errors while building the report.
It's an error in the test script @akuzm | 2020-05-31T14:00:25 | 2020-05-31T14:00:25 | {} | MEMBER |
ClickHouse | ClickHouse | 636,480,862 | 8,840 | alexey-milovidov | And
https://clickhouse-test-reports.s3.yandex.net/11318/70eaae38f2fe5c91d6a43b5166df623996826042/integration_tests_(asan).html#fail1 | 2020-05-31T14:45:30 | 2020-05-31T14:45:30 | {} | MEMBER |
ClickHouse | ClickHouse | 636,491,732 | 11,162 | alexey-milovidov | Ok.
I've edited the changelog entry to make it more comprehensible (so the user can understand the scope of this issue):
From:
> Fix data skipping indexes for columns with additional actions during merge
To:
> If data skipping index is dependent on columns that are going to be modified during background merg... | 2020-05-31T16:09:56 | 2020-05-31T16:09:56 | {} | MEMBER |
ClickHouse | ClickHouse | 636,519,907 | 9,473 | alexey-milovidov | I found that the root cause of slowdown is calculating visible widths of strings.
I have fixed it here: https://github.com/ClickHouse/ClickHouse/pull/11323 | 2020-05-31T19:50:44 | 2020-05-31T19:50:44 | {} | MEMBER |
ClickHouse | ClickHouse | 636,626,875 | 9,473 | filimonov | And one more related to Ctrl+C, really annoying: https://github.com/ClickHouse/ClickHouse/issues/9897 | 2020-06-01T05:50:15 | 2020-06-01T05:50:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,362,000 | 11,292 | azat | FWIW
There are also the following settings, that one may want to change in similar cases:
- `max_bytes_before_external_group_by`
- `max_bytes_before_external_sort`
- `max_bytes_before_remerge_sort`
And it looks like all of this can be archived by using separate user for distributed queries, by changing `user` ... | 2020-06-02T07:56:23 | 2020-06-02T07:56:23 | {} | MEMBER |
ClickHouse | ClickHouse | 637,465,435 | 10,357 | fibersel | @YiuRULE yes, take it, I have overestimated my free time. | 2020-06-02T11:08:05 | 2020-06-02T11:08:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 637,498,715 | 11,370 | akuzm | ```
SELECT groupUniqArray(v) FROM values('id int, v Array(int)', tuple(1048577)) GROUP BY id
```
Another variant of probably the same bug. | 2020-06-02T12:08:36 | 2020-06-02T12:08:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,484,883 | 10,484 | alexey-milovidov | Something wrong with build scripts:
```
File "../clickhouse-server_0.0.0.3562_all.deb" not found.
./utils/release/release_lib.sh: line 194: cd: clickhouse-server-0.0.0.3562: No such file or directory
``` | 2020-05-31T15:17:03 | 2020-05-31T15:17:03 | {} | MEMBER |
ClickHouse | ClickHouse | 636,485,035 | 11,293 | alexey-milovidov | Superseded by #11307 | 2020-05-31T15:18:17 | 2020-05-31T15:18:17 | {} | MEMBER |
ClickHouse | ClickHouse | 636,487,454 | 11,252 | alexey-milovidov | Data race in MemoryTracker - already fixed by @KochetovNicolai in #11249 | 2020-05-31T15:35:59 | 2020-05-31T15:35:59 | {} | MEMBER |
ClickHouse | ClickHouse | 637,016,024 | 10,212 | winofi | For me `ALTER TABLE $CLICKHOUSE_DB.rx DROP PARTITION toYYYYMM(subtractMonths(now(), 1));` stopped working after updating to version 20.4.4.18 (official build) recently and fails with the error described above and in #10911.
While `select toYYYYMM(subtractMonths(now(), 1))` returns 202005 `ALTER TABLE $CLICKHOUSE_DB.r... | 2020-06-01T17:58:37 | 2020-06-01T17:58:37 | {} | NONE |
ClickHouse | ClickHouse | 636,478,978 | 8,840 | alexey-milovidov | And
https://clickhouse-test-reports.s3.yandex.net/11315/3574dfe1db3e3d2bdaab1346b948357e6a9a1ec7/integration_tests_(asan).html#fail1 | 2020-05-31T14:30:21 | 2020-05-31T14:30:21 | {} | MEMBER |
ClickHouse | ClickHouse | 636,489,907 | 10,621 | alexey-milovidov | Ok. | 2020-05-31T15:55:32 | 2020-05-31T15:55:32 | {} | MEMBER |
ClickHouse | ClickHouse | 636,511,849 | 9,473 | alexey-milovidov | Actually it blocks only when client is receiving result. | 2020-05-31T18:47:10 | 2020-05-31T18:47:10 | {} | MEMBER |
ClickHouse | ClickHouse | 636,515,784 | 11,321 | alexey-milovidov | > Merge with master — Cannot fetch mergecommit
It is GitHub-related failure. | 2020-05-31T19:19:08 | 2020-05-31T19:19:08 | {} | MEMBER |
ClickHouse | ClickHouse | 636,681,040 | 8,505 | jackpgao | > > actually some of my query will execute exceed the max_execution_time
>
> All queries always execute longer than max_execution_time.
>
> set max_execution_time = 1 ;
> select sleepEachRow(3) from numbers(1000) settings max_block_size=20;
> DB::Exception: Timeout exceeded: elapsed 59.999714133 seconds, maximu... | 2020-06-01T08:02:08 | 2020-06-01T08:02:08 | {} | NONE |
ClickHouse | ClickHouse | 636,731,630 | 10,267 | Anshelen | v20.4.2.9-stable - after fix problem still exists | 2020-06-01T09:24:52 | 2020-06-01T09:24:52 | {} | NONE |
ClickHouse | ClickHouse | 637,260,722 | 10,357 | alexey-milovidov | @YiuRULE I think, it's perfectly Ok to grab this task :) | 2020-06-02T04:13:53 | 2020-06-02T04:13:53 | {} | MEMBER |
ClickHouse | ClickHouse | 637,406,273 | 11,356 | YiuRULE | Related to https://github.com/ClickHouse/ClickHouse/issues/10357 | 2020-06-02T09:12:16 | 2020-06-02T09:12:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 632,630,451 | 11,084 | alexey-milovidov | Not ready.
clickhouse-benchmark prints this message at startup:
```
<jemalloc>: Conf string ends with key
``` | 2020-05-22T10:55:30 | 2020-05-22T10:55:30 | {} | MEMBER |
ClickHouse | ClickHouse | 632,642,802 | 11,111 | alexey-milovidov | I have missed the release of gcc-10 (it was on May 7),
now we have to start support it anyway, because some users will use it.
| 2020-05-22T11:25:58 | 2020-05-22T11:25:58 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.