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 | 686,718,241 | 12,460 | azat | >@alexey-milovidov alexey-milovidov assigned KochetovNicolai on Aug 2
@alexey-milovidov I will close mine #13625 (anyway for now there is only test) | 2020-09-03T19:45:52 | 2020-09-03T19:46:10 | {} | MEMBER |
ClickHouse | ClickHouse | 686,834,341 | 9,336 | bohutang | > I have the same problem for `20.7.2.30`
>
> ```shell
> Code: 62, e.displayText() = DB::Exception: Syntax error: failed at position 5 ('@@'): @@wait_timeout = 2147483. Expected one of: ROLE, identifier, ROLE DEFAULT, DEFAULT ROLE (version 20.7.2.30 (official build))
> ```
>
> Is there any news about this?
N... | 2020-09-04T00:40:29 | 2020-09-04T00:40:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,836,969 | 14,414 | alexey-milovidov | @qoega Integration test about DNS cache is broken here: https://github.com/ClickHouse/ClickHouse/pull/14323 | 2020-09-04T00:51:33 | 2020-09-04T00:51:33 | {} | MEMBER |
ClickHouse | ClickHouse | 686,888,871 | 14,414 | qoega | > @qoega Integration test about DNS cache is broken here: #14323
More likely test_host_ip_change was flaky all time long and recent decrease in retries from 5 to 3 showed its flakiness.
When I tried to fix it in #14323 it seemed that there is no guarantee, that docker-compose dns removes old IP as soon as we stop o... | 2020-09-04T03:58:47 | 2020-09-04T03:58:47 | {} | MEMBER |
ClickHouse | ClickHouse | 687,110,136 | 14,447 | myrrc | > ReplicatedMergeTree expects two string literal (in single quotes) arguments. Your query has two identifier (in double quotes) arguments. The following works fine:
>
> ```
> :) CREATE TABLE table_repl (number UInt32) ENGINE = ReplicatedMergeTree('/clickhouse/{cluster}/tables/{shard}/table_repl', '{replica}') PARTI... | 2020-09-04T12:23:25 | 2020-09-04T12:23:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,511,516 | 14,446 | den-crane | No way.
You can set a default value - 86400
ALTER TABLE target_table on CLUSTER '{cluster}' MODIFY SETTING merge_with_ttl_timeout = 86400 | 2020-09-03T13:59:08 | 2020-09-03T13:59:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,527,880 | 14,446 | simPod | What if I change default value later? It would have got out of sync. The best option would be to be able to reset this so the value gets inherited. WDYT? | 2020-09-03T14:24:14 | 2020-09-03T14:24:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,860,736 | 14,397 | zhang2014 | Maybe need backport to 20.8? | 2020-09-04T02:20:27 | 2020-09-04T02:20:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 687,078,175 | 14,468 | bohutang | > Stress test - server was shut down by signal after 20 minutes of work. Infrastructure issue.
>
> ```
> 2020.08.05 22:42:52.263740 [ 287 ] {} <Debug> Application: Received termination signal.
> ```
>
> Integration - `test_ip_change_update_dns_cache` does not seem to be affected by this pr.
A small suggestio... | 2020-09-04T11:04:51 | 2020-09-04T11:04:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 687,145,003 | 14,491 | antisaa | > ```shell
> SELECT 'OPTIMIZE TABLE intraday_metrics PARTITION ID \'' || partition_id || '\' FINAL;' FROM system.parts WHERE (table = 'intraday_metrics') AND (partition LIKE '%2020%') GROUP BY partition_id
> ```
Thanks @filimonov , would executing ~300 generated queries produce heavy load on database and affect pe... | 2020-09-04T13:27:41 | 2020-09-04T13:27:41 | {} | NONE |
ClickHouse | ClickHouse | 686,486,707 | 14,436 | den-crane | It's kinda expected behavior and could happen sometimes. `Alter` could not start for 120000ms because it's blocked by other queries.
Try to increase user session parameter
lock_acquire_timeout - 120 ( How long locking request should wait before failing) | 2020-09-03T13:21:56 | 2020-09-03T13:22:24 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,491,593 | 14,419 | simPod | > select * from system.zookeeper where path = '/..zk_path/replicas
Gives result on both replicas.
But on the B, there's `replicas last added replica: A` (version: 1)
For fully replicated tables it's `replicas last added replica: B` (version: 4)
`system.replicas` says the `total_replicas` is 1. Seems like those... | 2020-09-03T13:29:12 | 2020-09-03T13:45:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,635,070 | 14,446 | simPod | https://github.com/ClickHouse/ClickHouse/issues/14449 | 2020-09-03T17:18:00 | 2020-09-03T17:18:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,816,289 | 14,282 | alexey-milovidov | Here is the checklist for the next release 20.8.
The latest stable release is 20.7 and the changelog is here:
https://github.com/ClickHouse/ClickHouse/blob/master/CHANGELOG.md
The changelog for 20.8 will be prepared when 20.8 will be released. | 2020-09-03T23:33:56 | 2020-09-03T23:34:05 | {} | MEMBER |
ClickHouse | ClickHouse | 686,935,526 | 14,457 | alexey-milovidov | Integration tests are broken in https://github.com/ClickHouse/ClickHouse/pull/14323 | 2020-09-04T06:14:46 | 2020-09-04T06:14:46 | {} | MEMBER |
ClickHouse | ClickHouse | 686,935,764 | 14,458 | alexey-milovidov | Integration tests are broken in https://github.com/ClickHouse/ClickHouse/pull/14323 | 2020-09-04T06:15:18 | 2020-09-04T06:15:18 | {} | MEMBER |
ClickHouse | ClickHouse | 687,038,664 | 14,485 | dongbin86 | does Clickhouse currently not support one aggregation method for merging another aggregation state ? | 2020-09-04T09:35:23 | 2020-09-04T09:35:23 | {} | NONE |
ClickHouse | ClickHouse | 687,049,205 | 14,476 | javisantana | A simple example to reproduce the issue
```
with (select groupArray(a) from (select [1, 2] as a)) as aaa
select aaa
``` | 2020-09-04T09:56:47 | 2020-09-04T09:56:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,491,892 | 11,005 | filimonov | Created pr to check that theory https://github.com/ClickHouse/ClickHouse/pull/14443 | 2020-09-03T13:29:36 | 2020-09-03T13:29:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,498,482 | 14,431 | alexey-milovidov | > ClickHouse special build check — Timeout :(
Known issue. | 2020-09-03T13:38:50 | 2020-09-03T13:38:50 | {} | MEMBER |
ClickHouse | ClickHouse | 686,643,850 | 14,447 | tavplubix | ReplicatedMergeTree expects two string literal (in single quotes) arguments. Your query has two identifier (in double quotes) arguments. The following works fine:
```
:) CREATE TABLE table_repl (number UInt32) ENGINE = ReplicatedMergeTree('/clickhouse/{cluster}/tables/{shard}/table_repl', '{replica}') PARTITION BY in... | 2020-09-03T17:35:25 | 2020-09-03T17:35:25 | {} | MEMBER |
ClickHouse | ClickHouse | 686,645,727 | 14,006 | alexey-milovidov | `base64Decode` is using AVX-512 with some shuffle operations. Memory sanitizer don't understand if there was uninitialized memory in SIMD register but it was not used in the result of shuffle. I remove the "bug" label because it's a false positive (still need to fix). | 2020-09-03T17:39:25 | 2020-09-03T17:39:54 | {} | MEMBER |
ClickHouse | ClickHouse | 686,483,758 | 11,005 | filimonov | > yum version resolver installed the latest clickhouse-common-static because clickhouse-server-20.3.8.53 depends on it. But the latest clickhouse-common-static-20.4.2.9 is require clickhouse-server-20.4.2.9
Bit more complicated, update of `clickhouse-common-static` triggered update of `clickhouse-client` (which were... | 2020-09-03T13:17:23 | 2020-09-03T13:17:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,957,605 | 12,494 | qoega | https://clickhouse-test-reports.s3.yandex.net/13647/5ada30df9616f7053a7bef183c92a8b9aa4e8150/fuzzer/report.html#fail1
``` sql
2020.09.04 02:41:09.157209 [ 117 ] {37fea028-54cf-42a3-ad2c-4c8963421173} <Error> : Logical error: 'Unknown numeric column of type: DB::ColumnAggregateFunction'.
clickhouse-server: ../src/C... | 2020-09-04T07:03:41 | 2020-09-04T07:03:41 | {} | MEMBER |
ClickHouse | ClickHouse | 687,069,441 | 14,442 | nikitamikhaylov | `Readme` will be translated into english later in another PR. | 2020-09-04T10:42:34 | 2020-09-04T10:42:34 | {} | MEMBER |
ClickHouse | ClickHouse | 686,464,226 | 14,002 | code-snail | I have the same problem in arm64
/usr/bin/objcopy:Zulu.o: Invalid bfd target | 2020-09-03T12:45:50 | 2020-09-03T12:45:50 | {
"confused": 1
} | NONE |
ClickHouse | ClickHouse | 686,469,944 | 14,435 | l1t1 | ```cpp
INSERT INTO t_enum(x)
SELECT x from t_source
WHERE x in ('hello', 'world')
```
> > how to print highlight syntax, i use < code > tag ,it print
> > `INSERT INTO t_enum(x) SELECT x from t_source WHERE x in ('hello', 'world')`
>
> Three backticks (`` ``` ``) on separate lines before and after, or `` ```s... | 2020-09-03T12:55:51 | 2020-09-03T12:55:51 | {} | NONE |
ClickHouse | ClickHouse | 686,477,019 | 14,403 | den-crane | Is it grow constantly? Did you try to restart CH ? | 2020-09-03T13:07:25 | 2020-09-03T13:07:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,840,020 | 14,452 | alexey-milovidov | Broken in https://github.com/ClickHouse/ClickHouse/pull/4634 | 2020-09-04T01:03:45 | 2020-09-04T01:03:45 | {} | MEMBER |
ClickHouse | ClickHouse | 686,922,566 | 13,156 | azat | @vitlibar can you take a look please? (I'm planning to rebase this once everything will be resolved, but until then it is better to keep all the history)
Btw any chance that this can be included into 20.9? | 2020-09-04T05:44:13 | 2020-09-04T05:44:13 | {} | MEMBER |
ClickHouse | ClickHouse | 687,168,863 | 14,491 | den-crane | OPTIMIZE command just initiates an unplanned merge for partition(s).
When you run `optimize table` it iterates through all partitions in similar way as `select ... from system.parts` and calls `optimize .... partition_id` for each partition. The same amount of job.
In general it's not recommended to use OPTIMIZE ... | 2020-09-04T14:09:15 | 2020-09-04T14:10:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 679,399,756 | 13,873 | den-crane | @zzl-wy try to upgrade to 20.4.9.110
Can you also show "show create table".
I failed to reproduce this issue.
```
create table deadlocktest(A Int64, D Date, S String) Engine=MergeTree partition by D order by A;
insert into deadlocktest select number, '2020-02-01', toString(cityHash64(number)) from numbers(1... | 2020-08-24T22:32:50 | 2020-09-03T14:07:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,524,107 | 14,419 | den-crane | Check at both servers that zhash is identical
```
select replica_name, total_replicas, active_replicas, zookeeper_path,
cityHash64(zookeeper_path) zhash, hex(zookeeper_path)
from system.replicas
where table = '....'
``` | 2020-09-03T14:18:37 | 2020-09-03T14:19:31 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,536,291 | 14,446 | den-crane | I have no idea. I am doing a triage.
Feel free to create a feature request to address it. | 2020-09-03T14:36:06 | 2020-09-03T14:36:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,865,611 | 9,800 | bharatnc | Sounds good 👍 thank you for confirming! | 2020-09-04T02:38:27 | 2020-09-04T02:38:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,883,691 | 14,251 | liangli1213 | @BohuTANG , have you reproduced this issue successfully? | 2020-09-04T03:39:01 | 2020-09-04T03:39:01 | {} | NONE |
ClickHouse | ClickHouse | 686,905,797 | 14,251 | liangli1213 | @BohuTANG I'm just wondering if you can add extra printing information when the exception occur, or add some code to capture the exception, e.g. what big-log message it handles when this issue happen. | 2020-09-04T04:55:56 | 2020-09-04T04:55:56 | {} | NONE |
ClickHouse | ClickHouse | 687,059,639 | 14,468 | KochetovNicolai | Stress test - server was shut down by signal after 20 minutes of work. Infrastructure issue.
```
2020.08.05 22:42:52.263740 [ 287 ] {} <Debug> Application: Received termination signal.
```
Integration - `test_ip_change_update_dns_cache` does not seem to be affected by this pr. | 2020-09-04T10:18:45 | 2020-09-04T10:18:45 | {} | MEMBER |
ClickHouse | ClickHouse | 687,061,392 | 14,490 | alesapin | Probably we can make `merge_with_ttl_timeout` less than 24 hours now because merges with TTL became less aggressive. | 2020-09-04T10:22:52 | 2020-09-04T10:22:52 | {} | MEMBER |
ClickHouse | ClickHouse | 687,104,986 | 14,476 | javisantana | sorry, full steps:
```
:) select version();
SELECT version()
┌─version()─┐
│ 20.7.2.30 │
└───────────┘
1 rows in set. Elapsed: 0.002 sec.
:) SELECT * FROM ( WITH ( SELECT groupArray(a) FROM ( SELECT [1, 2] AS a ) ) AS aaa SELECT aaa ) FORMAT JSON;
SELECT *
FROM
(
WITH
(
... | 2020-09-04T12:12:17 | 2020-09-04T12:12:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,703,968 | 5,319 | filimonov | @alexey-milovidov @amosbird are all the parts done in #12433? Can we close that? | 2020-09-03T19:14:31 | 2020-09-03T19:14:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,822,339 | 14,395 | alexey-milovidov | No backports for experimental features. | 2020-09-03T23:56:02 | 2020-09-03T23:56:02 | {} | MEMBER |
ClickHouse | ClickHouse | 686,835,674 | 14,323 | alexey-milovidov | Most likely this PR broke integration tests:
```
29b97ff | 03.09.2020 20:55:25 | report | task | Integration tests (release) | test_host_ip_change/test.py::test_ip_change_update_dns_cache
-- | -- | -- | -- | -- | --
6a32d30 | 03.09.2020 15:39:48 | report | task | Integration tests (release) | test_host_ip_change/... | 2020-09-04T00:46:14 | 2020-09-04T00:46:14 | {} | MEMBER |
ClickHouse | ClickHouse | 686,905,531 | 9,336 | bohutang | Comments:
Add `SET @@var=x` https://github.com/ClickHouse/ClickHouse/pull/14469 | 2020-09-04T04:55:02 | 2020-09-04T04:55:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,957,700 | 14,403 | shantam04 | 1) Is it grow constantly?
Yes, it grows constantly, and part deleting happens with few hours of delay so, in case of an increase in qps from the source part count increases by a lot.
2) Did you try to restart CH?
Yes, we tried restarting. On the restart, inactive parts are removed but it starts growing again. | 2020-09-04T07:03:52 | 2020-09-04T07:03:52 | {} | NONE |
ClickHouse | ClickHouse | 687,073,677 | 14,469 | bohutang | > @BohuTANG can you also briefly explain why this is needed? Is it just for better compatibility?
Some MySQL drivers send `set @@` query for setup after handshake https://github.com/ClickHouse/ClickHouse/issues/9336#issuecomment-686222422 | 2020-09-04T10:53:21 | 2020-09-04T10:53:49 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 687,152,582 | 14,476 | den-crane | it worked before, at least in 20.1
```
ClickHouse client version 20.1.4.14 (official build).
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 20.1.4 revision 54431.
:) CREATE VIEW tempView AS
:-] WITH(SELECT groupArray(a) FROM (SELECT [1, 2] AS a)) AS aaa
:-] SELECT aaa
... | 2020-09-04T13:40:41 | 2020-09-04T13:48:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,502,036 | 14,432 | alexey-milovidov | > test_storage_rabbitmq/test.py::test_rabbitmq_headers_exchange
@alesapin flaky test. | 2020-09-03T13:44:20 | 2020-09-03T13:44:20 | {} | MEMBER |
ClickHouse | ClickHouse | 686,524,395 | 14,445 | simPod | It has 447172626 rows 🤔 | 2020-09-03T14:19:02 | 2020-09-03T14:19:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,713,036 | 14,450 | filimonov | Also for columnar formats (parquet / arrow) it would be nice to skip / don't read the columns which exists in file / schema, but was not used in query. See also #13526 | 2020-09-03T19:34:11 | 2020-09-03T19:34:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 674,374,268 | 4,006 | bohutang | Update:
- [x] GTID replication support, depend on this pr #13672 finished | 2020-08-15T09:33:34 | 2020-09-03T23:39:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,822,357 | 14,394 | alexey-milovidov | No backports for experimental features. | 2020-09-03T23:56:05 | 2020-09-03T23:56:05 | {} | MEMBER |
ClickHouse | ClickHouse | 686,822,916 | 14,366 | alexey-milovidov | Unbundled build: don't care.
Cache dictionary test: it's wrong. | 2020-09-03T23:58:24 | 2020-09-03T23:58:24 | {} | MEMBER |
ClickHouse | ClickHouse | 686,856,811 | 9,800 | alexey-milovidov | Better to add microsecond variants for both event_time and query_start_time.
It may look somewhat redundant in presence of `query_duration_ms` but it's better for consistency.
Yes, let's do this change for all the tables. | 2020-09-04T02:05:50 | 2020-09-04T02:05:50 | {} | MEMBER |
ClickHouse | ClickHouse | 687,068,571 | 14,469 | KochetovNicolai | Looks like now statement like `set @@abc = 1;` says OK, but ignored.
Works for me. | 2020-09-04T10:40:13 | 2020-09-04T10:40:13 | {} | MEMBER |
ClickHouse | ClickHouse | 687,110,831 | 14,485 | filimonov | -MergeTree tranforms several states into a single state. But obviously it should be the same function.
You can't transform several states of one function (for example avg) to a single state of another function (for example uniq).
The same for count / max.
Also MV only 'see' the block which is currently inserte... | 2020-09-04T12:24:53 | 2020-09-04T12:24:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 687,112,728 | 14,489 | filimonov | You can create some bash/sql script which do some dummy select from odbc table function, and sleep for a while.
i.e. smth like
```
SELECT * FROM odbc('DSN=dummy', dummy); -- that will start clickhouse-odbc-bridge
``` | 2020-09-04T12:29:00 | 2020-09-04T12:29:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 687,116,575 | 14,491 | filimonov | That is not supported right now, it can be added but i'm not 100% sure if it's a good idea (security, etc)
Workaround is - do that outside:
```bash
clickhouse-client --query="SELECT 'OPTIMIZE TABLE intraday_metrics PARTITION ID \'' || partition_id || '\' FINAL;' FROM system.parts WHERE (table = 'intraday_metrics'... | 2020-09-04T12:37:10 | 2020-09-04T12:38:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,711,605 | 13,526 | filimonov | #14450 | 2020-09-03T19:31:12 | 2020-09-03T19:31:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,722,041 | 14,093 | UnamedRus | Disabling `min_merge_bytes_to_use_direct_io=0` fixed async write error and merging issue. | 2020-09-03T19:54:42 | 2020-09-03T19:54:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,827,224 | 5,319 | alexey-milovidov | No. Index analysis is not implemented for IS NULL, IS NOT NULL. | 2020-09-04T00:13:54 | 2020-09-04T00:13:54 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 686,827,768 | 5,319 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/pull/12455 | 2020-09-04T00:15:53 | 2020-09-04T00:15:53 | {} | MEMBER |
ClickHouse | ClickHouse | 686,840,282 | 14,401 | bohutang | > i make a MaterializeMySQL database like this
>
> ```sql
> CREATE DATABASE DB_NAME on cluster cluster01 ENGINE = MaterializeMySQL('ip:port', 'db', 'user, 'pass');
> ```
>
> and all server in cluster crashed, log detail is
>
> ```
> 2020.09.02 17:29:16.532879 [ 309295 ] {} <Fatal> BaseDaemon: (version 20.8.... | 2020-09-04T01:04:45 | 2020-09-04T01:04:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,900,510 | 14,251 | bohutang | > @BohuTANG , have you reproduced this issue successfully?
Yes, it happens sometimes, which is more complicated, and I try to locate the root cause.
From my test, if we do not use `begin/commit` multi-transaction it seems to be fine. | 2020-09-04T04:38:12 | 2020-09-04T04:38:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 687,070,650 | 14,469 | KochetovNicolai | @BohuTANG can you also briefly explain why this is needed? Is it just for better compatibility? | 2020-09-04T10:45:37 | 2020-09-04T10:45:37 | {} | MEMBER |
ClickHouse | ClickHouse | 686,646,143 | 14,221 | alexey-milovidov | AST fuzzer: #12460 | 2020-09-03T17:40:14 | 2020-09-03T17:40:14 | {} | MEMBER |
ClickHouse | ClickHouse | 686,702,323 | 12,433 | filimonov | #5319 | 2020-09-03T19:10:56 | 2020-09-03T19:10:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,824,553 | 12,587 | alexey-milovidov | No ETA. | 2020-09-04T00:04:21 | 2020-09-04T00:04:21 | {} | MEMBER |
ClickHouse | ClickHouse | 686,791,722 | 12,433 | yanchaoguo | > > Which version will be merged
>
> 20.7
你好 虽然20.7版本在create table 时解决了这个问题，但是新的MaterializeMySQL Feature是自动创建表的，如故mysql中存在索引为NULL时依然会报错
Upd (machine translation) : Hello. Although the 20.7 version solved this problem when creating table, the new MaterializeMySQL Feature automatically creates the table. If ther... | 2020-09-03T22:15:57 | 2020-09-04T05:03:12 | {} | NONE |
ClickHouse | ClickHouse | 687,059,737 | 14,468 | KochetovNicolai | Not waiting for perftests. | 2020-09-04T10:19:02 | 2020-09-04T10:19:02 | {} | MEMBER |
ClickHouse | ClickHouse | 687,101,413 | 14,476 | filimonov | > A simple example to reproduce the issue
>
> ```
> with (select groupArray(a) from (select [1, 2] as a)) as aaa
> select aaa
> ```
Tried to reproduce on several versions, but it works. What version you have? | 2020-09-04T12:04:33 | 2020-09-04T12:04:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 687,118,787 | 14,493 | filimonov | Duplicate of #8205 | 2020-09-04T12:41:39 | 2020-09-04T12:41:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,650,638 | 14,449 | den-crane | maybe implement something like
ALTER TABLE ... REMOVE SETTINGS
ALTER TABLE ... RESET SETTINGS | 2020-09-03T17:48:30 | 2020-09-03T17:48:50 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,665,999 | 14,454 | alexey-milovidov | (and because constexpr `n` is greater then zero) | 2020-09-03T18:18:52 | 2020-09-03T18:18:52 | {} | MEMBER |
ClickHouse | ClickHouse | 686,702,916 | 12,433 | filimonov | > Which version will be merged
20.7 | 2020-09-03T19:12:13 | 2020-09-03T19:12:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,787,907 | 10,373 | azat | >Performance — 4 faster, 5 slower, 449 unstable
FWIW here is previous test report (before rebase against upstream/master) - https://clickhouse-test-reports.s3.yandex.net/10373/9e4aa5954e45865d7756e1a6a174d7c1cf0ccfa0/performance_comparison/report.html#fail1
And it does not have those slow downs. | 2020-09-03T22:05:34 | 2020-09-03T22:05:53 | {} | MEMBER |
ClickHouse | ClickHouse | 686,935,976 | 14,455 | alexey-milovidov | Integration tests are broken in https://github.com/ClickHouse/ClickHouse/pull/14323 | 2020-09-04T06:15:48 | 2020-09-04T06:15:48 | {} | MEMBER |
ClickHouse | ClickHouse | 687,069,349 | 14,469 | bohutang | > Looks like now statement like `set @@abc = 1;` says OK, but ignored.
> Works for me.
`set @@` is mysql style, so ignored here. | 2020-09-04T10:42:19 | 2020-09-04T10:42:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,526,140 | 14,419 | simPod | Zookeeper paths differ in a small typo :/ Thank you very much for helping me debug this. I have setup a new monitoring query to catch this in the future. | 2020-09-03T14:21:46 | 2020-09-03T14:21:46 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,764,329 | 14,456 | den-crane | WA:
```
select
if(1 = 2, toUUID(
case when position(x, '-') = 0 then '00000000-0000-0000-0000-000000000000' else x end
), NULL) y
from (select '1dd20d3aa81350af566a117a23c80aba2fbf' x
union all select '1dd20d3a-8135-af56-a117-23c80aba2fbf')
┌────y─┐
│ ᴺᵁᴸᴸ │
└────... | 2020-09-03T21:05:07 | 2020-09-03T21:05:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,935,238 | 14,462 | alexey-milovidov | Integration tests are broken in https://github.com/ClickHouse/ClickHouse/pull/14323 | 2020-09-04T06:14:07 | 2020-09-04T06:14:07 | {} | MEMBER |
ClickHouse | ClickHouse | 687,083,564 | 14,469 | KochetovNicolai | Integration tests failure is not caused by this pr.
Not waiting for `performance` and `integration tests (release)` | 2020-09-04T11:18:52 | 2020-09-04T11:18:52 | {} | MEMBER |
ClickHouse | ClickHouse | 687,119,588 | 8,205 | filimonov | ALTER DICTIONARY SOURCE w/o dictionary recreation is needed for password rotation (atomic replace would be ok too) | 2020-09-04T12:43:08 | 2020-09-04T12:45:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,666,658 | 14,454 | alexey-milovidov | BTW, compiler will optimize this code (using the same assumptions).
Static analyzer tools are also OK with this code.
Removing this condition may hurt readability. Or may improve (as you have an idea to make this change). | 2020-09-03T18:20:13 | 2020-09-03T18:20:13 | {} | MEMBER |
ClickHouse | ClickHouse | 686,789,945 | 14,456 | shawel | I see counter-intuitive about fast performance in branches. I will try your method. thanks @den-crane | 2020-09-03T22:11:02 | 2020-09-03T22:11:02 | {} | NONE |
ClickHouse | ClickHouse | 686,810,349 | 14,454 | alexey-milovidov | > Functional stateless tests (release, DatabaseAtomic) — fail: 1, passed: 2204, skipped: 20
https://github.com/ClickHouse/ClickHouse/pull/14461 | 2020-09-03T23:13:41 | 2020-09-03T23:13:41 | {} | MEMBER |
ClickHouse | ClickHouse | 687,055,347 | 14,476 | vladsaraykin | @javisantana yes, this example is more simple. Thank you! | 2020-09-04T10:09:19 | 2020-09-04T10:09:19 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 687,139,699 | 11,870 | Akazz | AFAIK the failed test `00157_cache_dictionary` in [Functional stateful tests (thread)](https://github.com/ClickHouse/ClickHouse/pull/11870/commits/39d71edd0fd8d5d810a1d0ff911aeda023ab34a3) for 39d71ed takes a lot of time to run and times out under TSan | 2020-09-04T13:18:35 | 2020-09-04T13:18:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,481,069 | 14,441 | den-crane | As a workaround you can move database.sql from metadata folder to somewhere /home or /tmp
Start CH server and run sql from database.sql manually.
| 2020-09-03T13:13:23 | 2020-09-03T13:13:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,551,097 | 14,434 | den-crane | you can create 2 materialized view.
```sql
CREATE TABLE default.aaa (id Int64, mobile String, name String) ENGINE = MergeTree() ORDER BY id;
CREATE TABLE default.bbb (id Int64, mobile String, name String) ENGINE = MergeTree() ORDER BY id;
CREATE TABLE default.mv_ab(mobile String, cnt UInt64) ENGINE = SummingMer... | 2020-09-03T14:58:41 | 2020-09-03T14:58:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,757,850 | 14,456 | den-crane | Clickhouse always calculates all branches (if and else) for PERFORMANCE.
Probably it needs a for new function `toUUIDOrNull`
Before 20.7 it worked because
```
SELECT toUUID('1dd20d3aa81350af566a117a23c80aba2fbf')
┌─toUUID('1dd20d3aa81350af566a117a23c80aba2fbf')─┐
│ 1dd20d3a-8135-af56-a117-23c80ab... | 2020-09-03T20:50:07 | 2020-09-03T20:51:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,944,297 | 13,939 | fastio | Hi @filimonov, The performance check is failed. What should I do right now? Thanks. | 2020-09-04T06:34:22 | 2020-09-04T06:34:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 687,033,911 | 14,485 | dongbin86 | I first do the count aggregation of 1-second granularity, and then calculate the extreme value of 1-minute granularity on this basis.
| 2020-09-04T09:25:29 | 2020-09-04T09:25:48 | {} | NONE |
ClickHouse | ClickHouse | 687,043,486 | 14,412 | KochetovNicolai | BTW, for multiple `JOIN`s it doesn't work because they are rewritten as subqueries, so `CreatingSets` are independent for each subquery and not executed in parallel. E.g. for query
```
SELECT
number AS x,
s.y,
ss.z
FROM numbers(10) AS n
INNER JOIN
(
SELECT number AS y
FROM numbers(5000... | 2020-09-04T09:45:43 | 2020-09-04T09:48:50 | {} | MEMBER |
ClickHouse | ClickHouse | 686,790,075 | 10,368 | nvartolomei | related #2755 | 2020-09-03T22:11:22 | 2020-09-03T22:11:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,808,976 | 12,587 | azat | Looks promising! Any ETA for this? | 2020-09-03T23:09:34 | 2020-09-03T23:09:34 | {} | MEMBER |
ClickHouse | ClickHouse | 686,823,584 | 14,062 | alexey-milovidov | Not backported in 9 days - close. | 2020-09-04T00:01:08 | 2020-09-04T00:01:08 | {} | MEMBER |
ClickHouse | ClickHouse | 686,837,457 | 14,455 | alexey-milovidov | How to install clang-11:
```
git clone git@github.com:llvm/llvm-project.git
cd llvm-project
git checkout release/11.x
cd ..
mkdir build
cd build
CC=clang-10 CXX=clang++-10 cmake -GNinja -DCMAKE_BUILD_TYPE:STRING=Release -DLLVM_ENABLE_PROJECTS=all ../llvm-project/llvm/
ninja
sudo ninja install
hash clang
`... | 2020-09-04T00:53:35 | 2020-09-04T00:54:02 | {} | MEMBER |
ClickHouse | ClickHouse | 683,805,785 | 14,268 | pavitrakumar78 | @filimonov Thank you! `neighbor` function is what I was looking for for the second question related to `lag`!
I'll try to solve the EMA calculation problem in another way!
Edit: mis-tagged! | 2020-08-31T14:15:06 | 2020-08-31T14:15:28 | {} | NONE |
ClickHouse | ClickHouse | 683,931,676 | 12,339 | nikitamikhaylov | > The deadlock is fixed.
>
> The std::move changes I'm not sure are needed at all.
>
>
>
> What's left is the print function. @nikitamikhaylov do you want to continue? Maybe we should close this PR and concentrate on something more vital.
Of course, maybe I will create PR only with function in the near future. | 2020-08-31T17:51:33 | 2020-08-31T17:51:33 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.