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 | 829,526,243 | 23,625 | filimonov | If the client just drops the connection w/o waition for getting the whole responce you will can that effect. Check client side timeouts, and client code. | 2021-04-29T19:26:28 | 2021-04-29T19:26:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,529,361 | 8,912 | FArthur-cmd | Issue is taken. Will be fixed soon | 2021-04-29T19:31:46 | 2021-04-29T19:31:46 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,591,141 | 23,660 | filimonov | You have some bad xml syntax in the file mentioned in the error message. | 2021-04-29T20:57:30 | 2021-04-29T20:57:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,848,207 | 23,781 | filimonov | I think we should mark it as bug fix and backport (it is regressions for users) | 2021-04-30T05:27:20 | 2021-04-30T05:27:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,901,721 | 23,756 | keshy7 | This is very helpful! Thank you so much @filimonov! | 2021-04-30T07:30:15 | 2021-04-30T07:30:15 | {} | NONE |
ClickHouse | ClickHouse | 829,989,335 | 23,791 | alexey-milovidov | Don't use FORMAT Null to test the performance on synthetic benchmarks. | 2021-04-30T10:07:25 | 2021-04-30T10:07:52 | {} | MEMBER |
ClickHouse | ClickHouse | 830,046,903 | 23,794 | l1t1 | this sql can run.
with t1 as (select number id ,row_number()over(order by id) rn from (select * from system.numbers limit 3)),
t2 as (select number+1 id ,row_number()over(order by id) rn from (select * from system.numbers limit 4)),
t3 as (select number+2 id ,row_number()over(order by id) rn from (select * from sys... | 2021-04-30T12:04:16 | 2021-04-30T12:04:16 | {} | NONE |
ClickHouse | ClickHouse | 829,107,679 | 23,736 | asipincasp | Andrey, we are update from 21.2.5.5 to 21.4.5.46 an monday.
### schema
```
SHOW CREATE TABLE clicks
Query id: df0329f7-4272-4d5d-8cce-e5864f22aeaa
CREATE TABLE pcs_stats.clicks
(
`event_uuid` FixedString(60),
`group_id` UInt32,
`platform` String,
`os_name` String,
`os_version` UInt8,
`device` String,
... | 2021-04-29T10:10:12 | 2021-04-30T12:38:12 | {} | NONE |
ClickHouse | ClickHouse | 829,305,125 | 23,768 | den-crane | not a bug.
```
SELECT toFloat32('1619675611')
┌─toFloat32('1619675611')─┐
│ 1619675600 │
└─────────────────────────┘
``` | 2021-04-29T14:52:49 | 2021-04-29T14:52:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,458,827 | 23,746 | alexey-milovidov | A small typo:
```
2021-04-29 19:23:32 /ClickHouse/base/pcg-random/pcg_random.hpp:1646:18: error: use of undeclared identifier 'pgc_detail'; did you mean 'pcg_detail'?
2021-04-29 19:23:32 using ext_std8 = pgc_detail::extended<table_pow2, advance_pow2, BaseRNG,
2021-04-29 19:23:32 ^~~~~~~~~~
2021-04... | 2021-04-29T17:39:57 | 2021-04-29T17:39:57 | {} | MEMBER |
ClickHouse | ClickHouse | 829,573,999 | 23,764 | filimonov | @ucasFL can you fix it? | 2021-04-29T20:28:50 | 2021-04-29T20:28:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,990,436 | 23,791 | alexey-milovidov | @UnamedRus All issues where you test performance that way will be closed immediately. | 2021-04-30T10:09:19 | 2021-04-30T10:09:19 | {} | MEMBER |
ClickHouse | ClickHouse | 829,995,632 | 23,746 | alexey-milovidov | @bnaecker I heard that Illumos has Linux compatibility layer and can run Linux binaries, https://www.youtube.com/watch?v=TrfD3pC0VSs is that true? | 2021-04-30T10:18:32 | 2021-04-30T10:18:32 | {} | MEMBER |
ClickHouse | ClickHouse | 830,138,487 | 21,414 | vdimir | @amosbird `01414_push_predicate_when_contains_with_clause` failed :(
https://clickhouse-test-reports.s3.yandex.net/21414/51444be49f194ecd3d115d5527069862a8445d3f/fast_test/test_log.txt
<details>
<summary>Error</summary>
```
2021-04-29 17:23:54 01523_interval_operator_support_string_literal: ... | 2021-04-30T14:35:26 | 2021-04-30T14:35:26 | {} | MEMBER |
ClickHouse | ClickHouse | 830,142,726 | 23,806 | UnamedRus | https://github.com/ClickHouse/ClickHouse/issues/19627 | 2021-04-30T14:42:17 | 2021-04-30T14:42:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,071,197 | 19,970 | r-zenine | Hello everyone,
I am currently considering adding a support for [Lucene](https://lucene.apache.org/core/) or [Tantivy](https://github.com/tantivy-search/tantivy). I am early stage in the process and I just started looking into it.
My current approach is to build an external server that would wrap one of the two... | 2021-04-29T09:12:05 | 2021-04-29T18:17:41 | {
"+1": 6
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,535,522 | 20,955 | nikitamikhaylov | 1) Now the status is written in Zookeeper node (`task_path` + `/status`) in JSON format. Here is an example
```json
{"hits":{"all_partitions_count":5,"processed_partitions_count":5}}
```
where hits is a table task name from config. And `processed_partitions_count` is updated during the process of copying.
Final s... | 2021-04-29T19:41:23 | 2021-04-29T19:41:23 | {} | MEMBER |
ClickHouse | ClickHouse | 829,790,299 | 8,338 | olgarev | @alexey-milovidov Documented, please remove the `comp-documentation` tag.
| 2021-04-30T04:07:11 | 2021-04-30T04:07:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,969,182 | 23,785 | ronhuizer | Right, in this case I think I found the cause.
The lat and lon fields are floats, and I'm essentially querying them with a TinyInt, 4 instead of 4.0.
When I use a float in my query (4.0, 8.0 etc), the results are consistent between the 2 tables.
I still think its weird query results differ between the 2 table conf... | 2021-04-30T09:31:17 | 2021-04-30T09:31:17 | {} | NONE |
ClickHouse | ClickHouse | 829,998,065 | 78 | alexey-milovidov | Alright. | 2021-04-30T10:23:02 | 2021-04-30T10:23:02 | {} | MEMBER |
ClickHouse | ClickHouse | 830,117,561 | 23,803 | filimonov | The same is needed for `format_avro_schema_registry_url`
```
2021.04.30 14:00:04.277754 [ 21 ] {} <Error> void DB::StorageKafka::threadFunc(size_t): Code: 86, e.displayText() = DB::Exception: Received error from remote server /schemas/ids/1. HTTP status code: 401 Unauthorized, body: {"error_code": 40101, "message":... | 2021-04-30T14:02:42 | 2021-04-30T14:02:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,509,917 | 23,050 | alexey-milovidov | > Functional stateless tests flaky check (address) — Timeout, fail: 0, passed: 287
This is alright. | 2021-04-29T18:59:58 | 2021-04-29T18:59:58 | {} | MEMBER |
ClickHouse | ClickHouse | 829,854,208 | 4,745 | olgarev | @alexey-milovidov Documented already, please remove the `comp-documentation` label. | 2021-04-30T05:44:49 | 2021-04-30T05:44:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,989,970 | 23,791 | alexey-milovidov | ```
milovidov-desktop :) SELECT count() FROM numbers_mt(1000000000) WHERE NOT ignore(bitTest(materialize(1000), 4))
SELECT count()
FROM numbers_mt(1000000000)
WHERE NOT ignore(bitTest(materialize(1000), 4))
Query id: 164f1985-40d8-49c1-8704-0a4fde0f4557
┌────count()─┐
│ 1000000000 │
└────────────┘
1 ro... | 2021-04-30T10:08:29 | 2021-04-30T10:08:29 | {} | MEMBER |
ClickHouse | ClickHouse | 829,997,777 | 2,857 | alexey-milovidov | No, it is alright. | 2021-04-30T10:22:31 | 2021-04-30T10:22:31 | {} | MEMBER |
ClickHouse | ClickHouse | 829,999,138 | 7,644 | alexey-milovidov | @olgarev Yes, we need to provide description on the existing metrics. | 2021-04-30T10:25:08 | 2021-04-30T10:25:08 | {} | MEMBER |
ClickHouse | ClickHouse | 830,001,135 | 23,791 | alexey-milovidov | AFAIK, the pipeline with output is build,
there is single output processor,
and then there is an overhead in moving chunks between ports. | 2021-04-30T10:29:02 | 2021-04-30T10:29:02 | {} | MEMBER |
ClickHouse | ClickHouse | 830,022,556 | 23,791 | UnamedRus | > there is single output processor.
> But here we have 177.92 GB/s. for processing inside CPU cache and different effects matter (atomic ops in query pipeline).
Ok thank you. didn't know that.
| 2021-04-30T11:13:08 | 2021-04-30T11:13:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,312,934 | 23,674 | filimonov | Most probably you are doing something wrong. It is a user-level setting, you need to put it into users.xml inside the default profile.
If it does not help - please show what exactly did you do, and how the file looks like.
P.S. Relaxing that limitation (`max_partitions_per_insert_block`) is typically a very bad i... | 2021-04-29T15:02:23 | 2021-04-29T15:02:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,585,688 | 23,756 | filimonov | It cames from https://github.com/ClickHouse/ClickHouse/blob/c7bd69cc92b41ab6ab61b67956276baca0af1cbd/docker/server/entrypoint.sh#L117
That means you have some init DB script. And to run it docker entrypoint starts clickhouse, and wait up to configured timeout (default is 12 sec) to run those scripts.
Far from per... | 2021-04-29T20:48:16 | 2021-04-29T20:48:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,998,671 | 4,745 | alexey-milovidov | Perfect. | 2021-04-30T10:24:11 | 2021-04-30T10:24:11 | {} | MEMBER |
ClickHouse | ClickHouse | 829,990,141 | 23,791 | alexey-milovidov | ```
milovidov-desktop :) SELECT count() FROM numbers_mt(1000000000) WHERE NOT ignore(bitTest(materialize(1000), 4))
SELECT count()
FROM numbers_mt(1000000000)
WHERE NOT ignore(bitTest(materialize(1000), 4))
Query id: 807ed776-a304-4283-887b-a56f511b79ac
┌────count()─┐
│ 1000000000 │
└────────────┘
1 ro... | 2021-04-30T10:08:49 | 2021-04-30T10:08:49 | {} | MEMBER |
ClickHouse | ClickHouse | 830,001,777 | 23,791 | alexey-milovidov | If query performance is slower than around 10 GB/sec you can use queries with FORMAT Null safely.
But here we have 177.92 GB/s. for processing inside CPU cache and different effects matter (atomic ops in query pipeline). | 2021-04-30T10:30:09 | 2021-04-30T10:30:36 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 830,063,785 | 23,765 | filimonov | > this does not work in my case
what exactly do you mean?
> this works for local table but not for distributed table
What exactly do you mean?
Does it say 'i will not work in that case?' :)
| 2021-04-30T12:35:54 | 2021-04-30T12:35:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 830,150,303 | 23,806 | hagen1778 | Ah! Thank you, @UnamedRus! My finding skills failed me. Closing as duplicate. | 2021-04-30T14:53:21 | 2021-04-30T14:53:21 | {} | NONE |
ClickHouse | ClickHouse | 830,225,088 | 23,283 | excitoon | As a quick workaround I can suggest [URL-encoding special characters other than these: `/!-_.*'()`](https://docs.aws.amazon.com/AmazonS3/latest/userguide/object-keys.html) | 2021-04-30T16:51:39 | 2021-04-30T16:51:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,215,939 | 23,196 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-04-29T12:58:16 | 2021-04-29T12:58:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,587,018 | 23,719 | filimonov | > Question is, is it usual that it consumes such a big amount of memory?
No. | 2021-04-29T20:50:22 | 2021-04-29T20:50:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 830,114,525 | 23,431 | KochetovNicolai | ```
select number from remote('127.0.0.{3|2}', numbers(2)) where number global in (select number from numbers(1)) settings async_socket_for_remote=1, use_hedged_requests = 1, sleep_in_send_data_ms=10, receive_data_timeout_ms=1;
SELECT number
FROM remote('127.0.0.{3|2}', numbers(2))
WHERE number GLOBAL IN
(
... | 2021-04-30T13:58:10 | 2021-04-30T14:01:01 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 830,166,889 | 9,932 | filimonov | > @filimonov Could you please write a test using CREATE DICTIONARY query?
I was not able to reproduce it using CREATE DICTIONARY:
```
DROP TABLE IF EXISTS dictionary_source_en;
DROP TABLE IF EXISTS dictionary_source_ru;
DROP TABLE IF EXISTS dictionary_source_view;
DROP DICTIONARY IF EXISTS flat_dictionary;
... | 2021-04-30T15:15:43 | 2021-04-30T15:15:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,297,559 | 23,539 | filimonov | See https://github.com/ClickHouse/ClickHouse/issues/23444#issuecomment-824566121 | 2021-04-29T14:42:41 | 2021-04-29T14:42:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,660,195 | 23,775 | excitoon | @tritas Can you please attach error message and how Swift S3 responses look like (it can be seen at `trace` log level)? | 2021-04-29T23:20:10 | 2021-04-29T23:20:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,854,322 | 23,780 | amosbird | It's not yet implemented but already planned in the roadmap. It will be done in Q2 or Q3. | 2021-04-30T05:45:08 | 2021-04-30T05:45:08 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 829,855,237 | 2,857 | olgarev | @alexey-milovidov Do we need more information here? https://clickhouse.tech/docs/en/sql-reference/statements/select/join/ | 2021-04-30T05:47:39 | 2021-04-30T05:47:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,902,337 | 16,947 | azat | I'm going to split this PR into separate patches to make it easier to review (with a separate reproducers/tests), the first one is (should be the most obvious) - #23783 | 2021-04-30T07:31:24 | 2021-04-30T07:31:24 | {} | MEMBER |
ClickHouse | ClickHouse | 829,998,270 | 78 | alexey-milovidov | > please remove the comp-documentation label.
You mean: this task can be closed as fixed. | 2021-04-30T10:23:25 | 2021-04-30T10:23:25 | {} | MEMBER |
ClickHouse | ClickHouse | 830,191,547 | 23,746 | bnaecker | @alexey-milovidov Yes, that's right! Zones are a general tool on illumos for isolating runtime environments and resources, and LX zones allow running a Linux kernel without a full VM. It works by translating the system calls, and is a convenient way to get Linux-only applications running. It does have some drawbacks th... | 2021-04-30T15:54:26 | 2021-04-30T15:54:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,318,489 | 23,658 | filimonov | ```
WITH
[1, 2] AS a1,
arrayMap(x -> (x / 10), a1) AS a2
SELECT arrayFlatten(arrayMap((x, y) -> [x, y], a1, a2))
Query id: 11b3f1d2-4630-4c50-a27e-42a44635add4
┌─arrayFlatten(arrayMap(lambda(tuple(x, y), array(x, y)), a1, a2))─┐
│ [1,0.1,2,0.2] │
└─... | 2021-04-29T15:09:07 | 2021-04-29T15:09:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,338,762 | 23,682 | krebrikov | If there were direct functions like this it would be a perfect outcome !
Actually I didn't intend this Issue to be a feature request, but rather hoped that there is already some workaround to enrich inserts-related statistics (counted within derived MV) with audit data from system.query_log.
If it was a function -... | 2021-04-29T15:31:19 | 2021-04-29T15:31:19 | {} | NONE |
ClickHouse | ClickHouse | 829,940,812 | 23,761 | kssenii | `ClickHouse build check` - was broken in master
`Integration tests` - not related
`Functional stateless tests (ANTLR debug) ` - not related
`Yandex third-party checks (only for Yandex employees)` - not related | 2021-04-30T08:41:19 | 2021-04-30T08:41:19 | {} | MEMBER |
ClickHouse | ClickHouse | 829,993,664 | 23,791 | akuzm | > Don't use FORMAT Null to test the performance on synthetic benchmarks.
Why? It is used in a lot of performance tests. We have special tweak that doesn't send the result blocks to client in case of `format Null`, why is it different from calculating the same column for `WHERE` and discarding it? | 2021-04-30T10:15:00 | 2021-04-30T10:15:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 830,001,720 | 23,296 | crazyproger | We meet this problem almost every day( Currently fix it by detach/attach.
We have cluster with 4 shards, 2 mirrors in each shard. All CH version 20.8.14.4.
However table on which we meet this problems is bit different: it is ReplicatedMergeTree table which is mirrored on each instance(so 8 mirrors, `{shard}` in zk pa... | 2021-04-30T10:30:02 | 2021-04-30T10:48:01 | {} | NONE |
ClickHouse | ClickHouse | 830,050,388 | 22,398 | alexey-milovidov | ClickHouse vs. Redshift on time series data:
http://brandonharris.io/redshift-clickhouse-time-series/ | 2021-04-30T12:10:54 | 2021-04-30T12:11:00 | {} | MEMBER |
ClickHouse | ClickHouse | 829,503,513 | 19,823 | andrebsguedes | I think I may be hitting something related to this while working on a custom table engine. I have had IStorage::read being called 4 times sequentially when used in a sub query in the right side of the IN operator. | 2021-04-29T18:49:51 | 2021-04-29T18:49:51 | {} | NONE |
ClickHouse | ClickHouse | 829,529,419 | 23,555 | filimonov | > This seems similar to #22438, but the version we are running (21.4.4) should have the backported fix.
>
> Also the issue has not reoccurred on our clusters as of yet.
It is possible.
Are you seeing "Memory limit exceeded" exception? Does is happen that server has all allowed memory used by clickhouse? | 2021-04-29T19:31:53 | 2021-04-29T19:32:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 830,062,892 | 23,508 | filimonov | A couple of reports already. Needs attention. | 2021-04-30T12:34:21 | 2021-04-30T12:34:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 830,092,284 | 23,001 | SaltTan | #23800 | 2021-04-30T13:24:14 | 2021-04-30T13:24:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 830,115,812 | 23,800 | SaltTan | The dictionary source is MySQL and the expression for `update_field` looks like this:
```
<update_field>date_add(updated_date, interval 15 minute)</update_field>
```
I'm trying to create a test case, so far I found that `update_field` as an expression works really weird with ClickHouse as a source... | 2021-04-30T14:00:07 | 2021-04-30T14:00:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,446,019 | 23,736 | CurtizJ | But https://github.com/kshvakov/clickhouse-client is an empty repository. | 2021-04-29T17:20:07 | 2021-04-29T17:20:07 | {} | MEMBER |
ClickHouse | ClickHouse | 829,856,288 | 78 | olgarev | @alexey-milovidov Documented, please remove the `comp-documentation` label. | 2021-04-30T05:50:18 | 2021-04-30T05:50:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,864,824 | 23,539 | amosbird | > > https://sqlite.org/lang_with.html#mathint
>
> @amosbird what do you think about it?
Sounds good. Currently we only have implicit support for this feature: GLOBAL IN/JOIN. I'll try to implement the CTE hint. | 2021-04-30T06:10:36 | 2021-04-30T06:10:36 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 830,062,116 | 23,778 | filimonov | 1) clickhouse version (odbc bridge was reworked just a month ago, so try the latest clickhouse version)
2) share the list of OS processes and their RAM usage.
3) do `set send_logs_level='trace'` before the `INSERT ... SELECT ...` query and share the result.
| 2021-04-30T12:32:49 | 2021-04-30T12:32:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 830,094,198 | 23,794 | akuzm | `ORDER BY` has to read the entire input, and `system.numbers` is endless. This is the expected behavior. | 2021-04-30T13:27:29 | 2021-04-30T13:27:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 830,140,163 | 21,414 | amosbird | > @amosbird `01414_push_predicate_when_contains_with_clause` failed :(
> https://clickhouse-test-reports.s3.yandex.net/21414/51444be49f194ecd3d115d5527069862a8445d3f/fast_test/test_log.txt
>
> Error
> ```
> 2021-04-29 17:23:54 01523_interval_operator_support_string_literal: [ OK ]
> 2021-... | 2021-04-30T14:38:06 | 2021-04-30T14:38:06 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 829,214,330 | 23,734 | den-crane | Did you ever restore your table? Did you use attach partition?
https://github.com/ClickHouse/ClickHouse/issues/18804 | 2021-04-29T12:55:51 | 2021-04-29T12:55:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,519,083 | 23,682 | filimonov | > workaround to enrich inserts-related statistics (counted within derived MV) with audit data from system.query_log.
You can try joins (of course later, it is not possible at insert / MV time) . But it would be quite painful, non reliable and and non efficient.
Addind such function should be trivial. | 2021-04-29T19:14:49 | 2021-04-29T19:15:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,579,258 | 23,765 | filimonov | Aggregate functions are not aware of ordering (and i doubt they will be in close future), you can try smth like:
```
Select count() from (select distinct x from table)
```
Or
```
SET optimize_aggregation_in_order=1
Select count() from (select 1 from table group by x)
``` | 2021-04-29T20:37:12 | 2021-04-29T20:37:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,880,495 | 23,296 | Blackmorse | I have no system.part_log turned on..
Also this problem was gone after I deattached/attached part | 2021-04-30T06:46:32 | 2021-04-30T06:46:32 | {} | NONE |
ClickHouse | ClickHouse | 829,885,422 | 23,765 | qingping209 | > Select count() from (select distinct x from table)
this does not work in my case, neither local or distributed table;
> SET optimize_aggregation_in_order=1
> Select count() from (select 1 from table group by x)
this works for local table but not for distributed table | 2021-04-30T06:57:44 | 2021-04-30T06:57:57 | {} | NONE |
ClickHouse | ClickHouse | 829,984,207 | 23,296 | tavplubix | Hm, similar issue was reported for 20.8.2.3 (#18559) and 1.1.54394.3 ( #4833). Seems like the bug has always existed. | 2021-04-30T09:58:23 | 2021-04-30T09:58:23 | {} | MEMBER |
ClickHouse | ClickHouse | 830,095,782 | 23,778 | jhettler | Hi @filimonov,
I am using the latest docker image server version `ClickHouse server version 21.4.5 revision 54447.`
TOP of approx 1/4 of the INSERT ... SELECT ... process
```
top - 12:39:54 up 22:16, 0 users, load average: 2.33, 1.92, 1.49
Tasks: 5 total, 1 running, 4 sleeping, 0 stopped, 0 zombie
%... | 2021-04-30T13:30:08 | 2021-04-30T13:30:08 | {} | NONE |
ClickHouse | ClickHouse | 830,129,512 | 20,539 | myrrc | The first more-or-less working version is here. The slowdown for test processing (the data collection step only) is about 10x (0.2s -> 2s) or so. The test conversion takes about 10m for the test (but I guess the global symbols cache will speedup is, so it's not 10m for every test, but only for the first one).
[tes... | 2021-04-30T14:21:44 | 2021-04-30T14:21:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,527,509 | 23,759 | alexey-milovidov | Попробуйте выставить настройку
`SET join_algorithm = 'auto'` | 2021-04-29T19:28:35 | 2021-04-29T19:28:35 | {} | MEMBER |
ClickHouse | ClickHouse | 829,978,613 | 23,779 | tavplubix | Duplicate of #23508
Also related to #23700 | 2021-04-30T09:48:16 | 2021-04-30T09:48:16 | {} | MEMBER |
ClickHouse | ClickHouse | 830,017,227 | 23,296 | crazyproger | Attached output of query:
```
WITH
'202103_37087_37167_7' AS p1,
'202103_37167_37350_11' AS p2
SELECT materialize(hostName()) AS host,
*
FROM remote(
<all_hosts>,
'system.part_log')
WHERE (database = 'db')
AND (table = 'table')
AND ((part_name IN (p1, p2)) OR has(merged_fr... | 2021-04-30T11:02:03 | 2021-04-30T11:02:03 | {} | NONE |
ClickHouse | ClickHouse | 830,271,413 | 23,800 | kitaisreal | @SaltTan if you remove ```update field``` dictionary will load successfully ? | 2021-04-30T18:10:38 | 2021-04-30T18:12:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,231,285 | 9,195 | maxis42 | Looks like spark setting "spark.stage.maxConsecutiveAttempts" == 1 eliminates duplicates. | 2021-04-29T13:19:38 | 2021-04-29T13:19:38 | {} | NONE |
ClickHouse | ClickHouse | 829,298,697 | 23,539 | filimonov | > https://sqlite.org/lang_with.html#mathint
@amosbird what do you think about it? | 2021-04-29T14:44:12 | 2021-04-29T14:44:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,308,364 | 23,682 | filimonov | In other words you want to make things like query_id and initial_query_id available as functions, and use them inside your materialized views?
Something like:
```
CREATE MATERIALIZED VIEW foo AS SELECT hostName(), count(), median(), queryID(), initialQueryID() from bar;
```
? | 2021-04-29T14:56:51 | 2021-04-29T14:56:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,441,171 | 9,609 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/pull/23771 | 2021-04-29T17:12:57 | 2021-04-29T17:12:57 | {
"hooray": 1
} | MEMBER |
ClickHouse | ClickHouse | 829,795,172 | 7,644 | olgarev | @alexey-milovidov Do we need extra documentation here?
https://clickhouse.tech/docs/en/operations/system-tables/asynchronous_metric_log/
https://clickhouse.tech/docs/en/operations/system-tables/asynchronous_metrics/ | 2021-04-30T04:25:03 | 2021-04-30T04:25:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,852,115 | 23,781 | ucasfl | > I think we should mark it as bug fix and backport (it is regressions for users)
Alright! | 2021-04-30T05:38:30 | 2021-04-30T05:38:30 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 829,622,273 | 23,664 | azat | >ClickHouse build check — 14/15 builds are OK
Fixed in #23760
UPD: it hasn't been fixed there | 2021-04-29T21:57:11 | 2021-04-30T10:38:23 | {} | MEMBER |
ClickHouse | ClickHouse | 830,277,933 | 23,800 | SaltTan | > @SaltTan if you remove `update field` dictionary will load successfully ?
Yes.
Here is a test that demonstrates that `update_field` acts like `where`:
```
create dictionary system.complex_key_hashed_dict
(
name String,
code Int32,
value UInt64,
last_error_time DateTime,
last_error_message String,
remo... | 2021-04-30T18:22:28 | 2021-04-30T18:22:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,537,715 | 23,736 | asipincasp | with latest driver
https://github.com/ClickHouse/clickhouse-go/releases/tag/v1.4.3
same error
29.04.2021 20:20, Anton Popov пишет:
>
> But https://github.com/kshvakov/clickhouse-client
> <https://github.com/kshvakov/clickhouse-client> is an empty repository.
>
> —
> You are receiving this because you authored the thr... | 2021-04-29T19:44:57 | 2021-04-29T19:44:57 | {} | NONE |
ClickHouse | ClickHouse | 829,850,591 | 8,703 | MaxWk | is there has any update | 2021-04-30T05:34:14 | 2021-04-30T05:34:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 830,004,939 | 23,776 | alexey-milovidov | master: total errors 210.
this PR: total errors 204. | 2021-04-30T10:36:24 | 2021-04-30T10:36:24 | {} | MEMBER |
ClickHouse | ClickHouse | 830,126,695 | 23,778 | filimonov | Yes. So it is clickhouse-odbc process eats RAM (not clickhouse-server).
It can be a fault of the odbc driver used (it is loaded in the clickhouse-odbc process),
> I am using the latest docker image server version ClickHouse server version 21.4.5 revision 54447.
Not new enough :) https://github.com/ClickHouse... | 2021-04-30T14:17:13 | 2021-04-30T14:17:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 830,167,076 | 9,932 | filimonov | But may be behavior with xml differs. | 2021-04-30T15:16:01 | 2021-04-30T15:16:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,215,799 | 23,196 | kssenii | @Mergifyio update | 2021-04-29T12:58:03 | 2021-04-29T12:58:03 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 829,340,528 | 23,770 | lodthe | Created accidently | 2021-04-29T15:33:31 | 2021-04-29T15:33:54 | {} | NONE |
ClickHouse | ClickHouse | 829,352,985 | 9,609 | alexey-milovidov | We have finally found the culprit. The fix is on the way. | 2021-04-29T15:48:03 | 2021-04-29T15:48:03 | {} | MEMBER |
ClickHouse | ClickHouse | 829,368,032 | 23,139 | abraithwaite | Is there any more context for this comment?
https://github.com/ClickHouse/ClickHouse/pull/23139/files#diff-cbf46e10d88380e199f81d8c426b455b1a314f48d8fc0b9d97b9f55e815ca8e0R11
It seems logical that forcing a merge would reduce the active parts count. Why wouldn't that be the case? Thanks. | 2021-04-29T16:05:49 | 2021-04-29T16:05:49 | {} | NONE |
ClickHouse | ClickHouse | 829,592,688 | 23,734 | Alekky09 | Rebooting the replicas fixed it! So it looks like these mutations sometimes get stuck in the replicas without executing, reboot jumpstarted it | 2021-04-29T21:00:14 | 2021-04-29T21:00:45 | {} | NONE |
ClickHouse | ClickHouse | 829,787,601 | 12,066 | olgarev | @alexey-milovidov Documented, please remove the doc-alert and comp-documentation tags | 2021-04-30T03:57:49 | 2021-04-30T04:02:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 830,103,415 | 22,152 | filimonov | @tulid thanks for the report, try `altinity/clickhouse-server:21.6.1.6734-testing-arm `
```
ubuntu@ip-172-31-2-152:~/ClickHouse$ docker run -d altinity/clickhouse-server:21.6.1.6734-testing-arm
5e68c505175f9d562bf1eb328e7724b72d2f0a99bd33a68baec0dae63e532b39
ubuntu@ip-172-31-2-152:~/ClickHouse$ docker exec -it... | 2021-04-30T13:41:08 | 2021-04-30T13:41:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,274,246 | 24,157 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-05-18T15:32:52 | 2021-05-18T15:32:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,379,017 | 24,224 | mbtolou | check this query :
```sql
create table test.materialize_test
(
d DateTime,
a Int32,
b Int32,
c Int32 materialized a-b
)
ENGINE = MergeTree
ORDER BY d;
truncate table test.materialize_test;
insert into test.materialize_test values('2020-01-01 00:00:00',1,2)
select *,c
from test.materialize_tes... | 2021-05-18T17:23:48 | 2021-05-18T17:25:31 | {} | NONE |
ClickHouse | ClickHouse | 843,557,620 | 24,259 | kitaisreal | @azat I am working on bigger fix, this pull request fix only local problem. | 2021-05-18T20:56:40 | 2021-05-18T20:56:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 843,593,708 | 24,194 | alexey-milovidov | Tables should be also writable. | 2021-05-18T21:59:10 | 2021-05-18T21:59:10 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.