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 | 681,750,460 | 13,848 | alesapin | > At present, the upgrade version can solve this problem. We will upgrade the production environment cluster
Yes we had several bugs in early 20.3 versions with different metadata structure. Just try to update and check that issue reproduces.
I'll try to investigate, it reproduces only when the added column has a... | 2020-08-27T08:10:28 | 2020-08-27T08:13:17 | {} | MEMBER |
ClickHouse | ClickHouse | 681,815,660 | 10,373 | azat | @4ertus2 @alexey-milovidov friendly ping, can some one take a look? | 2020-08-27T08:52:38 | 2020-08-27T08:52:38 | {} | MEMBER |
ClickHouse | ClickHouse | 681,856,292 | 14,156 | filimonov | > N3orc13TimezoneErrorE
1) do you use some custom build?
2) Do you have a full stack trace of the error? In logs it should be just below the error, or you can run clickhouse-client with parameter `clickhouse-client --stacktrace` and retry the problematic command. (please paste the stack trace)
3) what is your OS... | 2020-08-27T10:10:17 | 2020-08-27T10:10:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,168,996 | 14,045 | radiatorius | Segmentation fault again ( now without "Warning : Removing temporary directory") after 10 s after service start:
```
2020.08.26 22:37:09.465913 [ 18766 ] {} <Fatal> BaseDaemon: ########################################
2020.08.26 22:37:09.465943 [ 18766 ] {} <Fatal> BaseDaemon: (version 20.5.4.40 (official build), no... | 2020-08-26T23:08:09 | 2020-08-26T23:08:09 | {} | NONE |
ClickHouse | ClickHouse | 681,245,898 | 14,151 | 4ertus2 | Related #14124 | 2020-08-27T01:32:01 | 2020-08-27T01:32:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,810,091 | 14,157 | l1t1 | i found two types of tables arenot supported, are there any other types more ?
https://github.com/ClickHouse/ClickHouse/issues/14114
```
DB::Exception: The ckdb.dzm cannot be materialized, because there is no primary keys..
DB::Exception: Sorting key cannot contain nullable columns.
``` | 2020-08-27T08:49:01 | 2020-08-27T08:50:41 | {} | NONE |
ClickHouse | ClickHouse | 681,932,476 | 14,155 | alexey-milovidov | The overall solution is limited because it does not support ZK clusters with authentication.
It also looks like it does not allow to connect to ZK with TLS.
(FYI in Yandex Cloud all connections are with TLS)
Maybe it will be better to have multiple named ZK clusters in config?
It will open up the possibility to... | 2020-08-27T12:58:58 | 2020-08-27T13:00:26 | {} | MEMBER |
ClickHouse | ClickHouse | 681,980,873 | 14,162 | ghost | @alexey-milovidov
OK, I will search the internet and do benchmark test by myself between clickhouse and other open source databases, such as infiniDB and crateDB.
By the way, if you know about recent benchmark test, please let me know so that I would use the result.
| 2020-08-27T14:22:16 | 2020-08-27T14:27:31 | {} | NONE |
ClickHouse | ClickHouse | 682,074,459 | 12,277 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/issues/13248 | 2020-08-27T17:03:39 | 2020-08-27T17:03:39 | {} | MEMBER |
ClickHouse | ClickHouse | 681,819,424 | 14,045 | radiatorius | Here's what one of the databaseY.tableX tables looks like through the user interface:
| 2020-08-27T08:57:31 | 2020-08-27T08:57:31 | {} | NONE |
ClickHouse | ClickHouse | 681,832,054 | 10,177 | oik741 | @PhantomPhreak checked my table, according to `show create table` it was created using new syntax
I think that it's more related to _format_version_ of table
In my case format_version is 1, according to code latest is 4 | 2020-08-27T09:22:10 | 2020-08-27T09:22:10 | {} | NONE |
ClickHouse | ClickHouse | 681,909,553 | 14,045 | radiatorius | Here are some mutations that cause
**Code: 173, e.displayText() = DB::ErrnoException: Allocator: Cannot mmap 16.00 EiB., errno: 12, strerror: Cannot allocate memory (version 20.6.4.44 (official build))** error:
`UPDATE id = 9, name = 'Contest july 2019', updated_at = toDateTime('2020-08-26 22:11:04') WHE... | 2020-08-27T12:12:26 | 2020-08-27T12:12:26 | {} | NONE |
ClickHouse | ClickHouse | 681,911,443 | 14,120 | alexey-milovidov | Because max_insert_threads is not enabled by default. | 2020-08-27T12:15:43 | 2020-08-27T12:15:43 | {} | MEMBER |
ClickHouse | ClickHouse | 681,945,435 | 14,179 | simPod | Stacktracke
```
0. Poco::Exception::Exception(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, int) @ 0x12405650 in /usr/bin/clickhouse
1. DB::Exception::Exception(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, int) @ 0xa24... | 2020-08-27T13:21:52 | 2020-08-27T13:21:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,969,027 | 14,162 | alexey-milovidov | It's also good to pick a couple of third-party benchmarks from the internet. | 2020-08-27T14:02:52 | 2020-08-27T14:02:52 | {} | MEMBER |
ClickHouse | ClickHouse | 681,973,551 | 13,991 | l1t1 | where is the document of MaterializeMySQL feature? | 2020-08-27T14:10:15 | 2020-08-27T14:10:15 | {} | NONE |
ClickHouse | ClickHouse | 681,990,005 | 14,179 | filimonov | > I can read those parts in the original table, I guess I can exclude 1.?
It depends on how did you get them from the original table. For example, simple `rsync` may easily create some incomplete parts if you copy them directly from the `/var/lib/clickhouse/data/db_name/table_name` folder from a running clickhouse i... | 2020-08-27T14:37:08 | 2020-08-27T14:37:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,150,411 | 14,120 | alexey-milovidov | > Functional stateless tests (unbundled) — fail: 0, passed: 0
Maybe something is wrong with start script. | 2020-08-26T22:11:52 | 2020-08-26T22:11:52 | {} | MEMBER |
ClickHouse | ClickHouse | 681,203,796 | 14,030 | bharatnc | Thank you @alexey-milovidov ! | 2020-08-27T01:11:13 | 2020-08-27T01:11:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,658,931 | 14,157 | liangli1213 | @BohuTANG @zhang2014 , need your confirmation. | 2020-08-27T07:09:00 | 2020-08-27T07:09:00 | {} | NONE |
ClickHouse | ClickHouse | 681,806,858 | 14,157 | l1t1 | the table is created
```
DESCRIBE TABLE dec2
┌─name─────┬─type─────────────────────┬─default_type─┬─default_expression─┬─comment─┬─codec_expression─┬─ttl_expression─┐
│ a │ Int32 │ │ │ │ │ │
│ b │ Nullable(De... | 2020-08-27T08:46:54 | 2020-08-27T08:46:54 | {} | NONE |
ClickHouse | ClickHouse | 681,818,890 | 14,143 | tavplubix | `00834_cancel_http_readonly_queries_on_client_close` with debug build - flacky because of slow stacktraces - 81dc22b1d01b7edaebfc45cc53988b100a45149e | 2020-08-27T08:56:30 | 2020-08-27T08:56:30 | {} | MEMBER |
ClickHouse | ClickHouse | 681,907,518 | 14,173 | alesapin | For some reason, our CI env have ancient version of git and fails with:
```fatal: unrecognized argument: --no-patch```
Without this option, the script is less optimal but still correct. because cannot have `^v` symbol in normal diff of `git show`. | 2020-08-27T12:08:29 | 2020-08-27T12:08:29 | {} | MEMBER |
ClickHouse | ClickHouse | 681,911,858 | 14,120 | alexey-milovidov | I will adapt the test. | 2020-08-27T12:16:28 | 2020-08-27T12:18:15 | {} | MEMBER |
ClickHouse | ClickHouse | 681,921,700 | 13,594 | siradjev | > @sd8089730 We need full info on how to reproduce this issue:
>
> * how to install Oracle;
> * how to install its ODBC driver;
> * how to configure DSN;
Seems like info from #13861 shall fix this one too. | 2020-08-27T12:36:52 | 2020-08-27T12:36:52 | {} | NONE |
ClickHouse | ClickHouse | 681,884,973 | 13,860 | qoega | splitter to #14167
| 2020-08-27T11:16:19 | 2020-08-27T15:41:55 | {} | MEMBER |
ClickHouse | ClickHouse | 681,176,574 | 14,121 | filimonov | May be it can have only dummy column, and inherit the structure from inserted block dynamically?
Describing the structure is also ok, but a bit more useless typing. | 2020-08-26T23:32:34 | 2020-08-26T23:34:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,304,238 | 14,045 | den-crane | mutation not mutations_
`mutation_586563.txt`
> Loading mutation: mutation_3.txt | 2020-08-27T02:24:28 | 2020-08-27T02:24:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,767,936 | 12,508 | alesapin | > @alesapin The file `skip_list.json` is in JSON format - we cannot have comments there. It's very important for development.
> I can add some very long tests to skip under TSan but refuse to do so if I cannot write comments.
https://github.com/ClickHouse/ClickHouse/pull/14159 | 2020-08-27T08:21:50 | 2020-08-27T08:21:50 | {} | MEMBER |
ClickHouse | ClickHouse | 681,995,544 | 13,966 | blinkov | @tac0x2a yes, the last commit looks green, thanks!
For testing locally see: https://github.com/ClickHouse/ClickHouse/blob/master/docs/tools/README.md | 2020-08-27T14:46:07 | 2020-08-27T14:46:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,172,059 | 14,045 | alexey-milovidov | Paste the contents of all mutations_*.txt files here.
Then remove these files from filesystem. | 2020-08-26T23:17:58 | 2020-08-26T23:17:58 | {} | MEMBER |
ClickHouse | ClickHouse | 681,179,313 | 14,139 | bharatnc | > "is lesser" does not sound natural for me (though I'm not a native english speaker)
Me neither ;) I've changed to use `<` symbol in the PR description instead! | 2020-08-26T23:42:53 | 2020-08-26T23:43:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,185,092 | 12,171 | filimonov | @taiyang-li can you explain why it is needed? it will increase the log sizes significantly, will brake the log parsers some people use... it would be ok if there is a good motivation for that, but i can't see it...
You easely can find the file & line number by greping the source code by message fragment. | 2020-08-27T00:04:36 | 2020-08-27T00:08:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,853,144 | 14,153 | filimonov | You use old version, please update first. If the issue reproduces on one of the supported versions please create new report. | 2020-08-27T10:04:00 | 2020-08-27T10:04:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,983,715 | 14,160 | Fanduzi | > Most probably It's cause JOIN/IN over Distributed table special logic https://clickhouse.tech/docs/en/sql-reference/operators/in/#select-distributed-subqueries
>
> You may expect that JOIN over Distributed would result the same as if tables are placed on one node but it's not always true. At least for now. By defa... | 2020-08-27T14:26:49 | 2020-08-27T14:26:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,985,044 | 14,113 | johnskopis | I'm not sure if there are any concerns using shared_ptr from InterserverIOHandler. I noticed there's not many shared pointers used and am wondering if it's frowned upon in this codebase?
I was thinking it would be better to "bind" the credential store to a handler but this would require a bit of effort (handler mana... | 2020-08-27T14:29:03 | 2020-08-27T14:29:03 | {} | NONE |
ClickHouse | ClickHouse | 681,991,037 | 14,179 | simPod | I did simple cp. But I'll explore those statements in further attempts, thanks for tip! | 2020-08-27T14:38:49 | 2020-08-27T14:38:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,887,396 | 6,051 | filimonov | 20.7 - still relevant | 2020-08-27T11:21:55 | 2020-08-27T11:21:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,908,519 | 14,120 | alexey-milovidov | The test `insert_parallel` did not speed up. | 2020-08-27T12:10:17 | 2020-08-27T12:10:17 | {} | MEMBER |
ClickHouse | ClickHouse | 682,036,633 | 14,160 | 4ertus2 | You may try rewrite LEFT JOIN with INNER JOIN and UNION ALL like this
```
SELECT a.id from dwh.app_users a
LEFT JOIN dwh.full_orders b
ON a.id = b.uid
```
```
SELECT a.id FROM dwh.app_users a
GLOBAL INNER JOIN dwh.full_orders b
ON a.id = b.uid
UNION ALL
SELECT a.id FROM dwh.app_users a
WHERE a.id GLOBAL NOT... | 2020-08-27T15:54:58 | 2020-08-27T15:56:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 680,777,666 | 12,155 | Rafael-Luo | one shard two replica also Distributed table is slower than query local table
there are two node
local table trace log
```
Limit
Expression
MergeSorting
PartialSorting
Expression
ParallelAggregating
Expression × 24
Filter
Expression
Expression
Fi... | 2020-08-26T09:46:48 | 2020-08-26T23:45:13 | {} | NONE |
ClickHouse | ClickHouse | 681,797,757 | 14,114 | l1t1 | if there is no 'bad' tables,drop ckdb and recreate it, all tables can be selected.
```
mysql> create database ckdb3;
Query OK, 1 row affected (0.02 sec)
mysql> use ckdb3;
Database changed
```
create db in clickhouse now
```
CREATE DATABASE ckdb3
ENGINE = MaterializeMySQL('127.0.0.1:3306', 'ckdb3', 'root', '... | 2020-08-27T08:41:07 | 2020-08-27T08:41:07 | {} | NONE |
ClickHouse | ClickHouse | 681,799,742 | 14,045 | radiatorius | The same errors are constantly repeated with with different values of databaseY and tableX. Maybe just upload all data to the newly installed server and expect this issue not to happen there again?
`2020.08.27 08:33:37.010166 [ 32604 ] {} <Error> databaseY.tableX: Cannot quickly remove directory /var/lib/clickhouse/... | 2020-08-27T08:42:24 | 2020-08-27T08:42:24 | {} | NONE |
ClickHouse | ClickHouse | 681,857,470 | 14,160 | Fanduzi | > ClickHouse version?
my fault
```
:) select version();
SELECT version()
┌─version()─┐
│ 20.3.5.21 │
└───────────┘
``` | 2020-08-27T10:12:52 | 2020-08-27T10:12:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,916,251 | 12,508 | alexey-milovidov | Massive build failures with "inner CI problem": https://clickhouse-builds.s3.yandex.net/0/478adb75ef53333b334b5b4a0a44c47e6ed52c32/clickhouse_build_check/report.html | 2020-08-27T12:25:24 | 2020-08-27T12:25:24 | {} | MEMBER |
ClickHouse | ClickHouse | 616,842,009 | 5,829 | alexey-milovidov | > It's way better to implement autodetect behaviour.
https://clickhouse.tech/docs/ru/whats-new/extended-roadmap/#internal-replication-auto | 2020-04-20T22:20:20 | 2020-08-27T12:31:14 | {} | MEMBER |
ClickHouse | ClickHouse | 682,084,857 | 12,552 | alexey-milovidov | BTW, how do you want to deal with:
> We also should support having single type of state for slightly different aggregate functions, for example sum and sumIf.
I think that we can convert user provided data type as if it is a synonim for another data type.
For example, the user provide
`AggregateFunction(sumIf... | 2020-08-27T17:23:15 | 2020-08-27T17:25:05 | {} | MEMBER |
ClickHouse | ClickHouse | 681,155,781 | 12,433 | calebeaires | Is this feature implemented? Based on the example above I cant create the table
```sql
CREATE TABLE nullable_key
(k Nullable(int), v int) ENGINE MergeTree ORDER BY k SETTINGS allow_nullable_key = 1;
```
Error
Code: 36, e.displayText() = DB::Exception: Unknown setting allow_nullable_key for storage Me... | 2020-08-26T22:27:04 | 2020-08-26T22:27:21 | {
"heart": 1
} | NONE |
ClickHouse | ClickHouse | 681,172,925 | 14,139 | alexey-milovidov | "is lesser" does not sound natural for me (though I'm not a native english speaker) | 2020-08-26T23:20:45 | 2020-08-26T23:20:45 | {} | MEMBER |
ClickHouse | ClickHouse | 681,640,408 | 14,041 | alesapin | Too many test failures in this version. Bug is very old, and I think we can backport it on demand. | 2020-08-27T06:56:29 | 2020-08-27T06:56:29 | {} | MEMBER |
ClickHouse | ClickHouse | 681,861,554 | 14,112 | vzakaznikov | This PR is ready for review. | 2020-08-27T10:21:31 | 2020-08-27T10:21:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,890,530 | 14,029 | nikitamikhaylov | From fuzz test:
```bash
2020.08.25 17:06:40.579994 [ 26899 ] {a2dd82f3-51f5-42f2-8070-11c03036654e} <Debug> executeQuery: (from [::1]:42316) SELECT * FROM file(( SELECT rankCorrelationSpearman('', '\0') ) AS k, ( SELECT extractAll([-33531624]) ) AS zfxi, -2.33987);
2020.08.25 17:06:40.580677 [ 26899 ] {a2dd82f3-51... | 2020-08-27T11:29:16 | 2020-08-27T11:29:16 | {} | MEMBER |
ClickHouse | ClickHouse | 681,929,529 | 14,167 | alesapin | Awesome | 2020-08-27T12:53:12 | 2020-08-27T12:53:12 | {} | MEMBER |
ClickHouse | ClickHouse | 681,944,038 | 14,179 | filimonov | 1) Maybe parts are broken?
2) check logs
3) add - - stacktrace parameter to clickhouse-client call | 2020-08-27T13:19:16 | 2020-08-27T13:19:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,968,019 | 14,162 | ghost | @alexey-milovidov
Thank you very much. | 2020-08-27T14:01:16 | 2020-08-27T14:01:16 | {} | NONE |
ClickHouse | ClickHouse | 682,006,517 | 13,991 | l1t1 | > It's merged into master, you can download the package here: https://clickhouse-builds.s3.yandex.net/0/cff92c8ad3dea2d96b1dfaadc76c3c10b3d701b2/clickhouse_build_check/report.html
i install the new rpms in the page, the version number is same as the rpms in https://repo.clickhouse.tech/rpm/testing/x86_64
```
#rpm ... | 2020-08-27T15:03:36 | 2020-08-27T15:03:36 | {} | NONE |
ClickHouse | ClickHouse | 682,048,414 | 14,157 | dengc367 | me too. decimal type is not supported when using the MaterializeMySQL engine. @filimonov @zhang2014 @BohuTANG
the stacktrace is below:
2020.08.28 00:01:42.718954 [ 7739 ] {cf4473a4-8b5a-4c3f-8aab-5f642e9bbf03} <Debug> executeQuery: (from [::ffff:127.0.0.1]:33014) select * from price_print;
2020.08.28 00:01:42.7... | 2020-08-27T16:15:46 | 2020-08-27T16:15:46 | {} | NONE |
ClickHouse | ClickHouse | 681,905,486 | 13,864 | filimonov | > Also, note, that you can't set `experimental_use_processors` setting in `SETTINGS` clause of query.
It's not the only setting ignored when you try to set it via SETTING in query, see https://github.com/ClickHouse/ClickHouse/issues/3178 | 2020-08-27T12:04:01 | 2020-08-27T12:04:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 682,042,742 | 14,045 | alesapin | Maybe you can try to install the same version with address sanitizer? You can download it from:
https://clickhouse-builds.s3.yandex.net/12470/9c642ff38f8d3180b88caa75228fa2052ddf0036/clickhouse_build_check/report.html
.html
>
> ```
> E QueryRuntimeException: Client failed! Return code: 243, stderr: Received exception from server (version 20.1.10):
> E ... | 2020-08-27T07:12:30 | 2020-08-27T07:22:32 | {} | MEMBER |
ClickHouse | ClickHouse | 681,766,277 | 14,159 | alesapin | GitHub highlighting is ugly for such hacked json:
| 2020-08-27T08:20:44 | 2020-08-27T08:20:44 | {} | MEMBER |
ClickHouse | ClickHouse | 681,856,784 | 14,160 | filimonov | ClickHouse version? | 2020-08-27T10:11:22 | 2020-08-27T10:11:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,894,036 | 14,160 | 4ertus2 | Most probably It's cause JOIN/IN over Distributed table special logic https://clickhouse.tech/docs/en/sql-reference/operators/in/#select-distributed-subqueries
You may expect that JOIN over Distributed would result the same as if tables are placed on one node but it's not always true. At least for now. By default Cl... | 2020-08-27T11:37:36 | 2020-08-27T11:37:36 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,922,582 | 5,829 | alexey-milovidov | See #14178 | 2020-08-27T12:38:42 | 2020-08-27T12:38:42 | {} | MEMBER |
ClickHouse | ClickHouse | 681,929,319 | 14,155 | alexey-milovidov | Does it respect `remote_host_filter`?
Otherwise we have SSRF and the possibility of DoS via ClickHouse. | 2020-08-27T12:52:48 | 2020-08-27T12:52:48 | {} | MEMBER |
ClickHouse | ClickHouse | 681,956,442 | 13,966 | blinkov | @tac0x2a there're some broken links:
```
WARNING:mkdocs.structure.pages:Documentation file 'getting-started/playground.md' contains a link to 'getting-started/interfaces/jdbc.md' which is not found in the documentation files.
WARNING:mkdocs.structure.pages:Documentation file 'getting-started/playground.md' contains ... | 2020-08-27T13:41:43 | 2020-08-27T13:41:43 | {
"eyes": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,996,648 | 11,125 | timofeytt | If you only need to select a single element from an array, you could do something like the following:
`SELECT arrayElement(range(3), ((rand() % length(range(3))) + 1))` | 2020-08-27T14:47:45 | 2020-08-27T14:47:45 | {} | NONE |
ClickHouse | ClickHouse | 682,096,030 | 14,091 | MyroTk | @vitlibar | 2020-08-27T17:45:49 | 2020-08-27T17:45:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,151,765 | 14,122 | alexey-milovidov | AST fuzzer: https://github.com/ClickHouse/ClickHouse/issues/13281 | 2020-08-26T22:15:31 | 2020-08-26T22:15:31 | {} | MEMBER |
ClickHouse | ClickHouse | 681,181,450 | 14,096 | filimonov | > it was intentional change #10049
> and \n is showed because default format is TSV
>
> I don't see UB here.
Agree :) changed because there is no bug here either :)
Actually the issue not in the newlines but in escaping which happens by default in TSV.
TSVRaw is the answer. | 2020-08-26T23:51:30 | 2020-08-26T23:51:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,669,383 | 14,045 | radiatorius | > mutation_586563.txt (mutation_586562.txt,mutation_586564.txt,mutation_586566.txt - are similar only with different uuid):
```
format version: 1
create time: 2020-02-19 06:40:05
commands: UPDATE source_deleted = 1 WHERE number IN (\'05e78327-1088-4565-a765-211c07f28ba2\', \'ce752309-28f7-4f35-bca4-5b1782f784a3\... | 2020-08-27T07:16:05 | 2020-08-27T07:16:05 | {} | NONE |
ClickHouse | ClickHouse | 681,911,678 | 14,120 | alexey-milovidov | But it helps.
```
milovidov@milovidov-desktop:~/work/ClickHouse/tests/queries/0_stateless$ clickhouse-benchmark <<< "INSERT INTO t SELECT * FROM numbers_mt(1000000000)"
Loaded 1 queries.
Queries executed: 7.
localhost:9000, queries 7, QPS: 6.956, RPS: 6958556803.329, MiB/s: 53089.575, result RPS: 0.000, resu... | 2020-08-27T12:16:09 | 2020-08-27T12:16:09 | {} | MEMBER |
ClickHouse | ClickHouse | 681,916,888 | 7,509 | alexey-milovidov | There are two ways to fix this issue.
1. Generic (still in discusstion): https://github.com/ClickHouse/ClickHouse/issues/12587
2. Provide specialization for integer division and modulo of division for nullable right hand side (local changes in code, no changes in overall design). | 2020-08-27T12:26:45 | 2020-08-27T12:28:44 | {} | MEMBER |
ClickHouse | ClickHouse | 681,967,709 | 14,121 | alexey-milovidov | No... INSERT checks table structure and it is implemented in general way. | 2020-08-27T14:00:42 | 2020-08-27T14:00:42 | {} | MEMBER |
ClickHouse | ClickHouse | 682,002,907 | 12,504 | pankajsoni19 | I am using clickhouse ami on aws `Debian GNU/Linux 9.13 (stretch)`
I recently updated to v20.6.4.44-stable using `apt upgrade`. This has effected the reports that we are using.
When will this be released and available for upgrade on the debian instance? | 2020-08-27T14:57:37 | 2020-08-27T14:57:37 | {} | NONE |
ClickHouse | ClickHouse | 681,201,730 | 13,888 | vladimir-golovchenko | Guys thanks for help ;) | 2020-08-27T01:04:12 | 2020-08-27T01:04:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,299,647 | 11,643 | amosbird | What's the difference between using subqueries order by and "Early Window Function" ?
```
select * from
( select * from ... order by ...)
group by ...
``` | 2020-08-27T02:09:14 | 2020-08-27T02:09:25 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 681,686,128 | 14,045 | radiatorius | We went back to version 20.6.4.44 - uptime has improved significantly but periodically (every 10-15 minutes) an automatic restart occurs with errors like this:
```
(version 20.6.4.44 (official build))
2020.08.27 07:21:33.682627 [ 2403 ] {} <Fatal> BaseDaemon: ########################################
2020.08.27 07... | 2020-08-27T07:27:13 | 2020-08-27T07:56:15 | {} | NONE |
ClickHouse | ClickHouse | 681,809,578 | 14,157 | liangli1213 | @l1t1 , table has been created, but data sync failed. | 2020-08-27T08:48:40 | 2020-08-27T08:48:40 | {} | NONE |
ClickHouse | ClickHouse | 681,917,515 | 12,508 | qoega | 01193_metadata_loading not fixed
https://clickhouse-test-reports.s3.yandex.net/14167/bfe4e94806444d1225d93a083a716ff8d770be3c/functional_stateless_tests_(release,_databaseatomic).html#fail1 | 2020-08-27T12:28:04 | 2020-08-27T12:28:04 | {} | MEMBER |
ClickHouse | ClickHouse | 681,959,920 | 14,110 | akuzm | `toDecimal256(1, 7)` gives `Wrong precision` error on the old server, so we can't compare.
Conversion from `Float32` is almost 6 time slower than from `Float64`, this is just weird:
https://clickhouse-test-reports.s3.yandex.net/14110/7e984f877b2d7e2d031dd26c8bd491daae55e8f5/performance_comparison/report.html#partia... | 2020-08-27T13:47:55 | 2020-08-27T13:47:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,965,466 | 14,110 | 4ertus2 | I'll add a new PR with just a division change | 2020-08-27T13:57:08 | 2020-08-27T13:57:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,967,476 | 13,966 | tac0x2a | @blinkov Thank you for reporting. But, it looks same comment to [this](https://github.com/ClickHouse/ClickHouse/pull/13966#issuecomment-679861857). Is it correct ?
By the way, how can I check broken links on my PC ? | 2020-08-27T14:00:22 | 2020-08-27T14:01:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,148,296 | 12,552 | alexey-milovidov | > Can we do without this part? If we speak about the version of serialized data, it exists only during storage or transfer of data.
Version in type name is needed to correctly deserialize data dumps in Native and other formats (scenarios 5 and 6), and to read tables with stored AggregateFunctions (scenarios 3 and 4)... | 2020-08-26T22:06:15 | 2020-08-26T22:11:19 | {} | MEMBER |
ClickHouse | ClickHouse | 681,170,323 | 13,888 | alexey-milovidov | Looks 100% alright :) | 2020-08-26T23:12:23 | 2020-08-26T23:12:23 | {} | MEMBER |
ClickHouse | ClickHouse | 681,724,376 | 13,848 | filimonov | I've got one more error report about the same:
> After attempting to update a DDL for one of our clickhouse tables, our alter query command halted. We rolled back the column addition, but after restarting one of the hosts we see this error: Table columns structure in ZooKeeper is different from local table | 2020-08-27T07:52:32 | 2020-08-27T07:52:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,847,897 | 14,155 | alesapin | `00834_cancel_http_readonly_queries_on_client_close` known flaky test. | 2020-08-27T09:53:26 | 2020-08-27T09:53:26 | {} | MEMBER |
ClickHouse | ClickHouse | 681,909,195 | 14,159 | alesapin | Ok, tests passed and `00834_cancel_http_readonly_queries_on_client_close` is known to be flaky. | 2020-08-27T12:11:43 | 2020-08-27T12:11:43 | {} | MEMBER |
ClickHouse | ClickHouse | 682,083,488 | 12,552 | alexey-milovidov | @akuzm I understand but actually - using different data types for different formats of binary data of the same aggregate function - still looks very natural for me.
And AggregateFunction is parametric data type. It will be almost the same but differ in version parameter.
And the user can write it without version bu... | 2020-08-27T17:20:25 | 2020-08-27T17:20:25 | {} | MEMBER |
ClickHouse | ClickHouse | 681,172,235 | 14,045 | alexey-milovidov | Still waiting for `dmesg` output and further details about your setup. | 2020-08-26T23:18:30 | 2020-08-26T23:18:30 | {} | MEMBER |
ClickHouse | ClickHouse | 681,180,469 | 12,155 | filimonov | > @filimonov
1) The suggestions above are relevant.
2) you didn't give answers
3) try to rerun your query from the same node where clickhouse server works. | 2020-08-26T23:47:32 | 2020-08-26T23:47:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 681,677,164 | 14,049 | alesapin | ```
AST fuzzer — Assertion `(n >= (static_cast<ssize_t>(pad_left_) ? -1 : 0)) && (n <= static_cast<ssize_t>(this->size()))' failed
```
Not related to changes? To be honest, I cannot find this assertion in log. | 2020-08-27T07:21:16 | 2020-08-27T07:21:30 | {} | MEMBER |
ClickHouse | ClickHouse | 681,180,115 | 14,045 | radiatorius | dmesg output:
```
[ 0.000000] Initializing cgroup subsys cpuset
[ 0.000000] Initializing cgroup subsys cpu
[ 0.000000] Initializing cgroup subsys cpuacct
[ 0.000000] Linux version 3.10.0-1127.18.2.el7.x86_64 (mockbuild@kbuilder.bsys.centos.org) (gcc version 4.8.5 20150623 (Red Hat 4.8.5-39) (GCC) ) #1 ... | 2020-08-26T23:46:06 | 2020-08-26T23:46:06 | {} | NONE |
ClickHouse | ClickHouse | 681,185,051 | 14,045 | radiatorius | There are no lines in the output of **cat /var/log/clickhouse-server/clickhouse-server.log | grep -F '[ 5606 ]'**
also can't find any mutation files **find / -name mutations_*.txt**
| 2020-08-27T00:04:28 | 2020-08-27T00:04:28 | {} | NONE |
ClickHouse | ClickHouse | 681,961,516 | 14,162 | alexey-milovidov | Yes, sure.
Don't forget to mention about the setup (what hardware was used) and keep in mind that some results are outdated (see disclaimer below). Also you can link to instructions to reproduce the results if needed: https://github.com/ClickHouse/ClickHouse/tree/master/benchmark | 2020-08-27T13:50:44 | 2020-08-27T13:50:44 | {} | MEMBER |
ClickHouse | ClickHouse | 681,981,442 | 14,155 | amosbird | @alexey-milovidov reimplemented as suggested. | 2020-08-27T14:23:10 | 2020-08-27T14:23:10 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 681,987,113 | 12,552 | akuzm | > Data type is what drives serialization/deserialization.
This may be true for our code, but in general I'd say that a data type is a set of allowed values + operations on them. Serialization is an orthogonal concept. For example, we have a lot of output formats in which we can read and write an Int64, but it's stil... | 2020-08-27T14:32:25 | 2020-08-27T14:32:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 693,211,366 | 14,858 | alesapin | Test failures caused by bugs in the CI. I'll merge, let's check it in the branch. | 2020-09-16T06:52:43 | 2020-09-16T06:52:43 | {} | MEMBER |
ClickHouse | ClickHouse | 693,264,996 | 14,872 | roycyz | > @roycyz did you register the function? see `registerFunctions.cpp`
thanks for replaying,It worked. | 2020-09-16T08:43:42 | 2020-09-16T09:32:54 | {} | NONE |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.