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 | 778,065,373 | 20,343 | filimonov | It looks like clickhouse tries to parse POST as a continuation of query, caused by comma at the beginning of the protobuf message, which makes clickhouse parse think that it's a continuation of SETTINGS. | 2021-02-12T08:54:39 | 2021-02-12T09:03:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,170,159 | 19,854 | Avogar | @fuwhu Add a test please. (See an example in tests/queries/0_stateless/00632_aggregation_window_funnel.sql) | 2021-02-12T12:35:21 | 2021-02-12T12:44:38 | {} | MEMBER |
ClickHouse | ClickHouse | 778,184,448 | 19,257 | nikitamikhaylov | 
| 2021-02-12T13:06:47 | 2021-02-12T13:06:47 | {} | MEMBER |
ClickHouse | ClickHouse | 777,612,451 | 20,197 | filimonov | See
https://clickhouse.tech/docs/en/operations/requirements/
https://blog.cloudflare.com/clickhouse-capacity-estimation-framework/ | 2021-02-11T16:16:53 | 2021-02-11T16:16:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,892,987 | 20,349 | l1t1 | https://github.com/ClickHouse/ClickHouse/pull/16787 | 2021-02-12T00:44:48 | 2021-02-12T00:44:48 | {} | NONE |
ClickHouse | ClickHouse | 778,192,849 | 20,363 | alesapin | Will check in branch | 2021-02-12T13:23:39 | 2021-02-12T13:23:39 | {} | MEMBER |
ClickHouse | ClickHouse | 778,497,723 | 20,346 | den-crane | It seems the issue is in skip indexes on transactions_temp_3
when we execute
```
SELECT count()
FROM temp
WHERE cityHash64(casino, game, round, trans, type) NOT IN
``` | 2021-02-12T22:47:28 | 2021-02-12T22:47:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,480,843 | 20,308 | alexey-milovidov | It looks like you accidentially mistype "catboost" instead of "clickhouse". | 2021-02-11T14:03:18 | 2021-02-11T14:03:18 | {} | MEMBER |
ClickHouse | ClickHouse | 777,594,083 | 20,283 | alexey-milovidov | Broken in #18145. | 2021-02-11T15:50:41 | 2021-02-11T15:50:41 | {} | MEMBER |
ClickHouse | ClickHouse | 777,621,496 | 19,317 | abyss7 | LGTM | 2021-02-11T16:30:28 | 2021-02-11T16:30:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,643,463 | 20,175 | amosbird | > > Test failure looks unrelated.
>
> It's related - it breaks master.
Oops. The test failure looks quite obscure w.r.t to parser changes. I still don't see how it's related. | 2021-02-11T17:02:05 | 2021-02-11T17:02:05 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 778,086,147 | 20,297 | alesapin | `test_distributed_ddl_parallel/test.py::test_all_in_parallel` -- not related to changes | 2021-02-12T09:36:54 | 2021-02-12T09:36:54 | {} | MEMBER |
ClickHouse | ClickHouse | 778,086,698 | 20,261 | alesapin | Out of support | 2021-02-12T09:37:59 | 2021-02-12T09:37:59 | {} | MEMBER |
ClickHouse | ClickHouse | 778,091,770 | 20,344 | alesapin | Fuzzer https://github.com/ClickHouse/ClickHouse/issues/20366.
Logical error better than segfault! | 2021-02-12T09:48:46 | 2021-02-12T09:48:46 | {} | MEMBER |
ClickHouse | ClickHouse | 778,293,457 | 20,346 | filimonov | | nnn | v20_12 | v21_2 | diff |
|:-|-:|-:|-:|
| Query | 1 | 1 | 0 |
| SelectQuery | 1 | 1 | 0 |
| FileOpen | 595 | 555 | -40 |
| Seek | 141 | 43 | -98 |
| ReadBufferFromFileDescriptorRead | 13916 | 13865 | -51 |
| ReadBufferFromFileDescriptorReadBytes | 14140342916 | 14089373955 | -50968961 |
| ReadCompressedB... | 2021-02-12T16:21:49 | 2021-02-12T16:23:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,333,022 | 19,580 | alesapin | ```
2021.02.12 14:20:58.800988 [ 17728 ] {de3e7894-b401-4f7d-8530-90cd5ab06682} <Debug> executeQuery: (from [::1]:45792, using production parser) (comment: /usr/share/clickhouse-test/queries/0_stateless/01520_client_print_query_id.expect) SELECT * FROM numbers(34599)
2021.02.12 14:20:58.916484 [ 17728 ] {de3e7894-b40... | 2021-02-12T17:29:22 | 2021-02-12T17:29:22 | {} | MEMBER |
ClickHouse | ClickHouse | 777,505,322 | 17,623 | nvartolomei | As I understand Kerberos could be implemented as just another mechanisms
for SASL.
On Thu, 11 Feb 2021, 12:39 Denis Glazachev, <notifications@github.com>
wrote:
> When integrating SASL in native TCP protocol, bear in mind that Kerberos
> support may eventually be added there too, so the integration better be
> reason... | 2021-02-11T14:36:00 | 2021-02-11T14:36:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,514,444 | 10,128 | filimonov | Forbid TTL for tables with old syntax: #20282 #17904 #19064 | 2021-02-11T14:47:40 | 2021-02-11T14:47:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,601,725 | 20,216 | filimonov | Aggregation function states internally just store some structures the most suitable to store correct & the most compact form of function state (because states are passed between clickhouse nodes during distributed queries).
Every aggregate function uses its own structures, very context-specific.
So client libra... | 2021-02-11T16:01:23 | 2021-02-11T16:01:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,613,873 | 20,208 | hodgesrm | Hi! Can we get some review on this one? We have a blog article pending and want to be able to point to the [great!] community docs. | 2021-02-11T16:19:02 | 2021-02-11T16:19:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,699,636 | 20,344 | azat | >Maybe it fixes #20046?
Indeed, will add Fixes tag into the description | 2021-02-11T18:31:59 | 2021-02-11T18:31:59 | {} | MEMBER |
ClickHouse | ClickHouse | 778,063,429 | 20,343 | filimonov | ```
2021.02.12 09:49:08.423834 [ 2472 ] {f6127b60-a00c-4537-99e7-e166f9bc828d} <Error> DynamicQueryHandler: Code: 62, e.displayText() = DB::Exception: Syntax error: failed at position 108 ('') (line 2, col 2): ����: 123456789B
87654322342�
�̣������. Un... | 2021-02-12T08:50:33 | 2021-02-12T08:50:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,181,687 | 20,350 | tavplubix | Maybe we should forbid subqueries as function arguments on parser level? We can allow it for table functions only by passing special flag (something like `is_table_function`) to `ParserFunction` constructor from `ParserTableExpression::parseImpl(...)`, `ParserCreateTableQuery::parseImpl(...)` and similar places. | 2021-02-12T13:01:07 | 2021-02-12T13:01:07 | {} | MEMBER |
ClickHouse | ClickHouse | 778,306,280 | 20,381 | filimonov | added here https://github.com/ClickHouse/ClickHouse/pull/12277
| 2021-02-12T16:44:05 | 2021-02-12T16:44:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,674,477 | 20,287 | alexey-milovidov | #20032 | 2021-02-11T17:50:08 | 2021-02-11T17:50:08 | {} | MEMBER |
ClickHouse | ClickHouse | 778,286,405 | 20,378 | alesapin | https://github.com/ClickHouse/ClickHouse/issues/18875 | 2021-02-12T16:09:26 | 2021-02-12T16:09:26 | {} | MEMBER |
ClickHouse | ClickHouse | 778,312,629 | 20,300 | alesapin | https://github.com/ClickHouse/ClickHouse/pull/20398 | 2021-02-12T16:55:03 | 2021-02-12T16:55:03 | {} | MEMBER |
ClickHouse | ClickHouse | 778,341,088 | 20,085 | gyuton | >
>
> I've only added more examples with different strings.
>
> Thank you for you PR.
@vdimir, you're welcome! | 2021-02-12T17:43:56 | 2021-02-12T17:43:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 753,023,102 | 18,649 | siradjev | > In case of allocating max_server_memory_usage (0.9*RAM) system file cache will be tiny and CH will be slow.
True, that is why it would be wise to make preallocation fraction configurable. So if there are many non RAM consuming queries - it makes sense to preallocate less, and if majority are RAM heavy queries it m... | 2020-12-31T18:33:19 | 2021-02-12T19:31:40 | {} | NONE |
ClickHouse | ClickHouse | 778,497,472 | 20,346 | den-crane | https://github.com/ClickHouse/ClickHouse/issues/20455 -- x USING (row_hash) -- produces the incorrect result.
| 2021-02-12T22:46:50 | 2021-02-12T22:46:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,484,165 | 20,308 | alexey-milovidov | Checkout release branch, for release you are insterested (the latest stable version is 21.2):
```
git checkout 21.2
git submodule init && git submodule sync && git submodule update
```
and repeat. | 2021-02-11T14:08:01 | 2021-02-11T14:08:01 | {} | MEMBER |
ClickHouse | ClickHouse | 777,535,832 | 20,334 | alexey-milovidov | Sometimes we put links to stackoverflow in comments as a reference, so people may read some discussions and answers. It does not necessarily means that the code is copy-pasted from stackoverflow or represent a derivative work of content from stackoverflow. | 2021-02-11T14:58:21 | 2021-02-11T14:58:21 | {} | MEMBER |
ClickHouse | ClickHouse | 777,567,268 | 20,175 | abyss7 | > Test failure looks unrelated.
It's related - it breaks master. | 2021-02-11T15:16:37 | 2021-02-11T15:16:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,589,848 | 19,580 | alesapin | https://github.com/ClickHouse/ClickHouse/pull/20175 | 2021-02-11T15:44:59 | 2021-02-11T15:44:59 | {} | MEMBER |
ClickHouse | ClickHouse | 777,604,944 | 20,188 | filimonov | Optimal partition size is between few gigs and 150 Gb (may also be bigger / smaller, if there are more pros than cons).
Optimal partition in a single table up to hundreds.
So if you collect 150 Gb of data per hour and you data retention is limited to couple of days then you can keep hourly.
| 2021-02-11T16:06:00 | 2021-02-11T16:07:42 | {
"+1": 5
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,618,209 | 20,166 | MaxJPowers | send_timeout/receive_timeout don't change the the timeout duration, I opened a seperate issue here #20213. | 2021-02-11T16:25:26 | 2021-02-11T16:25:26 | {} | NONE |
ClickHouse | ClickHouse | 777,692,726 | 20,344 | alexey-milovidov | Maybe it fixes #20046? | 2021-02-11T18:19:57 | 2021-02-11T18:19:57 | {} | MEMBER |
ClickHouse | ClickHouse | 778,084,943 | 20,147 | alesapin | Something wrong with division? | 2021-02-12T09:34:21 | 2021-02-12T09:34:21 | {} | MEMBER |
ClickHouse | ClickHouse | 778,094,803 | 20,085 | gyuton | @vdimir , Hi!
Can you review this, please? | 2021-02-12T09:54:37 | 2021-02-12T09:54:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,192,946 | 20,364 | alesapin | Will check in branch | 2021-02-12T13:23:49 | 2021-02-12T13:23:49 | {} | MEMBER |
ClickHouse | ClickHouse | 778,289,036 | 20,224 | alesapin | Introduces new error https://clickhouse-test-reports.s3.yandex.net/20376/424f72bde1a036eda6e411a6a1d330ab7f35fe25/stress_test_(address)/stderr.log | 2021-02-12T16:13:50 | 2021-02-12T16:13:50 | {} | MEMBER |
ClickHouse | ClickHouse | 778,331,614 | 19,580 | alesapin | `test_materialize_mysql_database/test.py::test_clickhouse_killed_while_insert_5_7[clickhouse_node1]`
`test_materialize_mysql_database/test.py::test_clickhouse_killed_while_insert_8_0[clickhouse_node1]`
cc @sundy-li | 2021-02-12T17:26:55 | 2021-02-12T17:57:45 | {} | MEMBER |
ClickHouse | ClickHouse | 778,465,049 | 20,388 | SaltTan | Interesting
0 rows in set. Elapsed: 62.633 sec.
Received exception from server (version 21.2.2):
Code: 60. DB::Exception: Received from localhost:9000. DB::Exception: Table db.tabB doesn't exist.
```
2021.02.12 21:22:50.672543 [ 17530 ] {23997cf8-7906-429e-a890-2f79ca06a4e2} <Debug> executeQuery: (from [::1]:4... | 2021-02-12T21:28:55 | 2021-02-12T21:28:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,783,865 | 20,095 | sevirov | @akuzm Please, review this PR. | 2021-02-11T20:56:08 | 2021-02-11T20:56:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,820,700 | 14,614 | 4ertus2 | > Is this fixed in the newer versions?
Probably not. I'm a bit out of context. | 2021-02-11T22:00:08 | 2021-02-11T22:00:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,960,618 | 20,345 | azat | >AST fuzzer (debug) — Logical error: 'Trying to get name of not a column: SelectWithUnionQuery'.
#20350 | 2021-02-12T04:26:17 | 2021-02-12T04:26:17 | {} | MEMBER |
ClickHouse | ClickHouse | 778,283,494 | 20,375 | alesapin | Let's check in branch | 2021-02-12T16:04:30 | 2021-02-12T16:04:30 | {} | MEMBER |
ClickHouse | ClickHouse | 777,674,190 | 20,287 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/issues/20342 | 2021-02-11T17:49:41 | 2021-02-11T17:49:41 | {} | MEMBER |
ClickHouse | ClickHouse | 778,083,988 | 20,339 | alesapin | This fix is required for releases. Tests ok, `Thread leak failure` is known https://github.com/ClickHouse/ClickHouse/issues/20032. | 2021-02-12T09:32:25 | 2021-02-12T09:32:25 | {} | MEMBER |
ClickHouse | ClickHouse | 778,287,826 | 18,618 | akuzm | `RANGE` frame is the default now. | 2021-02-12T16:11:54 | 2021-02-12T16:11:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,521,877 | 20,149 | alexey-milovidov | @keen-wolf Let's continue? | 2021-02-11T14:51:47 | 2021-02-11T14:51:47 | {
"rocket": 1
} | MEMBER |
ClickHouse | ClickHouse | 777,691,052 | 19,719 | inkrement | Ok, thanks! In this case, I guess my JSON is broken.. Thanks anyways (I'll close the issue for now). | 2021-02-11T18:17:13 | 2021-02-11T18:17:13 | {} | NONE |
ClickHouse | ClickHouse | 777,709,539 | 20,303 | azat | >AST fuzzer (TSan) — Lost connection to server. See the logs.
```
2021-02-11 12:50:51 Program terminated with signal SIGKILL, Killed.
```
>Functional stateless tests (ANTLR debug) — fail: 1, passed: 80, skipped: 3
https://github.com/ClickHouse/ClickHouse/commit/0854dccfdef61a1b8247f37bc2c5f716ed834b2d
>I... | 2021-02-11T18:49:00 | 2021-02-11T18:49:00 | {} | MEMBER |
ClickHouse | ClickHouse | 778,070,891 | 20,343 | filimonov | Simple workaround - move setting into http parameters:
```python
prefix = f"http://{CLICKHOUSE_IP}:{CLICKHOUSE_PORT}/?format_schema={PROTO_SCHEMA}:{PROTO_MESSAGE}&query="
query = f"INSERT%20INTO%20arbitrary_db.some_entrypoint%20FORMAT%20Protobuf"
``` | 2021-02-12T09:05:48 | 2021-02-12T09:05:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,087,952 | 20,343 | joein | @filimonov I've tested workaround on some other data which were crushing before, and it also works.
Thank you!
| 2021-02-12T09:40:48 | 2021-02-12T09:40:48 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 778,179,373 | 20,346 | filimonov | Trace logs:
```
21.2
2021.02.12 09:34:25.568315 [ 1016105 ] {ddc64156-1527-4f91-b4f4-2ab9d6bd5b42} <Debug> db1.transactions (SelectExecutor): Selected 501 parts by partition key, 501 parts by primary key, 214406 marks by primary key, 214406 marks to read from 501 ranges
2021.02.12 09:34:25.576833 [ 1016105 ] {d... | 2021-02-12T12:56:06 | 2021-02-12T12:58:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,272,014 | 18,282 | abyss7 | Also discussed following details:
- Setting is per-query based, it's not a special server mode. Each distinct insert query may be async and may be not.
- If sync insert query matches any queue from async insert, the queue is left untouched - no forced flush.
- Because insert queries are distinguished on syntax-bas... | 2021-02-12T15:44:58 | 2021-02-12T15:44:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,576,141 | 19,317 | abyss7 | Integration test results will be fixed on infrastructure side soon | 2021-02-11T15:27:36 | 2021-02-11T15:27:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,192,986 | 20,365 | alesapin | Will check in branch | 2021-02-12T13:23:55 | 2021-02-12T13:23:55 | {} | MEMBER |
ClickHouse | ClickHouse | 778,307,822 | 7,462 | den-crane | still relevant 21.3.1.5944 | 2021-02-12T16:46:46 | 2021-02-12T16:46:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,314,318 | 3,852 | gersmann | > > An interesting question would be if it is possible to reverse that behaviour and also consider null (to blank out previously set column values).
>
> There is no such setting or aggregate function, maybe workaround using `ifNull(null, some_default_value)` will help to do what you need.
> And AFAIK this is SQL st... | 2021-02-12T16:57:48 | 2021-02-12T16:57:48 | {} | NONE |
ClickHouse | ClickHouse | 778,336,181 | 19,580 | alesapin | No related failures, going to merge today or tomorrow. | 2021-02-12T17:34:59 | 2021-02-12T17:34:59 | {} | MEMBER |
ClickHouse | ClickHouse | 777,626,375 | 10,177 | filimonov | Adding TTLs to tables created in old syntax is forbidden now: #20282 #17904 #19064
If you face that issue on the table with old syntax on the versions where it was not forbidden - most probably you will need to fix the metadata in zookeeper by editing zookeeper node with some external tool (zkCli.sh / go-zkcli / zoo... | 2021-02-11T16:37:23 | 2021-02-11T16:39:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,634,231 | 20,293 | akuzm | clang-tidy is happy at last. | 2021-02-11T16:48:58 | 2021-02-11T16:48:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,719,983 | 20,216 | fpacifici | Hi, thanks for the reply. This is appreciated.
I see the issue with backward compatibility of the schema. Assuming you are talking about that when mentioning the challenges around migrations. Out of curiosity, how do you handle breaking changes on the aggregation state structure on existing aggregating merge trees? ... | 2021-02-11T19:06:50 | 2021-02-11T19:06:50 | {} | NONE |
ClickHouse | ClickHouse | 778,012,030 | 20,087 | ervikrant06 | Thanks @Felixoid for your inupt. sorry for late response.
Basically this should be sufficient to drop the older metrics.. right?
~~~
clickhouse2-0.clickhouse-hs.ns-vaggarwal.svc.cluster.local :) select COUNT(*) from shard_01.graphite where Path Like 'CPUFrequencyMHz_24.AsynchronousMetrics
.ClickHouse.clickho... | 2021-02-12T06:46:35 | 2021-02-12T06:46:35 | {} | NONE |
ClickHouse | ClickHouse | 778,290,137 | 19,689 | alesapin | Introduced new error with fix and reverted. | 2021-02-12T16:15:50 | 2021-02-12T16:15:59 | {} | MEMBER |
ClickHouse | ClickHouse | 778,418,090 | 20,230 | tavplubix | I suggest another approach: #20448. Could you please take a look? | 2021-02-12T19:53:01 | 2021-02-12T19:53:01 | {} | MEMBER |
ClickHouse | ClickHouse | 778,469,949 | 20,452 | UnamedRus | There is a parseDateTimeBestEffort function, which covers multiple date string formats.
```
SELECT
toDate(parseDateTimeBestEffort('02/21/1986')),
toDate(parseDateTimeBestEffort('19/May/1999'))
Query id: 384372ae-5318-4eeb-bf49-97bb09b9a8c9
┌─toDate(parseDateTimeBestEffort('02/21/1986'))─┬─toDate(par... | 2021-02-12T21:40:24 | 2021-02-12T21:40:24 | {
"hooray": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,506,217 | 19,856 | alexey-milovidov | There is also the following case:
```
milovidov-desktop :) EXPLAIN SYNTAX SELECT * FROM system.numbers AS t1, system.numbers AS t2 WHERE t1.number + 1 = t2.number + 2
EXPLAIN SYNTAX
SELECT *
FROM system.numbers AS t1
, system.numbers AS t2
WHERE (t1.number + 1) = (t2.number + 2)
Query id: 259dbb1f-ee0c-49... | 2021-02-12T23:12:49 | 2021-02-12T23:12:49 | {} | MEMBER |
ClickHouse | ClickHouse | 778,079,381 | 20,332 | alesapin | ```
$ zgrep -a 'Fatal' clickhouse-server.log.gz
2021.02.11 16:04:24.761795 [ 259 ] {} <Fatal> BaseDaemon: (version 20.12.6.5986, build id: 275A8EB5228619A4) (from thread 722) Terminate called without an active exception
2021.02.11 16:04:24.776287 [ 26864 ] {} <Fatal> BaseDaemon: ####################################... | 2021-02-12T09:23:04 | 2021-02-12T09:23:04 | {} | MEMBER |
ClickHouse | ClickHouse | 778,286,157 | 20,378 | alesapin | Stress test (thread) — Killed by signal (in clickhouse-server.log)
not related, buy looks dangerous:
```
2021.02.12 17:30:27.695164 [ 33288 ] {b45470c4-aa30-4822-ada8-2ae4880ef2da} <Debug> executeQuery: (from [::1]:39214, using production parser) SELECT intDiv(number, 2) AS k, count(), argMax(toString(number), num... | 2021-02-12T16:08:59 | 2021-02-12T16:08:59 | {} | MEMBER |
ClickHouse | ClickHouse | 777,512,557 | 10,128 | filimonov | > What I observed was:
> lots of merges in system.merges with is_mutation=1
> mutation in system.mutations with lots of names in parts_to_do_names (now the column is empty)
> increased disk usage roughly the size of the table (but it is hard to be sure)
I've checked it and It seems something is not working ... | 2021-02-11T14:45:21 | 2021-02-11T14:45:21 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,597,878 | 19,257 | nikitamikhaylov | :(
```sql
jakalletti-build.sas.yp-c.yandex.net :) select c, d, polygonsIntersectionGeographic(readWktMultiPolygon(a), readWktMultiPolygon(b)) from (select first.p as a, second.p as b, first.name as c, second.name as d from multipolygons_wkt as first cross join multipolygons_wkt as second limit 100) format Null;
... | 2021-02-11T15:56:01 | 2021-02-11T15:59:57 | {} | MEMBER |
ClickHouse | ClickHouse | 777,672,041 | 20,305 | alexey-milovidov | > AST fuzzer (TSan) — SUMMARY: ThreadSanitizer: heap-use-after-free obj-x86_64-linux-gnu/../src/Storages/MergeTree/MergeTreeIndexFullText.cpp:647:34 in DB::SplitT
#19233 | 2021-02-11T17:46:23 | 2021-02-11T17:46:23 | {} | MEMBER |
ClickHouse | ClickHouse | 777,688,291 | 19,451 | azat | >AST fuzzer (MSan) — Received signal 11
#20344
>AST fuzzer (UBSan) — ../contrib/libcxx/include/vector:687:31: runtime error: applying non-zero offset 18446744073709551568 to null pointer
#20345 | 2021-02-11T18:12:45 | 2021-02-11T18:12:45 | {} | MEMBER |
ClickHouse | ClickHouse | 777,784,152 | 20,154 | sevirov | @akuzm Please, review this PR. | 2021-02-11T20:56:39 | 2021-02-11T20:56:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,066,831 | 20,330 | alesapin | 20.11 out of support | 2021-02-12T08:57:48 | 2021-02-12T08:57:48 | {} | MEMBER |
ClickHouse | ClickHouse | 777,586,497 | 20,209 | filimonov | See also https://github.com/ClickHouse/ClickHouse/issues/19570
To extract some info about part clickhouse need to do `stat` system call. Lot of calls = slow. | 2021-02-11T15:40:45 | 2021-02-11T15:40:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,728,926 | 17,623 | stigsb | > @melin Is there some standard way to convert XML to JSON?
> How to convert attributes? How to not lose the order of fields (convert everything to arrays)?
If your want it to be reversible you end up with a lot of arrays, just like in DOM. Elements may be repeated, attributes too. And of top of that there's namespace... | 2021-02-11T19:21:40 | 2021-02-11T19:21:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,071,265 | 20,346 | filimonov | Please try to collect extra info (details are in chat) | 2021-02-12T09:06:33 | 2021-02-12T09:06:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 778,204,649 | 18,508 | nikitamikhaylov | @depressed-pho For you to know we have some prototype of window functions at the moment. Your function could be rewritten as following, for example.
```sql
select * from (
select sum(_ut_2) over (order by _ut_1 ASC), _ut_2
from (
select untuple(arrayJoin([(begin, 1), (end, -1)]))
from running_c... | 2021-02-12T13:47:04 | 2021-02-12T13:47:04 | {} | MEMBER |
ClickHouse | ClickHouse | 778,504,577 | 20,449 | alexey-milovidov | Decided to merge ASAP. | 2021-02-12T23:07:24 | 2021-02-12T23:07:24 | {} | MEMBER |
ClickHouse | ClickHouse | 777,509,986 | 17,623 | alexey-milovidov | @nvartolomei @bharatnc
Your help will be very appreciated.
Currently no one from my team is working on "Pluggable auth with tokens" task and it's not scheduled yet.
I'm also out of context of this task - I'm not familiar with internal details of OAuth, JWT.
It will be perfect if you prepare a prototype PR or... | 2021-02-11T14:42:05 | 2021-02-11T14:44:20 | {} | MEMBER |
ClickHouse | ClickHouse | 777,608,789 | 20,192 | filimonov | One more note: there was a regression in 20.12 related to backoff for retries.
See
https://github.com/ClickHouse/ClickHouse/issues/20203
https://github.com/ClickHouse/ClickHouse/pull/20335
There are good chances that it will also fix your issue (or at least partially fix). I will retest that once the above PR ... | 2021-02-11T16:11:40 | 2021-02-11T16:12:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,614,743 | 20,166 | filimonov | Sounds like network timeout issue. Check send_timeout / receive_timeout in clickhouse, also something similar should exist on client side. | 2021-02-11T16:20:12 | 2021-02-11T16:20:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,748,513 | 14,614 | yusufozturk | We have the same issue. Is this fixed in the newer versions? @4ertus2
Thanks. | 2021-02-11T19:52:11 | 2021-02-11T19:52:11 | {} | NONE |
ClickHouse | ClickHouse | 778,080,096 | 20,148 | alesapin | All failures fixed in branch | 2021-02-12T09:24:27 | 2021-02-12T09:24:27 | {} | MEMBER |
ClickHouse | ClickHouse | 778,186,238 | 20,017 | alexey-milovidov | Duplicate #20046. | 2021-02-12T13:10:30 | 2021-02-12T13:10:30 | {} | MEMBER |
ClickHouse | ClickHouse | 778,261,480 | 20,244 | filimonov | One more example:
```
clickhouse-client --query="create table a (a String, b String) engine=Null"
printf '1\t' | clickhouse-client --query="INSERT INTO a FORMAT TSV"
```
```
Code: 33, e.displayText() = DB::ParsingException: Unexpected end of stream, while parsing value of Nullable type: (at row 1)
: Buffer has... | 2021-02-12T15:26:59 | 2021-02-12T15:26:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,581,436 | 20,244 | filimonov | Minimal example:
```
printf '1\t' | clickhouse-local --query="SELECT * FROM table" --structure='a String, b String'
```
20.12 and before - works, 21.1 and newer - fails. | 2021-02-11T15:34:09 | 2021-02-11T15:34:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,928,635 | 17,623 | melin | > > @melin Is there some standard way to convert XML to JSON?
> > How to convert attributes? How to not lose the order of fields (convert everything to arrays)?
>
> If your want it to be reversible you end up with a lot of arrays, just like in DOM. Elements may be repeated, attributes too. And of top of that there'... | 2021-02-12T02:32:17 | 2021-02-12T02:32:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,959,844 | 9,800 | bharatnc | @filimonov @alexey-milovidov I think this can be closed. All have `event_time_microseconds` and `query_time_microseconds` field except the following case:
I added another new table a month back [system.distributed_ddl_queue](https://github.com/ClickHouse/ClickHouse/blob/master/docs/en/operations/system-tables/distr... | 2021-02-12T04:23:57 | 2021-02-12T04:30:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,963,640 | 20,303 | azat | >AST fuzzer (TSan) — Lost connection to server. See the logs.
https://clickhouse-test-reports.s3.yandex.net/20303/3993ad6f01c6f2f3ffd6eafba9eaad30999f316d/fuzzer_tsan/report.html#fail1
```
2021-02-12 01:24:18 Program terminated with signal SIGKILL, Killed.
```
>Performance — 1 errors, 1 too long, 7 faster, ... | 2021-02-12T04:34:41 | 2021-02-12T04:36:14 | {} | MEMBER |
ClickHouse | ClickHouse | 778,052,731 | 20,087 | Felixoid | > But if I look at Time precious is still 60s not 10m
That's what graphite-ch-optimizer was created for. | 2021-02-12T08:26:59 | 2021-02-12T08:26:59 | {} | MEMBER |
ClickHouse | ClickHouse | 778,086,529 | 20,295 | alesapin | Out of support | 2021-02-12T09:37:40 | 2021-02-12T09:37:40 | {} | MEMBER |
ClickHouse | ClickHouse | 778,253,560 | 20,179 | vmihailenco | Does anyone know what memory usage should I expect from multi-part upload? I would expect CH to use 100-200MB buffer but it does not seem to be the case. Does it try to merge parts before uploading or something?
It looks like `min_multi_part_upload_size=0` helps but memory usage is still several gigabytes... | 2021-02-12T15:13:38 | 2021-02-12T15:13:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 782,199,591 | 20,971 | alexey-milovidov | Users are not supposed to manually edit columns.txt file. | 2021-02-19T16:52:36 | 2021-02-19T16:52:36 | {} | MEMBER |
ClickHouse | ClickHouse | 782,226,278 | 20,919 | alexey-milovidov | Some builds failed due to
> Build failed without build log. It's inner CI problem. Maybe something wrong with docker or host. Just wait for build restart. | 2021-02-19T17:37:43 | 2021-02-19T17:37:43 | {} | MEMBER |
ClickHouse | ClickHouse | 782,569,182 | 9,545 | alexey-milovidov | No, but it should be rather small, as we enable it in our perf tests. | 2021-02-20T06:13:07 | 2021-02-20T06:13:07 | {
"+1": 1
} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.