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 | 773,647,375 | 19,451 | azat | Rebased (to fix conflicts and include #19886) | 2021-02-04T22:33:38 | 2021-02-04T22:33:38 | {} | MEMBER |
ClickHouse | ClickHouse | 773,760,903 | 20,001 | fuwhu | > > If it's less than about 100MB compressed, you can send it to [clickhouse-feedback@yandex-team.com](mailto:clickhouse-feedback@yandex-team.com) in attachment.
> > If it's more, you can upload to Google Drive or Dropbox or https://disk.yandex.com/ or other file sharing service, and send a link to [clickhouse-feedbac... | 2021-02-05T03:29:05 | 2021-02-05T03:29:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,832,750 | 20,092 | alexey-milovidov | That will be awesome! | 2021-02-05T06:49:56 | 2021-02-05T06:49:56 | {} | MEMBER |
ClickHouse | ClickHouse | 773,184,740 | 19,886 | azat | >I'm not able to carefully review the parts about Kafka and RabiitMQ but the code LGTM.
Before 46f5268135ec92b248f5246b2a7a574fdb49f7b7 tests for RabbitMQ fails - https://clickhouse-test-reports.s3.yandex.net/19886/f3a56b965c541c7527b3505da6f4d0f6bfc16f46/integration_tests_(asan).html#fail1
But after - everything... | 2021-02-04T10:00:18 | 2021-02-04T10:00:18 | {} | MEMBER |
ClickHouse | ClickHouse | 773,229,761 | 19,603 | JackyWoo | @alexey-milovidov need help
There are 3 CI checking failures. I tested with master compiled binary file in local which raise the same failures. So I guess maybe the checking failures are not from my pr.
What can I do for the checking failures.
| 2021-02-04T11:13:32 | 2021-02-04T11:13:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,250,714 | 20,063 | alexey-milovidov | @hexiaoting It is slightly more complicated because different programming languages and database engines may use different definitions of modulo of division.
In most math applications, the result of modulo of division by x is convenient to be defined that it always falls into `[0, x)`.
`-1 % 100 = 99`
But there ... | 2021-02-04T11:51:31 | 2021-02-04T11:52:34 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 773,303,057 | 17,708 | AbelVM | I'd go with the whole set of H3 functions, as they provide almost all the needed geospatial functionality to analyze location data.
Adding only type conversion does not improve anything, imho
> Completely agree.
> It would be great to get H3ToGeo at least so data can be easily converted both ways within Click... | 2021-02-04T13:28:33 | 2021-02-04T13:28:33 | {} | NONE |
ClickHouse | ClickHouse | 773,306,886 | 20,062 | filimonov | `git pull --recurse-submodules` <- you don't need it. | 2021-02-04T13:35:19 | 2021-02-04T13:35:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,361,285 | 18,377 | vdimir | Looks like failed tests failed in master recently too.
<details>
<summary>CI Info</summary>
```
SELECT
any(check_name),
max(check_start_time) ,
splitByString('::', test_name)[1],
groupUniqArray(test_status),
count(pull_request_number)
FROM `gh-data`.checks
WHERE
test_status != 'OK'
AND che... | 2021-02-04T14:43:08 | 2021-02-04T14:43:08 | {} | MEMBER |
ClickHouse | ClickHouse | 773,419,073 | 20,076 | pan3793 | Thanks for @filimonov @den-crane explanation, and I do did some tries with data type conversion before open the question, but didn't got an expected equivalent result.
For example:
```
SELECT murmurHash3_32(toInt32(2))
┌─murmurHash3_32(toInt32(2))─┐
│ 1085422463 │
└────────────────────────────... | 2021-02-04T16:04:42 | 2021-02-04T16:06:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,507,236 | 19,999 | kreuzerkrieg | > Yes, it's 100% green :)
Rare natural phenomena, I guess today stars are aligned right :)
| 2021-02-04T18:15:10 | 2021-02-04T18:15:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,174,135 | 19,766 | ghost | I added `sequenceFirstNode` to get the starting points.
When events arguments are omitted in `sequenceNextNode`, it is replaced with `sequenceFirstNode` automatically. | 2021-02-04T09:43:46 | 2021-02-04T09:43:46 | {} | NONE |
ClickHouse | ClickHouse | 773,215,540 | 20,001 | alexey-milovidov | If it's less than about 100MB compressed, you can send it to clickhouse-feedback@yandex-team.com in attachment.
If it's more, you can upload to Google Drive or Dropbox or https://disk.yandex.com/ or other file sharing service, and send a link to clickhouse-feedback@yandex-team.com.
If it's even larger and does not fi... | 2021-02-04T10:49:54 | 2021-02-04T10:50:13 | {} | MEMBER |
ClickHouse | ClickHouse | 773,227,665 | 19,252 | alexey-milovidov | #20064 | 2021-02-04T11:09:43 | 2021-02-04T11:09:43 | {} | MEMBER |
ClickHouse | ClickHouse | 773,308,322 | 19,315 | OmarBazaraa | Thanks @den-crane for your suggestion!
I tried it but it's giving me the same error...
```sql
SELECT *
FROM system.mutations
ORDER BY create_time DESC
LIMIT 1
FORMAT Vertical
Row 1:
──────
database: clarity
table: page_data
mutation_id: mutation_17... | 2021-02-04T13:37:39 | 2021-02-04T13:37:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,331,615 | 20,062 | filimonov | You need to update all submodules by creating a branch from the current commit, and committing URL change on the top of that.
So smth like
1) edit /.gitmodules & commit
2) go to every /contrib/XXX having submodules and do there
- `git checkout -b fixing_submodules`
- edit URLs in .gitmodules
- comm... | 2021-02-04T14:14:02 | 2021-02-04T14:14:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,349,083 | 20,061 | filimonov | Fixed in https://github.com/ClickHouse/ClickHouse/pull/17858
Please update to 21.1 or newer. | 2021-02-04T14:25:03 | 2021-02-04T14:25:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,567,526 | 19,784 | lehasm | May I change PR status from Draft to Ready for review before tests execution? | 2021-02-04T19:58:27 | 2021-02-04T19:58:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,719,721 | 19,407 | templarzq | SUMMARY: ThreadSanitizer: thread leak (/usr/bin/clickhouse+0x8adb1cb) in pthread_create
==================
Stats: SizeClassAllocator64: 3M mapped (1M rss) in 20287 allocations; remains 20159 | 2021-02-05T01:34:44 | 2021-02-05T01:34:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,721,545 | 20,063 | hexiaoting | @alexey-milovidov You are so nice, Thank you very much. your fix is ok, and I have added it into my pr. | 2021-02-05T01:39:46 | 2021-02-05T02:37:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,745,605 | 20,076 | pan3793 | Still not got expected result with `murmurHash3_32`, but succeed with `xxHash64`
```
SELECT toInt32(murmurHash3_32('abc'))
┌─toInt32(murmurHash3_32('abc'))─┐
│ -1277324294 │
└────────────────────────────────┘
SELECT toInt64(xxHash64('a'))
┌─toInt64(xxHash64('a'))─┐
│ -32924777353505... | 2021-02-05T02:45:56 | 2021-02-05T02:47:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,848,316 | 20,063 | alexey-milovidov | I just merged my PR (and also I applied one change to it). You can either: - close; - rebase to master so we can merge more tests. | 2021-02-05T07:23:07 | 2021-02-05T07:23:07 | {} | MEMBER |
ClickHouse | ClickHouse | 773,154,643 | 20,061 | kindred77 | SELECT version()
┌─version()─┐
│ 20.8.3.18 │
└───────────┘
| 2021-02-04T09:14:12 | 2021-02-04T09:14:12 | {} | NONE |
ClickHouse | ClickHouse | 773,182,755 | 20,063 | zhangjmruc | Do we need to limit the fix to cases where type A is signed and type B is usigned? | 2021-02-04T09:57:17 | 2021-02-04T09:57:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,311,690 | 20,043 | ghost | I'm also testing it on a very tiny node, 1-core + 2GB RAM if that can help. | 2021-02-04T13:43:37 | 2021-02-04T13:43:37 | {} | NONE |
ClickHouse | ClickHouse | 773,554,620 | 19,580 | alesapin | Actually not a deadlock, but sometimes it takes a gigantic amount of time (about 5-15 min or so) to respond to a simple request (checked only with thread sanitizer, maybe in release build it works well). Seems like it happens when the node decides to send a request to the leader but the leader changes to this node bef... | 2021-02-04T19:36:00 | 2021-02-04T19:36:27 | {} | MEMBER |
ClickHouse | ClickHouse | 773,683,211 | 20,088 | nvartolomei | Why would it timeout at `DROP TABLE`? 🤔 | 2021-02-05T00:01:05 | 2021-02-05T00:01:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,717,181 | 19,896 | depressed-pho | And now only the Yandex-exclusive checks are failing. I can't do anything about that. | 2021-02-05T01:27:28 | 2021-02-05T01:27:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,844,223 | 20,068 | akuzm | fuzzer probably https://github.com/ClickHouse/ClickHouse/issues/20017 | 2021-02-05T07:14:37 | 2021-02-05T07:14:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,288,633 | 20,062 | leosunli | > Question nr. 1: why do you build 20.8.7? You don't need it, believe me. If you want lts version - pick the latest 20.8. (20.8.12 currently).
I have now updated to the latest version (v20.8.11.17-lts)
> Question nr. 2: why don't you use official builds?
Need this version for historical reasons. Is it the lts v... | 2021-02-04T13:03:27 | 2021-02-04T13:12:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,358,598 | 20,001 | fuwhu |
> If it's less than about 100MB compressed, you can send it to [clickhouse-feedback@yandex-team.com](mailto:clickhouse-feedback@yandex-team.com) in attachment.
> If it's more, you can upload to Google Drive or Dropbox or https://disk.yandex.com/ or other file sharing service, and send a link to [clickhouse-feedback@... | 2021-02-04T14:39:17 | 2021-02-04T14:40:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,374,472 | 20,062 | leosunli | After i use 'git submodule update --init --recursive'
leo@B-59ZDMD6M-0211 ClickHouse % git status
位于分支 branch-20.8.11.17
尚未暂存以备提交的变更：
（使用 "git add <文件>..." 更新要提交的内容）
（使用 "git restore <文件>..." 丢弃工作区的改动）
（提交或丢弃子模组中未跟踪或修改的内容）
修改： contrib/cassandra (新提交, 修改的内容)
修改： contrib/cctz (新提交, 修改的内容)
修改： ... | 2021-02-04T15:02:03 | 2021-02-04T15:02:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,720,982 | 19,407 | templarzq | and other slowdowns do not appear in previous performance test report which is the same code.. | 2021-02-05T01:38:14 | 2021-02-05T01:38:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,843,728 | 20,017 | akuzm | https://clickhouse-test-reports.s3.yandex.net/20068/b0fba3decef216e32c8813c0fe321210915159c6/fuzzer_ubsan/report.html#fail1
Found a new similar report under UBSan, but can't reproduce:
```
CREATE TABLE tracking_events_tmp (`APIKey` UInt32, `EventDate` Date) ENGINE = MergeTree PARTITION BY toYYYYMM(EventDate) ORDER... | 2021-02-05T07:13:40 | 2021-02-05T07:13:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,229,055 | 20,064 | alexey-milovidov | According to git blame, it's already fixed: c39ee3f5, #19868. CC @amosbird | 2021-02-04T11:12:14 | 2021-02-04T11:12:30 | {} | MEMBER |
ClickHouse | ClickHouse | 773,242,664 | 20,063 | alexey-milovidov | @hexiaoting I am preparing the same fix at this time.
It looks like this:
```
-/** Division with remainder you get a number with the same number of bits as in divisor.
- */
+/** Division with remainder you get a number with the same number of bits as in divisor,
+ * or larger in case of signed type.
+ */
... | 2021-02-04T11:37:10 | 2021-02-04T11:37:10 | {} | MEMBER |
ClickHouse | ClickHouse | 773,408,851 | 5,160 | mzealey | This is an issue when using `remote()` over connections with a bit of latency even if only 1 server is specified w/o sharding key. I suggest if 1 server then it should be a much longer connect timeout (eg a few seconds) | 2021-02-04T15:50:23 | 2021-02-04T15:50:23 | {} | NONE |
ClickHouse | ClickHouse | 773,578,192 | 20,094 | alexey-milovidov | #20096 | 2021-02-04T20:17:33 | 2021-02-04T20:17:33 | {} | MEMBER |
ClickHouse | ClickHouse | 773,591,540 | 20,043 | den-crane | can you show the output of
from 2G instance
```
set send_logs_level = 'debug';
SELECT count(*) FROM test WHERE title LIKE '%as' format Null; | 2021-02-04T20:42:26 | 2021-02-04T20:42:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,727,054 | 20,018 | JackyWoo | hi alexey-milovidov vitlibar , I have fixed the bug, today I will commit the pr. | 2021-02-05T01:56:44 | 2021-02-05T01:56:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,840,521 | 20,068 | akuzm | It works. | 2021-02-05T07:07:15 | 2021-02-05T07:07:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,233,183 | 20,064 | alexey-milovidov | @filimonov I cannot reproduce your report on 21.2.1.
But I found that you are using some unknown build variant, earlier than prestable.
Prestable version is:
```
ClickHouse 21.2.1.5869 with revision 54447, build id: 17AA52CB220EC095D4C6579FD7196243A0CDAEBF.
Checksum of the binary: 28216FC61E2F4FB287D643B95888331... | 2021-02-04T11:19:49 | 2021-02-04T11:20:39 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 773,321,921 | 18,945 | qoega | 00159_parallel_formatting_http Stateful thread https://clickhouse-test-reports.s3.yandex.net/0/dbcddbcebdc22425f0ada3b7e569489c2279ffd6/functional_stateful_tests_(thread).html
```
2021-02-02 13:42:58 00159_parallel_formatting_json_and_friends: [ FAIL ] 600.52 sec. - Timeout!
2021-02-02 13... | 2021-02-04T14:00:05 | 2021-02-04T14:00:05 | {} | MEMBER |
ClickHouse | ClickHouse | 773,312,827 | 20,062 | leosunli | I have one question:
I updated url in .gitsubmodule on master branch of submodule arrow. and pushed it internal gitlab.
But i use 'git submodule update --init --recursive ' ,
the url is not updated in .gitsubmodule on separating branch of submodule arrow (https://github.com/ClickHouse-Extras/arrow/blob/master/.gi... | 2021-02-04T13:45:28 | 2021-02-04T14:06:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,492,104 | 19,999 | kreuzerkrieg | > Tests is broken. We should convert enum to string properly
> [test_disk_types](https://github.com/ClickHouse/ClickHouse/blob/830b0eb36eecac54aba06c6744c1f521b0855d8c/tests/integration/test_disk_types/test.py#L5)
ok, how it is now? are we good to go? | 2021-02-04T17:50:33 | 2021-02-04T17:50:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,501,776 | 19,784 | lehasm | Noticed in src\AggregateFunctions\AggregateFunctionMannWhitney.h typos in text messages:
* throw Exception("Unknown parameter in aggregate function " + getName() + ". It must be one of: **'two sided'**, 'less', 'greater'", ErrorCodes::BAD_ARGUMENTS);
Should be two-sided (with dash)
* In several places word require ... | 2021-02-04T18:06:08 | 2021-02-04T18:06:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,539,967 | 20,086 | filimonov | Kafka don't know if the first level table is real MergeTree, or for example engine=Null (then only second level MVs matter). | 2021-02-04T19:10:28 | 2021-02-04T19:10:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,553,875 | 19,252 | filimonov | Maybe that need to be checked https://github.com/ClickHouse/ClickHouse/pull/19518#issuecomment-771118083
To avoid performance issues for people using plain IPv6. | 2021-02-04T19:34:43 | 2021-02-04T19:34:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,830,604 | 20,001 | alexey-milovidov | Yes, I've got the files. | 2021-02-05T06:45:19 | 2021-02-05T06:45:19 | {} | MEMBER |
ClickHouse | ClickHouse | 773,272,392 | 19,995 | filimonov | It seems you are doing something wrong with submodules. Submodule points to the concrete commit, not to a master
https://github.com/msgpack/msgpack-c/tree/46684265d50b5d1b062d4c5c428ba08462844b1d | 2021-02-04T12:33:02 | 2021-02-04T12:33:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,304,391 | 20,043 | ghost | Actually I can see the same behavior in clickhouse-client for some reason - you can see that with growth of rows processed, progress percent is volatile (sorry for a lot of images, could not find a way to make it shorter):
<img width="969" alt="Снимок экрана 2021-02-04 в 15 24 30" src="https://user-images.githubuser... | 2021-02-04T13:31:01 | 2021-02-04T13:31:01 | {} | NONE |
ClickHouse | ClickHouse | 773,505,329 | 17,278 | alexey-milovidov | Ok, thank you! | 2021-02-04T18:11:55 | 2021-02-04T18:11:55 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 773,551,380 | 20,076 | filimonov | Big endian (java) vs little endian (x86)?
Yes, try with strings. | 2021-02-04T19:30:14 | 2021-02-04T19:30:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,841,877 | 19,321 | bharatnc | Fixed via https://github.com/ClickHouse/ClickHouse/pull/19090. Looks correct now. | 2021-02-05T07:10:05 | 2021-02-05T07:10:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,342,960 | 20,061 | filimonov | can not reproduce on 21.2. | 2021-02-04T14:20:20 | 2021-02-04T14:20:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,356,322 | 19,315 | den-crane | @OmarBazaraa
HMm, I think this error from the previous mutations.
Try remove failed mutations first:
```
kill mutation where not is_done;
ALTER TABLE clarity.page_data DELETE WHERE joinHas(, , (ProjectId, serId, SessionId) )
``` | 2021-02-04T14:35:41 | 2021-02-04T14:35:41 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,357,139 | 20,062 | leosunli | > git checkout -b fixing_submodules
fixing_submodules should be **the branch** in which the url is updated in .gitsubmodule of submodule repo.
Should be **the branch** pointed to origin/HEAD or other branch ?
| 2021-02-04T14:37:00 | 2021-02-04T14:48:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,469,212 | 20,076 | den-crane | @pan3793 check the string first 'abc'
```
SELECT murmurHash3_32('abc')
┌─murmurHash3_32('abc')─┐
│ 3017643002 │
└───────────────────────┘
```
Need to verify that algo & seed are consistent
Also Java does not have Unsigned
```
SELECT toInt32(murmurHash3_32('abc'))
┌─toInt32(murmurHash3_... | 2021-02-04T17:14:56 | 2021-02-04T17:24:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,549,638 | 20,039 | abyss7 | Assert can't cause to fail non-debug integration tests. Seems unrelated. | 2021-02-04T19:27:01 | 2021-02-04T19:27:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,748,123 | 20,027 | bharatnc | @tavplubix would you mind reviewing ? This is one last PR (system.part_log was left to add `event_time_microseconds` field) and is similar to: https://github.com/ClickHouse/ClickHouse/pull/14760 that you had reviewed in the past. Thank you! | 2021-02-05T02:53:19 | 2021-02-05T02:53:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,315,943 | 20,062 | leosunli | Since our build machine cannot connect to the external network, I put all dependent project internal gitlab.
But the url is not updated in .gitsubmodule on separating branch of submodule, when i has udpated url in .gitsubmodule on master branch of submodule. | 2021-02-04T13:50:37 | 2021-02-04T13:50:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,375,521 | 20,059 | den-crane | Nothing wrong here. It's another limit.
Single array max size = 1mil
https://github.com/ClickHouse/ClickHouse/blob/486418d112b8d97eccbed1bf9317e5b5870da381/src/Columns/ColumnArray.cpp#L44
Range function = 100 mil per block.
With range per block you'll get **A call to function range would produce 100016599 ar... | 2021-02-04T15:03:33 | 2021-02-04T15:03:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,172,855 | 20,048 | ben-efiz | @alexey-milovidov ok understood. I think `system.zookeeper_log` makes sense then. Maybe also naming it different then error/exceptions? Because this is also what i see when querying `system.errors`
```
SELECT *
FROM system.errors
ORDER BY value DESC
┌─name───────────────────┬─code─┬─value─┐
│ KEEPER_EXCEPTION... | 2021-02-04T09:41:50 | 2021-02-04T09:41:50 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 773,277,894 | 19,995 | leosunli | Hi @filimonov
I don't understand that you say Submodule points to the concrete commit.
Could you help me explain in more detail？
Thanks | 2021-02-04T12:43:41 | 2021-02-04T12:43:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,363,666 | 19,981 | filimonov | Also https://github.com/ClickHouse/ClickHouse/pull/7426 | 2021-02-04T14:46:35 | 2021-02-04T14:46:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,368,850 | 20,075 | kilik52 | ````
SELECT * FROM system.mutations WHERE is_done = 0
````
Mutations get blocked since i called this:
````SQL
UPDATE `取消原因` = joinGet('order_join', 'r1', `第三方订单ID`), `取消原因二级分类` = joinGet('order_join', 'r2', `第三方订单ID`) WHERE `日期` >= '2021-01-01 00:00:00'
````
After that all mutations such as:
````SQL
... | 2021-02-04T14:53:53 | 2021-02-04T14:59:24 | {} | NONE |
ClickHouse | ClickHouse | 773,392,893 | 20,076 | den-crane | ```sql
select murmurHash3_32((1));
┌─murmurHash3_32(1)─┐
│ 3831157163 │
└───────────────────┘
select murmurHash3_32(toUInt16(1));
┌─murmurHash3_32(toUInt16(1))─┐
│ 2511813514 │
└─────────────────────────────┘
select murmurHash3_32(toUInt32(1));
┌─murmurHash3_32(toUInt32(1))─┐
│ ... | 2021-02-04T15:27:46 | 2021-02-04T15:27:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,396,076 | 19,580 | alesapin | Looks like (dead|live)lock? All SessionID requests waiting for some other request which stucked in `send_msg_to_leader`.
```
| | | 155 0x0000000015b4fd18 in Poco::ThreadImpl::runnableEntry at /home/alesap/code/cpp/ClickHouse/contrib/poco/Foundation/src/Thread_POSIX.cpp:345
| | | | 55 0x0000000015b51510 in Poco::(... | 2021-02-04T15:32:07 | 2021-02-04T15:35:05 | {} | MEMBER |
ClickHouse | ClickHouse | 773,501,841 | 19,999 | alexey-milovidov | Yes, it's 100% green :)
Let @vdimir finish review and merge. | 2021-02-04T18:06:14 | 2021-02-04T18:06:14 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 773,687,653 | 20,088 | nvartolomei | ```
2021.02.05 01:06:38.857000 [ 36 ] {} <Error> DatabaseCatalog: Cannot drop table default.t (bb18f622-0fc3-464f-b653-0b358c2f1c20). Will retry later.: Code: 242, e.displayText() = DB::Exception: Can't drop readonly replicated table (need to drop data in ZooKeeper as well)
```
It seems then ZK connections did tim... | 2021-02-05T00:11:15 | 2021-02-05T00:13:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,717,557 | 18,508 | depressed-pho | Now AST fuzzers are failing due to seemingly unrelated parts of the code base. | 2021-02-05T01:28:32 | 2021-02-05T01:28:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,276,583 | 20,062 | filimonov | Question nr. 1: why do you build 20.8.7? You don't need it, believe me. If you want lts version - pick the latest 20.8. (20.8.12 currently).
Question nr. 2: why don't you use official builds?
Regarding the issue - my guess is that you are doing something wrong with submodules ([again](https://github.com/ClickH... | 2021-02-04T12:41:13 | 2021-02-04T12:42:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,311,135 | 20,062 | filimonov | > My requirement is to put the dependent projects in the internal gitlab, so I need to update the url in the submodule.
You need to sync the internal repo with the used inside clickhouse.
After that just update URLs in https://github.com/ClickHouse/ClickHouse/blob/master/.gitmodules and commit that file.
w/o doing... | 2021-02-04T13:42:44 | 2021-02-04T13:42:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,355,956 | 20,076 | filimonov | The hashing functions are the same.
Binary represenataion of number `1` may differ significantly. Number of bits (8/16/32/64/...), endianness etc.
Try using it on some byte sequence with fixed / platform independent representation. | 2021-02-04T14:35:05 | 2021-02-04T14:35:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,239,525 | 17,708 | alexey-milovidov | @hombit Do you want to help adding these functions? | 2021-02-04T11:31:08 | 2021-02-04T11:31:08 | {} | MEMBER |
ClickHouse | ClickHouse | 773,253,758 | 20,016 | tavplubix | Distributed DDL queries used to be executed in the main thread before #14684. Then we added a pool of worker threads to allow parallel execution of queue tasks. However, default pool size (`distributed_ddl.pool_size`) is 1 and it's not recommended to increase number of worker threads, because greater value causes arbit... | 2021-02-04T11:57:11 | 2021-02-04T11:57:11 | {} | MEMBER |
ClickHouse | ClickHouse | 773,361,031 | 19,315 | OmarBazaraa | > Try remove failed mutations first
Thanks @den-crane, it worked!
And what is more interesting now is that my original query is working now too without having to use `Join` table.
Also, the column `latest_fail_reason` of `system.mutations` table has been cleared, I can no longer find the old error messages. | 2021-02-04T14:42:48 | 2021-02-04T14:42:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,365,629 | 20,062 | filimonov | > > git checkout -b fixing_submodules
>
> fixing_submodules should be **the branch** in which the url is updated in .gitsubmodule of submodule repo.
> Should be **the branch** pointed to origin/HEAD or other branch ?
That command creates branch from the current state / current commit.
| 2021-02-04T14:49:18 | 2021-02-04T14:49:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,411,876 | 20,084 | den-crane | You have to use AggregateFunctions / States / Merge
```
CREATE MATERIALIZED VIEW ohlc
ENGINE = AggregatingMergeTree
PARTITION BY toDate(bucket)
ORDER BY (stock_id, bucket, slice)
POPULATE AS
SELECT
stock_id,
fromUnixTimestamp64Milli(toInt64(round((toUnixTimestamp64Milli(now64(3)) + 50) / 100) * 100))... | 2021-02-04T15:54:47 | 2021-02-04T15:54:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,436,928 | 20,043 | ghost | I like how your test instance if more powerful than each of our hundred production servers :)
Yes, low RAM available confirmed to be the reason for the issue. Tried on 32G/8-core and everything worked perfectly:
```
root@vps:~# curl -i 'http://localhost:8123/?query=SELECT+count%28%2A%29+FROM+default.test2+WHERE+ti... | 2021-02-04T16:29:20 | 2021-02-04T16:29:20 | {} | NONE |
ClickHouse | ClickHouse | 773,229,396 | 20,064 | alexey-milovidov | It does not reproduce on master:
```
milovidov-desktop :) SELECT partition_id, argMaxIf(max_block_number, max_block_number, level = 0) FROM system.parts WHERE (database = 'test') AND (table = 'hits') GROUP BY partition_id
SELECT
partition_id,
argMaxIf(max_block_number, max_block_number, level =... | 2021-02-04T11:12:53 | 2021-02-04T11:12:53 | {} | MEMBER |
ClickHouse | ClickHouse | 773,299,158 | 20,043 | ghost | Ok, I've installed prerelease:
```
labs-den-personal :) select version();
SELECT version()
Query id: 441d00d4-641c-48a2-90e7-a36fcbbbd481
┌─version()───┐
│ 21.2.1.5869 │
└─────────────┘
```
Created table:
```
CREATE TABLE default.test2
(
`id` UInt64,
`title` String,
`time` DateTime
... | 2021-02-04T13:21:50 | 2021-02-04T13:21:50 | {} | NONE |
ClickHouse | ClickHouse | 773,418,851 | 20,084 | sasha-id | Yes, I'm using AggregateFunctions / States / Merge, just edited it manually and forgot to add -State.
Thank you, that what I thought.
| 2021-02-04T16:04:23 | 2021-02-04T16:04:23 | {} | NONE |
ClickHouse | ClickHouse | 773,231,408 | 20,064 | alexey-milovidov | Need info from @amosbird:
- is your fix relevant?
Need info from @filimonov:
- could you please prepare the reproducing example? | 2021-02-04T11:16:27 | 2021-02-04T11:16:27 | {} | MEMBER |
ClickHouse | ClickHouse | 773,261,837 | 19,580 | alesapin | Realized during refactoring that watches from non-leader host doesn't work... | 2021-02-04T12:12:43 | 2021-02-04T12:12:43 | {} | MEMBER |
ClickHouse | ClickHouse | 773,357,337 | 20,075 | filimonov | ```
SELECT * FROM system.mutations WHERE ...
SELECT * FROM system.replication_queue WHERE ...
SELECT * FROM system.merges WHERE ...
SELECT * FROM system.replicas WHERE ...
``` | 2021-02-04T14:37:18 | 2021-02-04T14:37:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,414,544 | 20,084 | den-crane | MV is an insert trigger it never reads original table
Check this:
https://den-crane.github.io/Everything_you_should_know_about_materialized_views_commented.pdf
https://youtu.be/ckChUkC3Pns?list=PLO3lfQbpDVI-hyw4MyqxEk3rDHw95SzxJ&t=9371 | 2021-02-04T15:58:27 | 2021-02-04T15:58:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,479,155 | 20,086 | den-crane | By design.
Insert is successful only all inserts into dependent materialized views succeeded.
Then commit into kafka happen.
| 2021-02-04T17:30:07 | 2021-02-04T17:30:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,566,376 | 19,784 | lehasm | Fixed source code messages in this pull request
@CurtizJ
Needs 'can be tested' label to test changes
| 2021-02-04T19:56:17 | 2021-02-04T19:56:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,602,914 | 20,087 | den-crane | https://github.com/bzed/carbon-schema-to-clickhouse/issues/1
` graphite_rollup <function>avg</function> `
I think avg works incorrectly in graphite_rollup. Because graphite_rollup does not store sum+count.
It just stores values. So instead of 1000/10 = 100, you can get 700/2=350
| 2021-02-04T21:05:04 | 2021-02-04T21:05:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,719,912 | 19,407 | templarzq | too slow???
3.078 | 3.082 | +1.001x | 0.001 | 0.004 | reinterpret_as | 11 | \n SELECT\n reinterpretAsString(a),\n reinterpretAsString(b),\n reinterpretAsString(c),\n reinterpretAsString(d),\n reinterpretAsString(f),\n reinterpretAsString(g),\n\n toUInt64(number) as a,\n toUInt256(number) as b,\n toInt128(number) as ... | 2021-02-05T01:35:18 | 2021-02-05T01:35:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,187,999 | 20,063 | hexiaoting | It seems that my result is wrong?
@zhangjmruc
-199 % 200 should be 1 ?
-199 % -200 should be -199 ?
199 % -200 should be -1 ? | 2021-02-04T10:05:06 | 2021-02-04T10:14:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,233,443 | 17,708 | Draconic | Completely agree.
It would be great to get H3ToGeo at least so data can be easily converted both ways within ClickHouse. | 2021-02-04T11:20:14 | 2021-02-04T11:20:14 | {} | NONE |
ClickHouse | ClickHouse | 773,273,121 | 20,018 | alexey-milovidov | @vitlibar Could you please take a look? | 2021-02-04T12:34:30 | 2021-02-04T12:37:43 | {} | MEMBER |
ClickHouse | ClickHouse | 773,359,863 | 20,043 | den-crane | > I'm also testing it on a very tiny node, 1-core + 2GB RAM if that can help.
Probably this is the reason. My test instance has 128GB.
my settings are
```
SELECT
name,
value
FROM system.settings
WHERE changed
┌─name──────────────────────────────────────┬─value────────────┐
│ optimize_move_to_p... | 2021-02-04T14:41:13 | 2021-02-04T14:41:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,503,389 | 17,278 | bharatnc | @alexey-milovidov I have started working on this. /assign to me ? | 2021-02-04T18:08:47 | 2021-02-04T18:08:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,740,092 | 20,092 | ucasfl | I want to do this. | 2021-02-05T02:30:56 | 2021-02-05T02:30:56 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 765,628,512 | 19,409 | den-crane | can you show ping (latency) from one clickhouse server to another ?
try to add setting connect_timeout_with_failover_ms=1000
```
cat /etc/clickhouse-server/users.d/connect_timeout_with_failover_ms.xml
<?xml version="1.0"?>
<yandex>
<profiles>
<default>
<connect_timeout_with_fail... | 2021-01-22T19:14:34 | 2021-01-22T19:14:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 765,801,877 | 19,392 | den-crane | You can pre-aggregate and find the first 1 and the last 1.
Connected to ClickHouse server version 21.1.1 revision 54443.
```
CREATE TABLE items (
time DateTime,
group_id UInt16,
value UInt32()
) ENGINE = MergeTree()
PARTITION BY toYYYYMM(time)
ORDER BY (group_id, time);
insert into items select ... | 2021-01-23T00:44:01 | 2021-01-23T00:44:01 | {
"+1": 7
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 765,968,115 | 19,491 | kitaisreal | accurateCastOrNull function was not implemented in 20.11 | 2021-01-23T12:40:46 | 2021-01-23T12:40:52 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.