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 | 750,636,569 | 18,438 | l1t1 | where can i get the 20.13.1.5509 version? | 2020-12-24T00:29:42 | 2020-12-24T00:29:42 | {} | NONE |
ClickHouse | ClickHouse | 750,782,844 | 18,452 | hit-lacus | +1 | 2020-12-24T07:14:09 | 2020-12-24T07:14:09 | {} | NONE |
ClickHouse | ClickHouse | 750,814,318 | 18,438 | filimonov | Cc @traceon | 2020-12-24T09:03:03 | 2020-12-24T09:03:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,926,338 | 18,438 | traceon | This will be fixed by #17211 as a side effect. | 2020-12-24T16:52:35 | 2020-12-24T16:52:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,946,148 | 18,307 | ghuname | @den-crane At the moment I am trying to diagnose if we have a network problem.
IP address of node5 is 10.255.0.145.
In node5's log I can see network errors towards all other nodes except node6.
What confuses me is that in node5's log, I can find network errors related to node5's IP address as it is shown bellow:
... | 2020-12-24T19:01:01 | 2020-12-24T19:01:01 | {} | NONE |
ClickHouse | ClickHouse | 751,128,166 | 18,438 | traceon | @l1t1
<img width="992" alt="Screen Shot 2020-12-25 at 03 05 55" src="https://user-images.githubusercontent.com/10261534/103108137-51500180-465e-11eb-8a64-4dec24bd1b52.png">
| 2020-12-24T23:07:57 | 2020-12-24T23:07:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,309,494 | 18,283 | splichy | OK, that's pitty :/ | 2020-12-23T13:52:39 | 2020-12-23T13:52:39 | {} | NONE |
ClickHouse | ClickHouse | 750,487,273 | 17,292 | nikitamikhaylov | I grepped server's logs, there is no stacktrace from memory sanitizer assertion failure. | 2020-12-23T21:51:53 | 2020-12-23T21:51:53 | {} | MEMBER |
ClickHouse | ClickHouse | 750,638,800 | 14,297 | nikitamikhaylov | Thread leak in live view? Some thread doesn't join? @vzakaznikov
```
Stats: LargeMmapAllocator: allocated 1061217 times, remains 49 (130820 K) max 1240 M; by size logs: 17:43485; 18:91665; 19:178646; 20:734439; 21:6755; 22:5260; 23:824; 24:99; 25:44;
==================
WARNING: ThreadSanitizer: thread leak (pid=3... | 2020-12-24T00:31:24 | 2020-12-24T00:34:06 | {} | MEMBER |
ClickHouse | ClickHouse | 750,820,442 | 18,457 | amosbird | https://github.com/ClickHouse/ClickHouse/pull/17858 | 2020-12-24T09:22:13 | 2020-12-24T09:22:13 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 751,163,742 | 18,438 | alexey-milovidov | Here it is: https://clickhouse-builds.s3.yandex.net/0/7d8a307bb22efce5a24ab039aab5c47f82d9136f/clickhouse_build_check/report.html
Download the latest build: `clang-11 | relwithdebuginfo | none | bundled | unsplitted`
```
wget https://clickhouse-builds.s3.yandex.net/0/7d8a307bb22efce5a24ab039aab5c47f82d9136f/clic... | 2020-12-25T04:03:49 | 2020-12-25T04:03:49 | {} | MEMBER |
ClickHouse | ClickHouse | 750,321,961 | 17,838 | weeds085490 | > @weeds085490
> How this temporary inconsistency possible and how it is related to the query_log?
During the SQL execution of clickhouse, the input data is not based on the same snapshot. For the same filter, we may get two different data in the same SQL when other threads are inserting new data. Let‘s look at ex... | 2020-12-23T14:23:26 | 2020-12-23T14:23:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,354,055 | 18,241 | SaltTan | You don't have to duplicate in the base data if you use `expression`
```
<structure>
<key>
<attribute>
<name>network</name>
<type>String</type>
</attribute>
</key>
<attribute>
<n... | 2020-12-23T15:38:52 | 2020-12-23T15:38:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 749,500,422 | 18,340 | alesapin | Probably you are using compact parts or in-memory parts? | 2020-12-22T11:44:39 | 2020-12-23T20:07:54 | {} | MEMBER |
ClickHouse | ClickHouse | 750,731,554 | 18,448 | amosbird | > To run clickhouse in docker
Perhaps we can extend those functions so that they are aware of containers? | 2020-12-24T04:12:27 | 2020-12-24T04:12:27 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 750,791,181 | 13,707 | alex-zaitsev | Fixed by #16856 | 2020-12-24T07:41:36 | 2020-12-24T07:41:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,878,836 | 18,312 | alesapin | Failed to disable `-fno-inline` for libcxx, libcxx-abi and libunwind. Even when specified `-finline-functions -finline-hint-functions ` in corresponding files `target_compile_options(cxx PRIVATE -finline-functions -finline-hint-functions)` and so on. So exceptions unwinding extremely slow:
. And enable `always_fetch_merged_part`.
_Add setting always_fetch_merged_part which restrict replica to merge parts by itself and always prefer dowloading from other replicas. #10379 (alesapin)_ | 2020-12-24T16:17:15 | 2020-12-24T16:18:32 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,927,058 | 18,328 | sevirov | @KochetovNicolai Please, review this PR. | 2020-12-24T16:57:20 | 2020-12-24T16:57:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,936,248 | 18,326 | sevirov | @vitlibar Please, review this PR. | 2020-12-24T17:57:34 | 2020-12-24T17:57:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,986,320 | 15,988 | traceon | @vitlibar `ExpireLRUCache` seemingly updates the expire time on access. This is not what I need. It is poorly documented (even has bugs in the docs.) It is not used anywhere in ClickHouse. I will move the caching to the `ExternalAuthenticators` as per your request, but not going to use `ExpireLRUCache`, it is completel... | 2020-12-24T20:23:49 | 2020-12-24T21:52:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,132,860 | 18,475 | nikitamikhaylov | Omg, Arcadia :( | 2020-12-25T00:07:46 | 2020-12-25T00:07:46 | {} | MEMBER |
ClickHouse | ClickHouse | 751,141,577 | 18,481 | alexey-milovidov | Flaky test fixed in #18493. | 2020-12-25T01:24:35 | 2020-12-25T01:24:35 | {} | MEMBER |
ClickHouse | ClickHouse | 750,362,854 | 18,241 | vdimir | > You don't have to duplicate in the base data if you use `expression`
Seems now it doesn't work with `file` source (one of common usecase)
https://github.com/ClickHouse/ClickHouse/blob/51481c9c5fa5235d2afcfdc98470dfc3fdb86cae/src/Dictionaries/FileDictionarySource.cpp#L85
But it works with `clickhouse` source.
... | 2020-12-23T15:58:09 | 2020-12-23T15:58:09 | {} | MEMBER |
ClickHouse | ClickHouse | 750,370,222 | 18,241 | mzealey | (also a note; i don't believe there is any documentation around `<expression>`) | 2020-12-23T16:17:21 | 2020-12-23T16:17:21 | {} | NONE |
ClickHouse | ClickHouse | 750,490,171 | 18,247 | SaltTan | The default DDL cleaner parameters are:
```
<yandex>
<distributed_ddl>
<max_tasks_in_queue>1000</max_tasks_in_queue>
<task_max_lifetime>604800</task_max_lifetime>
<cleanup_delay_period>60</cleanup_delay_period>
</distributed_ddl>
</yandex>
``` | 2020-12-23T22:02:16 | 2020-12-23T22:02:16 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,041,002 | 7,649 | deripas | Hi!
Thank you very much, very useful!
In what release will this be seen? | 2020-12-24T21:08:10 | 2020-12-24T21:08:10 | {} | NONE |
ClickHouse | ClickHouse | 751,142,667 | 18,401 | akuzm | Looks like it's even going to work, until somebody decides to call `Logger::destroy()`.
The Poco logging framework is hard to use correctly, and we just have to replace it altogether, but maybe we can merge this change... But if there is some other way to solve your problem, I would prefer not to merge it, because ... | 2020-12-25T01:33:29 | 2020-12-25T01:33:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,454,815 | 18,306 | alesapin | `01584_distributed_buffer_cannot_find_column` -- not related to changes. | 2020-12-23T20:09:53 | 2020-12-23T20:09:53 | {} | MEMBER |
ClickHouse | ClickHouse | 750,484,021 | 18,406 | SaltTan | This will completely disable "read backoff"
```
<yandex>
<profiles>
<default>
<!-- Pay attention only to reads, that took at least this amount of time. If set to 0 - means backoff is disabled. -->
<read_backoff_min_latency_ms>0</read_backoff_min_latency_ms>
</default>
... | 2020-12-23T21:39:41 | 2020-12-23T21:39:41 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,748,311 | 18,448 | weeds085490 | > > To run clickhouse in docker
>
> Perhaps we can extend those functions so that they are aware of containers?
@amosbird
In my opinion, Obtaining resource information from environment variables is only a way to avoid getting information about the physical/virtual machine. There are others:
* From /proc (... | 2020-12-24T05:28:36 | 2020-12-24T05:29:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,817,379 | 18,456 | akuzm | Never mind, looks like I don't need this anymore. | 2020-12-24T09:13:07 | 2020-12-24T09:13:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,379,153 | 18,433 | alexey-milovidov | Yes, I want to add some options:
- explicitly enable or disable watchdog;
- enable autorestart;
- restart delay. | 2020-12-23T16:39:48 | 2020-12-23T16:39:48 | {} | MEMBER |
ClickHouse | ClickHouse | 750,809,349 | 18,443 | alexey-milovidov | @nikitamikhaylov I checked it and appeared I was wrong:
```
~/arc/arcadia/contrib/libs/poco$ grep -r POCO_CLICKHOUSE
Foundation/include/Poco/Config.h:#define POCO_CLICKHOUSE_PATCH 1
```
Nevertheless, Arcadia is using some version of Poco that is neither upstream neither the same as in ClickHouse. | 2020-12-24T08:44:54 | 2020-12-24T08:44:54 | {} | MEMBER |
ClickHouse | ClickHouse | 750,950,963 | 11,512 | webparadoxer | HttpCode:500 ; ;Code: 32, e.displayText() = DB::Exception: Attempt to read after eof: while executing 'FUNCTION CAST(_dummy_0 : 0, 'Nullable(Decimal(9, 2))' :: 1) -> CAST(_dummy_0, 'Nullable(Decimal(9, 2))') Nullable(Decimal(9, 2)) : 2' (version 20.12.4.5 (official build))
| 2020-12-24T19:34:33 | 2020-12-24T19:34:33 | {} | NONE |
ClickHouse | ClickHouse | 751,139,458 | 18,481 | alexey-milovidov | Fuzzer: #18492 | 2020-12-25T01:08:53 | 2020-12-25T01:08:53 | {} | MEMBER |
ClickHouse | ClickHouse | 750,717,106 | 18,397 | fuwhu | @den-crane
Got it, thanks a lot Denny.
Just created a new issue https://github.com/ClickHouse/clickhouse-jdbc/issues/530 | 2020-12-24T03:02:26 | 2020-12-24T03:02:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,757,500 | 18,448 | weeds085490 | > > > > > To run clickhouse in docker
> > > >
> > > >
> > > > Perhaps we can extend those functions so that they are aware of containers?
> > >
> > >
> > > @amosbird
> > > In my opinion, Obtaining resource information from environment variables is only a way to avoid getting information about the physical/vi... | 2020-12-24T05:58:16 | 2020-12-24T05:58:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,840,428 | 18,166 | alesapin | `01149_zookeeper_mutation_stuck_after_replace_partition` flaky | 2020-12-24T10:32:39 | 2020-12-24T10:32:39 | {} | MEMBER |
ClickHouse | ClickHouse | 751,150,909 | 18,500 | alexey-milovidov | The difference in virtual memory usage does not matter, only physical memory usage matters.
(possible reasons: we allocate memory in a different way, probably more efficiently, by reserving more virtual memory pages)
Another possible reason: #11139
If the difference in physical memory usage is in range of megaby... | 2020-12-25T02:28:23 | 2020-12-25T02:28:23 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 750,365,069 | 18,294 | abyss7 | Known problem with flaky mutation tests | 2020-12-23T16:03:55 | 2020-12-23T16:03:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,367,531 | 18,397 | den-crane | ClickHouse/clickhouse-jdbc#529
<img width="578" alt="Screen Shot 2020-12-23 at 12 09 26 PM" src="https://user-images.githubusercontent.com/19737682/103016049-d086ed80-4517-11eb-9610-5d45e205d585.png">
| 2020-12-23T16:10:20 | 2020-12-23T16:10:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,470,391 | 18,443 | azat | >Ok, let's guard displayText with POCO_CLICKHOUSE_PATCH then
I'm not sure that this will help, since seems that it also has some different Poco?
https://github.com/ClickHouse/ClickHouse/pull/17641#issuecomment-744542579
> Yandex synchronization check failed, because I didn't patch Poco in Arcadia :((
Does... | 2020-12-23T20:54:17 | 2020-12-23T20:54:17 | {} | MEMBER |
ClickHouse | ClickHouse | 750,473,294 | 18,443 | alexey-milovidov | Arcadia does not define `POCO_CLICKHOUSE_PATCH` and it's using something more close to the original Poco (also very old). | 2020-12-23T21:03:20 | 2020-12-23T21:03:30 | {} | MEMBER |
ClickHouse | ClickHouse | 750,476,264 | 18,437 | den-crane | Please grep logs for a history of 20201223_6410604_6410604_0 | 2020-12-23T21:13:10 | 2020-12-23T21:13:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,714,398 | 18,397 | fuwhu | > The issue is in JDBC
> Open an issue where https://github.com/ClickHouse/clickhouse-jdbc
>
> ```
> curl 'localhost:8123?query=explain+select+1+format+TabSeparatedWithNamesAndTypes'
> explain
> String
> Expression (Projection)
> Expression (Before ORDER BY and SELECT)
> SettingQuotaAndLimits (Set limit... | 2020-12-24T02:50:39 | 2020-12-24T02:50:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,812,887 | 12,744 | filimonov | > @den-crane
> hi!
> I tried to use KafkaEngine to consume the message from Kafka topic. The messages format is CSV, and the delimiter is `\1` (\u0001), a special character. We use it cause the message will not contain this special character. But when I execute the DDL , the '\1' will change to `\ \1` like this:
>
... | 2020-12-24T08:58:16 | 2020-12-24T08:58:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,316,608 | 18,397 | den-crane | The issue is in JDBC
Open an issue where https://github.com/ClickHouse/clickhouse-jdbc
```
curl 'localhost:8123?query=explain+select+1+format+TabSeparatedWithNamesAndTypes'
explain
String
Expression (Projection)
Expression (Before ORDER BY and SELECT)
SettingQuotaAndLimits (Set limits and quota after r... | 2020-12-23T14:10:18 | 2020-12-23T14:11:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,471,846 | 18,406 | matthieugouel | It seems to work like a charm with your advice. Many thanks, it helps me a lot. | 2020-12-23T20:58:45 | 2020-12-23T20:58:45 | {} | NONE |
ClickHouse | ClickHouse | 750,722,855 | 18,450 | awesomeleo | There is a related one https://github.com/ClickHouse/ClickHouse/issues/6556, but just a workaround. | 2020-12-24T03:28:02 | 2020-12-24T03:28:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,126,513 | 18,438 | l1t1 | > @l1t1 You can get it from testing repository: https://repo.yandex.ru/clickhouse/deb/testing/main/
> Or from master per-commit builds: https://github.com/ClickHouse/ClickHouse/commits/master
I opened the second link, but no builds link found there | 2020-12-24T22:44:06 | 2020-12-24T22:44:06 | {} | NONE |
ClickHouse | ClickHouse | 750,431,375 | 18,376 | azat | >It stopped to work.
Yes, and this is because of another strange thing that I found in debian manifests - https://github.com/ClickHouse/ClickHouse/blob/97ffc4032150990554a70f03f527d88117269b52/debian/rules#L8
It enables hardening and it should fail, but it will work only since v9 compat level, so they should be d... | 2020-12-23T19:00:35 | 2020-12-23T19:00:35 | {} | MEMBER |
ClickHouse | ClickHouse | 750,563,128 | 18,443 | nikitamikhaylov | But why it compiles successfully in my PR? | 2020-12-23T23:29:55 | 2020-12-23T23:29:55 | {} | MEMBER |
ClickHouse | ClickHouse | 750,702,551 | 12,542 | weiqxu | If we need this function, it looks implement 3 parts.
1. AggregateFunction need to be supported Nullable.
2. For outer JOIN, the no match row should return NULL , so the data type for the result should be Nullable(AggregateFunction)
3. implement cloneResize() for AggregateFunction
I tried to POC changed ,it wor... | 2020-12-24T01:58:49 | 2020-12-24T01:58:49 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,804,503 | 12,744 | BigRantLing | @den-crane
hi!
I tried to use KafkaEngine to consume the message from Kafka topic. The messages format is CSV, and the delimiter is `\1` (\u0001), a special character. We use it cause the message will not contain this special character. But when I execute the DDL , the '\1' will change to `\ \1` like this:
```sql
... | 2020-12-24T08:28:47 | 2020-12-24T08:28:47 | {} | NONE |
ClickHouse | ClickHouse | 750,826,679 | 18,340 | ghost | @alesapin
I will try to find some test-cases reproducing this exception and then share the test-cases.
Wish you a Marry Christmas. :)
| 2020-12-24T09:44:07 | 2020-12-24T10:41:09 | {} | NONE |
ClickHouse | ClickHouse | 750,321,695 | 17,872 | KochetovNicolai | https://github.com/ClickHouse/ClickHouse/blob/9be5fa9ef23ab8b82e79601daa5d2ecffc71667e/src/Interpreters/InterpreterInsertQuery.cpp#L280-L290
#14005
```
INSERT INTO test SELECT SUM(((((materialize(number) AS x) + materialize(x + 1)) + materialize(x + 2)) + materialize(x + 3)) + materialize(x + 4)) AS s
FROM numb... | 2020-12-23T14:22:47 | 2020-12-23T14:22:47 | {} | MEMBER |
ClickHouse | ClickHouse | 750,385,667 | 18,227 | millin | Same issue happened for me in 20.12 | 2020-12-23T16:53:45 | 2020-12-23T16:53:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,723,122 | 18,450 | awesomeleo | The table is created by the MaterializedMySQL engine, so we can not do the workaround. | 2020-12-24T03:29:09 | 2020-12-24T03:29:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,728,978 | 18,437 | amosbird | Sounds dangerous. Do you observe any issue with data integrity? | 2020-12-24T03:58:30 | 2020-12-24T03:58:30 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 750,766,536 | 18,450 | awesomeleo | I have fixed it locally. I will submit a patch asap. | 2020-12-24T06:26:27 | 2020-12-24T06:26:27 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,861,926 | 18,299 | alesapin | Manually cherry-picked fix for background remove of empty into 20.12 https://github.com/ClickHouse/ClickHouse/pull/17292 | 2020-12-24T11:56:36 | 2020-12-24T11:56:36 | {} | MEMBER |
ClickHouse | ClickHouse | 750,874,451 | 16,790 | Avogar | @olgarev Hi! I figured out how to resolve conflicts and merge these changes. So, now you can add fixes. | 2020-12-24T12:52:48 | 2020-12-24T12:52:48 | {} | MEMBER |
ClickHouse | ClickHouse | 750,879,544 | 18,312 | alesapin | When I've disabled `-fno-inline` globally everything speed up significantly.
`set (CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} -O0 -g3 -ggdb3 ${CMAKE_CXX_FLAGS_ADD}")` | 2020-12-24T13:12:34 | 2020-12-24T13:12:34 | {} | MEMBER |
ClickHouse | ClickHouse | 750,901,958 | 18,381 | nikitamikhaylov | Arcadia check is fixed in master and I hope this PR won't break it down again. | 2020-12-24T14:46:59 | 2020-12-24T14:46:59 | {} | MEMBER |
ClickHouse | ClickHouse | 750,919,619 | 18,453 | den-crane | Clickhouse does not support SQL scripts.
We does not support python jdbc driver. Ask in python jdbc repo. | 2020-12-24T16:12:46 | 2020-12-24T16:12:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,467,262 | 18,376 | alexey-milovidov | Hardening (when needed) should be enabled by global compile options in CMake, not via debian scripts. | 2020-12-23T20:45:42 | 2020-12-23T20:45:42 | {} | MEMBER |
ClickHouse | ClickHouse | 750,469,016 | 18,376 | azat | >Hardening (when needed) should be enabled by global compile options in CMake, not via debian scripts.
Indeed (I was very surprised when I saw `_FORTIFY_SOURCE`) | 2020-12-23T20:50:03 | 2020-12-23T20:50:03 | {} | MEMBER |
ClickHouse | ClickHouse | 750,706,719 | 18,449 | vladimir-golovchenko | Related bug: https://github.com/ClickHouse/ClickHouse/issues/18402 | 2020-12-24T02:17:15 | 2020-12-24T02:17:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,878,986 | 18,312 | alesapin | I think as daily job we can merge it as is (5 hours per day is not too much). And trying to optimise is different PR :( | 2020-12-24T13:10:19 | 2020-12-24T13:10:19 | {} | MEMBER |
ClickHouse | ClickHouse | 750,892,190 | 18,312 | alesapin | Istructions which looks like coverage collection
| 2020-12-24T14:05:06 | 2020-12-24T14:05:06 | {} | MEMBER |
ClickHouse | ClickHouse | 751,139,499 | 17,829 | alexey-milovidov | @CurtizJ One corner case has been found: #18492 | 2020-12-25T01:09:20 | 2020-12-25T01:09:20 | {} | MEMBER |
ClickHouse | ClickHouse | 750,327,977 | 5,837 | den-crane | https://github.com/ClickHouse/clickhouse-jdbc/releases
**clickhouse-jdbc-0.2.4-shaded.jar** --- 5.41 MB | 2020-12-23T14:37:57 | 2020-12-23T14:37:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,354,315 | 18,433 | amosbird | When running on k8s, I'd like to start clickhouse-server with tty attached so that the log is directly available. Perhaps we can add some option to force enable watchdog? | 2020-12-23T15:39:29 | 2020-12-23T15:39:40 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 750,709,009 | 18,421 | hexiaoting | version: 20.13.1.1
support floating point partition key.
I will implement this. | 2020-12-24T02:26:00 | 2020-12-24T02:26:00 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,710,329 | 18,310 | bohutang | hi,
#11996 is addressed MySQL engine enum type.
MaterializedMySQL enum is not releated to it. | 2020-12-24T02:31:59 | 2020-12-24T02:31:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,756,766 | 18,448 | weeds085490 | > > > > To run clickhouse in docker
> > >
> > >
> > > Perhaps we can extend those functions so that they are aware of containers?
> >
> >
> > @amosbird
> > In my opinion, Obtaining resource information from environment variables is only a way to avoid getting information about the physical/virtual machine. ... | 2020-12-24T05:55:44 | 2020-12-24T05:55:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,757,375 | 18,097 | akuzm | Do I get it right that the frame in Splunk ('window' in their terms) always ends with current row?
> a cumulative sum per ordered group (== streamstats sum by ...). I have seen avg too
> a cumulative count per ordered group (== streamstats count by )
This is just `count/sum/whatever over (partition by ...)`.
... | 2020-12-24T05:57:53 | 2020-12-24T05:57:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,790,897 | 18,403 | tobberguo | @BohuTANG @zhang2014 | 2020-12-24T07:40:32 | 2020-12-24T07:40:32 | {} | NONE |
ClickHouse | ClickHouse | 751,152,249 | 18,276 | fuwhu | @alexey-milovidov @amosbird @akuzm
Hi All,
could you please help review on this? or some other idea about this change ? | 2020-12-25T02:35:09 | 2020-12-25T02:35:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,404,141 | 18,436 | CurtizJ | BTW,` setCancellationHook` is also unused now. But do we need it in processors pipeline to allow early exit from aggregator? | 2020-12-23T17:41:45 | 2020-12-23T17:41:45 | {} | MEMBER |
ClickHouse | ClickHouse | 750,715,250 | 18,397 | den-crane | @fuwhu The issue is not related to CH server.
Use my fix for JDBC-driver ClickHouse/clickhouse-jdbc#529 | 2020-12-24T02:54:26 | 2020-12-24T02:54:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,793,408 | 5,225 | ddoneGit | I meet this question too.the real reason is that we had started one more clickhouse-server.
if you " ps -ef | grep clickhouse-server " and "service clickhouse-server status" you will find
the pid is different.so you just need kill the started pid and restart clickhouse-server
this is my solution
cat ./status
... | 2020-12-24T07:50:07 | 2020-12-24T07:50:07 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 750,812,505 | 18,436 | alexey-milovidov | @KochetovNicolai answered that the function is unused and can be removed. | 2020-12-24T08:56:39 | 2020-12-24T08:56:39 | {} | MEMBER |
ClickHouse | ClickHouse | 750,464,951 | 18,443 | alexey-milovidov | --Could you please also remove all entries of `POCO_CLICKHOUSE_PATCH` macro from code?--
Upd. No we cannot do it, because there is Arcadia build. | 2020-12-23T20:39:19 | 2020-12-23T20:41:09 | {} | MEMBER |
ClickHouse | ClickHouse | 750,481,853 | 18,227 | filimonov | We saw something similar.
zkCli at the same time was reporting smth like that:
```
2020-12-05 05:17:00,514 [myid:] - INFO [main-SendThread(127.0.0.1:2181):ClientCnxn$SendThread@852] - Socket connection established to 127.0.0.1/127.0.0.1:2181, initiating session
2020-12-05 05:17:00,516 [myid:] - INFO [main-Se... | 2020-12-23T21:31:50 | 2020-12-23T21:31:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,309,218 | 18,283 | splichy | PS: listBox is available on some pages, like: https://clickhouse.tech/docs/en/interfaces/
but when you click it will redirect you to https://clickhouse.tech/docs/en/ | 2020-12-23T13:51:58 | 2020-12-23T13:51:58 | {} | NONE |
ClickHouse | ClickHouse | 750,309,411 | 18,406 | den-crane | probably a duplicate of https://github.com/ClickHouse/ClickHouse/issues/18137
try to set a user parameter `read_backoff_min_latency_ms=10000000` and check results | 2020-12-23T13:52:25 | 2020-12-23T13:52:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,314,350 | 18,400 | den-crane | It's the expected behavior. CH does not deallocate memory immediately to save CPU on allocations. | 2020-12-23T14:04:28 | 2020-12-23T14:04:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,393,585 | 18,241 | SaltTan | > (also a note; i don't believe there is any documentation around <expression>)
There is https://clickhouse.tech/docs/en/sql-reference/dictionaries/external-dictionaries/external-dicts-dict-structure/#ext_dict_structure-attributes | 2020-12-23T17:13:23 | 2020-12-23T17:13:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,466,934 | 18,376 | alexey-milovidov | > export DEB_BUILD_MAINT_OPTIONS=hardening=+all,-pie
Let's remove it. BTW we disable `pie` explicitly. | 2020-12-23T20:44:44 | 2020-12-23T20:44:44 | {} | MEMBER |
ClickHouse | ClickHouse | 750,753,171 | 18,448 | amosbird | > > > To run clickhouse in docker
> >
> > Perhaps we can extend those functions so that they are aware of containers?
>
> @amosbird
> In my opinion, Obtaining resource information from environment variables is only a way to avoid getting information about the physical/virtual machine. There are others:
>
> * Fro... | 2020-12-24T05:43:31 | 2020-12-24T05:43:31 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 750,758,936 | 18,097 | akuzm | I think we'll have to make a test with some simplified popular examples of window functions, and also the CH way of emulating it.
It will help to see how much the usability improves, and how much the performance changes. Some CH features may be more convenient and fast than window functions, e.g. `LIMIT BY`. | 2020-12-24T06:02:21 | 2020-12-24T06:04:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,927,539 | 18,438 | alexey-milovidov | @l1t1 You can get it from testing repository: https://repo.yandex.ru/clickhouse/deb/testing/main/
Or from master per-commit builds: https://github.com/ClickHouse/ClickHouse/commits/master | 2020-12-24T17:00:06 | 2020-12-24T17:00:06 | {} | MEMBER |
ClickHouse | ClickHouse | 750,940,734 | 18,248 | sevirov | @Akazz Please, review this PR. | 2020-12-24T18:23:07 | 2020-12-24T18:23:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 750,950,898 | 7,690 | webparadoxer | HttpCode:500 ; ;Code: 32, e.displayText() = DB::Exception: Attempt to read after eof: while executing 'FUNCTION CAST(_dummy_0 : 0, 'Nullable(Decimal(9, 2))' :: 1) -> CAST(_dummy_0, 'Nullable(Decimal(9, 2))') Nullable(Decimal(9, 2)) : 2' (version 20.12.4.5 (official build))
| 2020-12-24T19:34:05 | 2020-12-24T19:34:05 | {} | NONE |
ClickHouse | ClickHouse | 750,953,071 | 8,295 | azat | @matwey there were lots of changes and right I don't see that this is an issue:
```
$ fgrep -r link_job_pool_src_ch_clickhouse_utils build.ninja CMakeFiles/rules.ninja
build.ninja: pool = link_job_pool_src_ch_clickhouse_utils
build.ninja: pool = link_job_pool_src_ch_clickhouse_utils
CMakeFiles/rules.ninja:pool... | 2020-12-24T19:50:43 | 2020-12-24T19:50:43 | {} | MEMBER |
ClickHouse | ClickHouse | 766,878,513 | 19,558 | alesapin | Don't affect integration tests. | 2021-01-25T15:03:45 | 2021-01-25T15:03:45 | {} | MEMBER |
ClickHouse | ClickHouse | 767,012,768 | 18,549 | alexey-milovidov | I will take this. | 2021-01-25T18:15:17 | 2021-01-25T18:15:17 | {} | MEMBER |
ClickHouse | ClickHouse | 767,029,108 | 15,395 | alexey-milovidov | Minimal example:
```
SELECT arrayMap(x -> arrayMap(x -> (x.1), [(1, 1), (2, 2), (3, 3)]), [(4, 4), (5, 5), (6, 6)])
┌─arrayMap(lambda(tuple(x), arrayMap(lambda(tuple(x), tupleElement(x, 1)), array(tuple(1, 1), tuple(2, 2), tuple(3, 3)))), array(tuple(4, 4), tuple(5, 5), tuple(6, 6)))─┐
│ [[1,2,3],[1,2,3],[1,2,3... | 2021-01-25T18:41:48 | 2021-01-25T18:41:48 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.