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 | 874,662,707 | 25,609 | zxc111 | @vdimir
The option is enable already.
Anything else I need to enable?
Maybe you push incorrect repository, https://github.com/zxc111/ClickHouse-1 is right and https://github.com/zxc111/ClickHouse is golang driver. 😭
 ... | 2021-07-06T11:17:22 | 2021-07-06T11:18:08 | {} | NONE |
ClickHouse | ClickHouse | 874,724,921 | 21,184 | tavplubix | The reason is that ClickHouse sends `ERR_Packet` with incorrect sequence number, so MySQL client does not recognize it as a part of `COM_QUERY_Response` and resets connection. It happens if exception is thrown during query execution:
```
mysql> select throwIf(number) from numbers(2) settings max_block_size=1;
ERROR ... | 2021-07-06T12:41:17 | 2021-07-06T12:41:17 | {} | MEMBER |
ClickHouse | ClickHouse | 874,949,865 | 25,917 | excitoon | Test fails without a fix: https://clickhouse-test-reports.s3.yandex.net/25917/0e5a1c9b0bff3e50edf4bf16568cc9b46cd8777f/fast_test/runlog.out.log | 2021-07-06T17:31:07 | 2021-07-06T17:31:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,987,302 | 26,020 | amosbird | > To fix it I changed the ordinary pipe to use the main plan / query when the projection doesn't use aggregation and that seems to be working fine.
We cannot use the original plan since filtering is done in projection and we have to apply it to the ordinary pipe too. | 2021-07-06T18:28:25 | 2021-07-06T18:28:25 | {
"+1": 2
} | COLLABORATOR |
ClickHouse | ClickHouse | 874,274,649 | 26,001 | alexey-milovidov | Does it work if you enable `allow_experimental_window_functions` globally? | 2021-07-05T18:32:06 | 2021-07-05T18:32:06 | {} | MEMBER |
ClickHouse | ClickHouse | 874,567,713 | 26,015 | tavplubix | > deleting data of zookeeper
Do I understand correctly that you have removed some data from ZooKeeper manually? Such error means that there are some tables in replicated database, but no metadata in ZooKeeper (it should never happen on normal operation).
Do you use debug build? Usually logical errors cause crash ... | 2021-07-06T08:29:37 | 2021-07-06T08:29:37 | {} | MEMBER |
ClickHouse | ClickHouse | 874,715,079 | 26,024 | depyronick | > can you share `select timezone()` from both servers?
Server on left has timezone config set to UTC on config.xml but the server on the right uses default configurations.
Server on left shows `UTC` and `Europe/Berlin` on right.
I've made a few more tests, it seems like the issue is that if I use `remote` func... | 2021-07-06T12:25:54 | 2021-07-06T12:25:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,731,425 | 25,895 | tavplubix | @Mergifyio update | 2021-07-06T12:51:02 | 2021-07-06T12:51:02 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 874,883,122 | 26,017 | alexey-milovidov | @DimasKovas Logical error is a simple exception in release build.
But in debug build it terminates the server (as failed assertion). | 2021-07-06T15:57:38 | 2021-07-06T15:57:38 | {} | MEMBER |
ClickHouse | ClickHouse | 875,036,196 | 18,794 | UnamedRus | It's also make sense to add check about too old kernel.
Clickhouse could have some hard to debug issues with randomly closed sockets on 3.10 kernel. | 2021-07-06T19:48:36 | 2021-07-06T19:48:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,080,547 | 26,029 | den-crane | Can you show the same for
```
select shopid, sum(qty), sum(qty)
from xxxxxx
where identity(goodsid) = '23576'
group by shopid
```
I think indexHint disables ind_goodsid bloom_filter index. Because in this case this index is bad. | 2021-07-06T21:04:43 | 2021-07-06T21:04:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,161,054 | 25,609 | zxc111 | thanks a lot.
I've rebased on you branch. | 2021-07-05T14:35:54 | 2021-07-05T14:35:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,338,966 | 21,435 | ilejn | Ubuntu 21.04 has glibc 2.33 and ClickHouse built in a plain way has number of symbols marked GLIBC_2.32 and GLIBC_2.33,
namely
```
objdump -T ./build/programs/clickhouse | egrep "GLIBC_2.32|GLIBC_2.33"
0000000000000000 DF *UND* 0000000000000000 GLIBC_2.33 mknodat
0000000000000000 DF *UND* 00000000000... | 2021-07-05T21:33:02 | 2021-07-05T21:33:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,667,192 | 26,020 | kitaisreal | @amosbird can you take a look at it ? | 2021-07-06T11:05:38 | 2021-07-06T11:05:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,695,481 | 25,589 | TszKitLo40 | @vdimir The distinct on test failed. | 2021-07-06T11:54:20 | 2021-07-06T11:54:20 | {
"eyes": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,709,953 | 26,024 | den-crane | can you share `select timezone()` from both servers? | 2021-07-06T12:17:42 | 2021-07-06T12:17:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,887,972 | 26,024 | den-crane | >I am trying to insert from one server with Berlin timezone to another with UTC, without changing any value...
It seems you insert datetime incorrectly using HTTP.
CH stores datetime in UTC only, no matter what.
TZ are applied when string are rendered or parsed.
`INSERT INTO table SELECT * FROM remote()` -- d... | 2021-07-06T16:03:32 | 2021-07-06T16:03:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,429,659 | 23,743 | alexey-milovidov | @vdimir @KochetovNicolai This refactoring introduced performance regression: https://github.com/ClickHouse/ClickHouse/issues/24075 | 2021-07-06T03:19:47 | 2021-07-06T03:19:54 | {} | MEMBER |
ClickHouse | ClickHouse | 874,536,965 | 24,148 | kssenii | `Yandex synchronization check (only for Yandex employees)` - `E Exception: 01889_sql_json_functions: [ FAIL ] - return code 48` - fixed in master.
`Integration tests` - not related.
`Functional stateless tests (release, wide parts enabled) — fail: ` - not related | 2021-07-06T07:43:23 | 2021-07-06T07:43:23 | {} | MEMBER |
ClickHouse | ClickHouse | 874,975,318 | 26,036 | kitaisreal | @tavplubix can you take a look at it ? | 2021-07-06T18:09:59 | 2021-07-06T18:09:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,190,660 | 25,990 | kitaisreal | @EgorKurito is there any change that these MATERIALIZED columns were added with ALTER command ?
Can you please provide server characteristics ?
How many records are in table 2 billions ?
Maybe query is not slow.
| 2021-07-05T15:22:22 | 2021-07-05T15:22:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,387,044 | 25,962 | alexey-milovidov | @mergifyio update | 2021-07-06T00:54:18 | 2021-07-06T00:54:18 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 875,083,113 | 26,025 | rsaphala | config.xml: https://gist.github.com/rsaphala/e7d40124498d75ef9f80bb7572bdbc1c
metrica.xml: https://gist.github.com/rsaphala/23c9ad7007912e2673d612668336d2a8
Yes I use both | 2021-07-06T21:09:04 | 2021-07-06T21:09:04 | {} | NONE |
ClickHouse | ClickHouse | 875,148,832 | 19,897 | BradSwain | @alexey-milovidov
Your understanding is still correct.
The `openlib.optimal` setting is designed to find all possible races by default, even in code that may potentially be called in parallel in the future. This leads to a high number of false positives being reported as well.
The tool is designed to be config... | 2021-07-06T23:26:03 | 2021-07-06T23:26:03 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 874,140,178 | 25,991 | kitaisreal | @LHertz for your specific query you can use something like:
```sql
SELECT arrayFilter(array -> length(array) > 0, arrayFlatten(extractAllGroups('ABCA', '(A)B|C(A)')))
┌─result────┐
│ ['A','A'] │
└───────────┘
``` | 2021-07-05T14:05:41 | 2021-07-05T14:05:41 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,247,261 | 26,000 | den-crane | 20.12.3 is obsolete, check with 21.5 or 21.6
and check result with parseDateTime64BestEffortOrNull
```
>= parseDateTime64BestEffortOrNull('2020-05-09T17:00:00.000Z', 3)
``` | 2021-07-05T17:18:35 | 2021-07-05T19:03:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,370,889 | 26,001 | vzakaznikov | The strange part is that it used to work so something got changed. | 2021-07-05T23:44:58 | 2021-07-05T23:44:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,429,499 | 24,075 | alexey-milovidov | This commit is slow: 08f10dc
This commit is good: 6f08f945e84f614e64d3cb2e36ea925c3267c508
| 2021-07-06T03:19:10 | 2021-07-06T03:19:10 | {} | MEMBER |
ClickHouse | ClickHouse | 874,681,082 | 25,589 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-07-06T11:28:09 | 2021-07-06T11:28:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,712,311 | 26,020 | Algunenano | Fixed with #26026. The problem wasn't that the materialization of the projection is in progress, but that it couldn't merge the result of the 2 pipes (the `ordinary_pipe` that reads the parts that aren't materialized yet, and the `projection_pipe` that reads projected parts) because the ordinary was using the projectio... | 2021-07-06T12:21:35 | 2021-07-06T12:21:35 | {} | MEMBER |
ClickHouse | ClickHouse | 875,050,531 | 26,037 | alexey-milovidov | Ok.
Workaround: hash both keys into single UInt64 key if the number of keys is small (e.g millions) and you can tolerate 1 / 2^64 potential collision probability for every key. | 2021-07-06T20:12:57 | 2021-07-06T20:12:57 | {} | MEMBER |
ClickHouse | ClickHouse | 875,159,962 | 25,700 | alexey-milovidov | You can try Query Profiler: https://clickhouse.tech/docs/en/operations/optimizing-performance/sampling-query-profiler/
It should be already enabled by default, just replace the query id in the [example](https://clickhouse.tech/docs/en/operations/optimizing-performance/sampling-query-profiler/#example). | 2021-07-06T23:57:10 | 2021-07-06T23:57:24 | {} | MEMBER |
ClickHouse | ClickHouse | 874,137,024 | 24,148 | l1tsolaiki | > @l1tsolaiki will this feature be compatible with materialized views?
As far as I understand, it should work with materialized views. | 2021-07-05T14:01:37 | 2021-07-05T14:01:37 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,413,605 | 24,075 | alexey-milovidov | Version 21.5 is Ok. | 2021-07-06T02:24:36 | 2021-07-06T02:24:36 | {} | MEMBER |
ClickHouse | ClickHouse | 874,414,847 | 24,832 | alexey-milovidov | @mergifyio update | 2021-07-06T02:28:58 | 2021-07-06T02:28:58 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 875,065,342 | 23,503 | Algunenano | I can't reopen the issue myself. Can one of the maintainers do it, please? Otherwise, should I open a different ticket instead? | 2021-07-06T20:38:07 | 2021-07-06T20:38:07 | {} | MEMBER |
ClickHouse | ClickHouse | 875,117,570 | 25,917 | kitaisreal | @excitoon can you please check 01903_correct_block_size_prediction_with_default fail with Functional stateless tests (thread) . | 2021-07-06T22:13:53 | 2021-07-06T22:13:53 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,161,766 | 26,033 | alexey-milovidov | What happens if you run `curl 'http://localhost:8123/' -d 'SHOW DATABASES'`?
Could you please paste server logs corresponding to this query? | 2021-07-07T00:01:52 | 2021-07-07T00:01:52 | {} | MEMBER |
ClickHouse | ClickHouse | 874,153,421 | 25,994 | electrical | @kirillDanshin we haven't done any `OPTIMIZE FINAL` yet anywhere.
After digging into this further we might have found the issue ( I hope )
Although the docker container shows with ulimit to be set to `1048576`, the `/proc/sys/fs/file-max` file shows a maximum of `65536` which i'm guessing we are hitting.
Need to s... | 2021-07-05T14:23:48 | 2021-07-05T14:23:48 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 874,551,584 | 25,994 | electrical | @den-crane thank you for that! I'll have a look at that. since our metrics are collected into prometheus it doesn't make much sense to keep them long in clickhouse it self indeed :-) | 2021-07-06T08:06:15 | 2021-07-06T08:06:15 | {} | NONE |
ClickHouse | ClickHouse | 874,552,710 | 26,014 | kitaisreal | @alexey-milovidov unstable performance tests are unrelated, fix for disabling JIT for aggregate functions in perf tests is already in master. | 2021-07-06T08:07:58 | 2021-07-06T08:07:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,588,371 | 26,015 | zouyonghao | > > deleting data of zookeeper
>
> Do I understand correctly that you have removed some data from ZooKeeper manually? Such error means that there are some tables in replicated database, but no metadata in ZooKeeper (it should never happen on normal operation).
I stop the zookeeper, delete its data and then restar... | 2021-07-06T09:00:45 | 2021-07-06T09:00:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,613,954 | 26,015 | tavplubix | So everything works as expected. Maybe we should change error code from `LOGICAL_ERROR` to something else. | 2021-07-06T09:37:20 | 2021-07-06T09:37:20 | {} | MEMBER |
ClickHouse | ClickHouse | 874,680,679 | 25,589 | vdimir | @mergifyio update | 2021-07-06T11:27:22 | 2021-07-06T11:27:22 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 874,867,046 | 19,897 | alexey-milovidov | @BradSwain Thank you!
The report is not clear to me. For example,
it said that if `finalize()` and `nextImpl()` called concurrently the trouble may happen. This is true - these met... | 2021-07-06T15:37:22 | 2021-07-06T15:38:26 | {} | MEMBER |
ClickHouse | ClickHouse | 874,885,978 | 25,994 | alexey-milovidov | @electrical You have too low system-wide limit on the number of opened files:
```
EMFILE The per-process limit on the number of open file descriptors has been reached (see the description of RLIMIT_NOFILE in getrlimit(2)).
ENFILE The system-wide limit on the total number of open files has been reac... | 2021-07-06T16:01:02 | 2021-07-06T16:01:02 | {} | MEMBER |
ClickHouse | ClickHouse | 875,041,030 | 19,897 | alexey-milovidov | Ok. 111 cases are not too many, it is possible to check all of them... | 2021-07-06T19:57:06 | 2021-07-06T19:57:06 | {} | MEMBER |
ClickHouse | ClickHouse | 875,077,274 | 26,035 | den-crane | >For example, I modified the configuration file of clickhouse, how can it be hot-loaded without restarting the clickhouse service?
Some parameters are applied automatically, no actions required other parameters require full restart `service clickhouse-server restart` .
```
Settings & restart
All users setti... | 2021-07-06T20:59:14 | 2021-07-06T20:59:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,108,952 | 26,025 | den-crane | You don't need to use `/etc/metrica.xml` https://kb.altinity.com/altinity-kb-setup-and-maintenance/altinity-kb-server-config-files
Just copy this xml file into `/etc/clickhouse-server/config.d/`
this is exactly https://github.com/ClickHouse/ClickHouse/issues/21905#issuecomment-803456597
| 2021-07-06T21:58:16 | 2021-07-06T21:58:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,235,835 | 26,001 | vzakaznikov | @akuzm, can you take a look? | 2021-07-05T16:53:02 | 2021-07-05T16:53:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,425,891 | 25,983 | shadowDy | > ```
> cat /etc/clickhouse-server/users.d/read_backoff_min_latency_ms.xml
> <?xml version="1.0"?>
> <yandex>
> <profiles>
> <default>
> <read_backoff_min_latency_ms>0</read_backoff_min_latency_ms>
> </default>
> </profiles>
> </yandex>
> ```
thanks，This configuration se... | 2021-07-06T03:07:37 | 2021-07-06T03:07:37 | {} | NONE |
ClickHouse | ClickHouse | 874,657,110 | 26,022 | alesapin | Fast test build ok | 2021-07-06T10:47:47 | 2021-07-06T10:47:47 | {} | MEMBER |
ClickHouse | ClickHouse | 874,370,609 | 26,001 | vzakaznikov | > Does it work if you enable `allow_experimental_window_functions` globally?
Yes, if I add the setting to users.xml.
```xml
<yandex>
<profiles>
<default>
<allow_experimental_window_functions>1</allow_experimental_window_functions>
</default>
</profiles>
</yandex>
``` | 2021-07-05T23:43:33 | 2021-07-05T23:43:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,485,230 | 26,000 | Kev1ntan | @den-crane thanks, working properly with 21.6 | 2021-07-06T06:04:06 | 2021-07-06T06:04:06 | {} | NONE |
ClickHouse | ClickHouse | 874,549,833 | 25,999 | vpanfilov | Seems to be related to https://github.com/ClickHouse/ClickHouse/issues/23444 | 2021-07-06T08:03:40 | 2021-07-06T08:03:40 | {} | NONE |
ClickHouse | ClickHouse | 874,626,537 | 25,609 | vdimir | @zxc111
Oh, I broke build with my comment. I'll fix it, please, let me push into this PR not to ask you to rebase branch every time.
Also I looked to `hex` function always adds leading zero (like `hex(1) = '01'`), so we can add leading zeros to `bin` too, to be consistent. I'll make it too. | 2021-07-06T09:57:43 | 2021-07-06T10:00:22 | {} | MEMBER |
ClickHouse | ClickHouse | 874,731,627 | 25,895 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-07-06T12:51:18 | 2021-07-06T12:51:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,052,736 | 18,794 | UnamedRus | > Just check for some minimal version?
Yes. (at least for start)
> What threshold do you suggest?
Dunno,
I faced an issues with sockets and 3.10, i heard that was some bug in kernel + ext4, which often lead to filesystem corruption in versions pre 3.17.
@den-crane said that it's better to start from 3.17+
... | 2021-07-06T20:16:42 | 2021-07-06T20:16:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,160,191 | 25,700 | alexey-milovidov | If you have any difficulties, upgrade ClickHouse to version 21.6 (latest) and check again. | 2021-07-06T23:57:48 | 2021-07-06T23:57:48 | {} | MEMBER |
ClickHouse | ClickHouse | 874,174,490 | 25,988 | tobiaslins | @kitaisreal amazing! thank you for the fast responses and solution. I just upgraded to 21.6 and it seems to work fine now :) | 2021-07-05T14:58:28 | 2021-07-05T14:58:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,199,511 | 25,990 | kitaisreal | @EgorKurito could you please provide query_log, query_thread_log, trace_log data for queries that we try to optimise ? | 2021-07-05T15:38:17 | 2021-07-05T15:39:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,387,122 | 25,962 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-07-06T00:54:35 | 2021-07-06T00:54:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,395,543 | 24,075 | alexey-milovidov | It is somehow related to query parallelism. With max_threads = 1 there is no performance difference.
New version is using 100% CPU (single core) for some reason.
Old version is using 1400% CPU on 16 core machine. | 2021-07-06T01:24:01 | 2021-07-06T01:24:01 | {} | MEMBER |
ClickHouse | ClickHouse | 874,415,022 | 24,832 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-07-06T02:29:37 | 2021-07-06T02:29:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,415,917 | 24,075 | alexey-milovidov | Maybe #23743 or #24574. | 2021-07-06T02:32:56 | 2021-07-06T02:32:56 | {} | MEMBER |
ClickHouse | ClickHouse | 874,652,588 | 26,022 | alesapin | Waiting for fast test only. | 2021-07-06T10:39:50 | 2021-07-06T10:39:50 | {} | MEMBER |
ClickHouse | ClickHouse | 874,709,610 | 26,019 | den-crane | Yes, there are plans to implement grace-join and merge-join in the future. | 2021-07-06T12:17:07 | 2021-07-06T12:17:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,074,674 | 26,025 | den-crane | @rsaphala do you use metrica.xml or incl ?
https://github.com/ClickHouse/ClickHouse/issues/21905#issuecomment-803456597
https://github.com/ClickHouse/ClickHouse/pull/25083/files
Can you share your config files? | 2021-07-06T20:54:49 | 2021-07-06T20:54:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,160,782 | 25,700 | alexey-milovidov | `create_table_query` requires opening and reading .sql metadata files.
It's expected to be slow if you have slow or overloaded IO.
20 rows per second does not look normal anyhow but it is possible with slow HDD or EBS. | 2021-07-06T23:59:29 | 2021-07-06T23:59:41 | {} | MEMBER |
ClickHouse | ClickHouse | 874,663,549 | 25,884 | alesapin | @mergifyio update | 2021-07-06T10:59:44 | 2021-07-06T10:59:44 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 875,039,932 | 18,794 | alexey-milovidov | Yes, should be easy to check.
Just check for some minimal version? What threshold do you suggest?
Or a list of version ranges with explanation of known problems? | 2021-07-06T19:55:14 | 2021-07-06T19:55:29 | {} | MEMBER |
ClickHouse | ClickHouse | 875,046,511 | 19,897 | alexey-milovidov | @BradSwain
I don't understand how it is supposed to work.
Currently looks like it assumes that every function can work concurrently with every another.
To prove with static analysis that something is synchronized, it requires very powerful inference, equivalent to solving the halting problem (in extreme case).
... | 2021-07-06T20:06:20 | 2021-07-06T20:06:20 | {} | MEMBER |
ClickHouse | ClickHouse | 874,122,399 | 25,994 | electrical | @den-crane
First query: `46886`
Second query:
```
24120 metric_log
7992 own_table1
7315 own_table2
3658 query_log
920 part_log
340 trace_log
256 own_table3
224 own_table4
198 own_table5
185 own_table6
```
I had to replace the names of our own tables as you can understand. | 2021-07-05T13:39:40 | 2021-07-05T13:39:40 | {} | NONE |
ClickHouse | ClickHouse | 874,158,460 | 25,994 | kitaisreal | @electrical ok, if there will be any additional issues feel free to post them here. | 2021-07-05T14:31:54 | 2021-07-05T14:31:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,194,471 | 25,805 | gyuton | @kitaisreal Hi! Can you review this, please? | 2021-07-05T15:28:50 | 2021-07-05T15:28:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,247,949 | 25,994 | den-crane | BTW @electrical you can disable metric_log/query_log or configure TTL
https://kb.altinity.com/altinity-kb-setup-and-maintenance/altinity-kb-system-tables-eat-my-disk | 2021-07-05T17:20:09 | 2021-07-05T17:20:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,398,262 | 24,075 | alexey-milovidov | Minimal query to reproduce:
```
SELECT * FROM
(SELECT 1 AS k) t1
LEFT JOIN
(SELECT 1 AS k FROM numbers_mt(10000000000) WHERE number = 1) t2
USING k
```
The right hand side of JOIN does not get required number of threads. | 2021-07-06T01:34:37 | 2021-07-06T01:35:00 | {
"heart": 1
} | MEMBER |
ClickHouse | ClickHouse | 874,531,366 | 25,798 | azat | >And AFAICS this test (test_distributed_queries_stress) hadn't passed, and it should be marked as FAIL?
Checked in test_results.tsv, and according to it the test marked correctly as FLAKY (I see retries there). | 2021-07-06T07:34:02 | 2021-07-06T07:34:02 | {} | MEMBER |
ClickHouse | ClickHouse | 874,651,784 | 26,008 | kitaisreal | Data type Map was experimental at 21.6. toJSONString function was not introduced. | 2021-07-06T10:38:26 | 2021-07-06T10:38:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,702,374 | 25,585 | SargtLa | Hello! Sorry, renamed tables for this issue, because real name does not matter here. I have no problem with inserting the data and get no errors about naming. The insert queries are all the same and the name of the table does not matter. The issue is that some data were flushed to the main table from buffer and some da... | 2021-07-06T12:05:58 | 2021-07-06T12:05:58 | {} | NONE |
ClickHouse | ClickHouse | 874,739,088 | 26,024 | depyronick | So, i've tried using `--use_client_time_zone` setting with `clickhouse-client` it seems it now inserts what it reads.
Is this the correct way to do that? I am trying to insert from one server with Berlin timezone to another with UTC, without changing any value... | 2021-07-06T13:01:35 | 2021-07-06T13:01:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,136,860 | 25,994 | kitaisreal | @electrical it does not seem to be a lot of columns parts.
Can you please share logs of queries ?
Maybe you use OPTIMIZE FINAL a lot ?
Can you share table configuration ? | 2021-07-05T14:01:24 | 2021-07-05T14:01:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,328,658 | 25,800 | traceon | Sounds reasonable.
You mean block users from authenticating, if no roles were **discovered** for them, right?
Because there can be roles that are assigned unconditionally in the config - these should not be taken into account, when the authentication decision is made (but still assigned, if user is authenticated si... | 2021-07-05T21:06:40 | 2021-07-05T21:28:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,707,923 | 25,995 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-07-06T12:14:34 | 2021-07-06T12:14:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 863,039,660 | 21,320 | ilejn | `HashJoin::RightTableData::maps` is now `std::vector<MapsVariant>`, not just a single `MapsVariant`. All `MapsVariant` have the same type, it is possible to have several maps for e.g. `Type::String` or for `Type::UInt32`, while not together.
If types are different everything is converted to `Type::hashed`.
To preve... | 2021-06-17T08:25:21 | 2021-07-06T13:35:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,932,568 | 19,897 | BradSwain | Your understanding is correct.
> But how it proves that they can be called concurrently in our code? It does not look true.
When the `openlib.optimal` setting is used the tool searches for function pairs that could potentially be called in parallel, but the tool cannot prove that these functions will be called i... | 2021-07-06T17:04:57 | 2021-07-06T17:09:15 | {} | NONE |
ClickHouse | ClickHouse | 875,071,468 | 18,794 | den-crane | Unfortunately I was talking about 4.16.4
_EXT4-fs error (device sda2): ext4_validate_inode_bitmap:96: comm AsyncBlockInput: Corrupt inode bitmap - block_group = 38449, inode_bitmap = 1259864081_
_Обновления ядра с 4.16 на 4.17 на Debian из backports решило вопрос, пофиксили в 4.16.4 (https://cdn.kernel.org/pub/li... | 2021-07-06T20:49:20 | 2021-07-06T20:49:20 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,651,879 | 26,006 | kitaisreal | Data type Map was experimental at 21.6. toJSONString function was not introduced. | 2021-07-06T10:38:36 | 2021-07-06T10:38:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 874,706,838 | 25,995 | alesapin | Too many failures, let's check one more time.
@Mergifyio update | 2021-07-06T12:12:56 | 2021-07-06T12:12:56 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 874,724,814 | 26,014 | CurtizJ | Arcadia sync is broken by this PR. | 2021-07-06T12:41:07 | 2021-07-06T12:41:07 | {} | MEMBER |
ClickHouse | ClickHouse | 874,777,223 | 19,897 | BradSwain | Hi @alexey-milovidov
Here is a link to another report for clickhouse: http://cloud.coderrect.com/6ed934f377614be290e4a71c7999fed9/index.html
This report was generated with some special configurations set.
Instead of starting from main (which can potentially miss code paths), coderrect was configured to search... | 2021-07-06T13:49:53 | 2021-07-06T13:49:53 | {} | NONE |
ClickHouse | ClickHouse | 874,380,228 | 25,999 | l1t1 | what version do you use? if your version is newer than 21.4, try version 21.4
https://github.com/ClickHouse/ClickHouse/issues/24075 | 2021-07-06T00:26:43 | 2021-07-06T00:26:43 | {} | NONE |
ClickHouse | ClickHouse | 874,414,780 | 24,832 | alexey-milovidov | > Maybe then we should use max_distributed_connections as upper limit only
Yes, it is intended to work this way. | 2021-07-06T02:28:44 | 2021-07-06T02:28:44 | {} | MEMBER |
ClickHouse | ClickHouse | 871,086,216 | 25,733 | compasses | Hi, @KochetovNicolai do you have any idea or any clue? | 2021-06-30T04:27:36 | 2021-06-30T04:27:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 871,212,377 | 25,820 | alesapin | One more time | 2021-06-30T08:44:23 | 2021-06-30T08:44:23 | {} | MEMBER |
ClickHouse | ClickHouse | 870,810,935 | 25,823 | azat | >I hope it helps. But docker-compose runs iptables commands as well
@qoega I don't see that it runs iptables directly, only via docker, which AFAICS uses --wait already:
https://github.com/docker/docker-ce/blob/acac13f511d3c907fe0b6cd084cae84a5b389666/components/engine/libnetwork/iptables/iptables.go#L537 | 2021-06-29T18:13:56 | 2021-06-29T18:13:56 | {} | MEMBER |
ClickHouse | ClickHouse | 870,818,144 | 25,825 | alexey-milovidov | I've got:
```
Unknown table engine s3. Maybe you meant: ['S3']
```
Which answers the question. | 2021-06-29T18:25:27 | 2021-06-29T18:25:27 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.