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 | 659,519,079 | 12,494 | akuzm | ```
SELECT 1 WHERE 1 AND arrayJoin([NULL])
2020.07.16 19:17:27.063307 [ 617442 ] {d3acd8a7-3fcb-4cff-9826-4befa98bb4f2} <Error> : Logical error: 'Unknown numeric column of type: DB::ColumnNullable'.
clickhouse-server: /home/akuzm/ch4/ch/src/Common/Exception.cpp:45: DB::Exception::Exception(const std::string &, i... | 2020-07-16T16:18:26 | 2020-07-16T16:18:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,588,296 | 12,522 | akuzm | ```
SELECT toRelativeDayNum(( SELECT dictGetUInt8() ) AS a), NULL, ( SELECT * FROM hdfs(['B:q']) ) AS st, [];
```
sql_fuzzy.pl finds an assertion right away | 2020-07-16T18:22:28 | 2020-07-16T18:22:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,600,169 | 12,460 | akuzm | Another variety, same error on the server:
```
/4/ :) SELECT anyLast(s) FROM (SELECT '1' s UNION ALL SELECT '2') GROUP BY 1 WITH TOTALS HAVING length(anyLast(s)) > arrayJoin([1, 1])
2020.07.16 21:45:40.032881 [ 686655 ] {b63e2f9f-0b48-42b4-b650-84e0f447269b} <Error> : Logical error: 'Invalid number of rows in Ch... | 2020-07-16T18:46:18 | 2020-07-16T18:46:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 660,006,374 | 12,558 | peaksnail | Goods idea, It worth a try | 2020-07-17T09:52:52 | 2020-07-17T09:52:52 | {} | NONE |
ClickHouse | ClickHouse | 658,823,143 | 11,940 | Slach | @KochetovNicolai @alexey-milovidov @filimonov
I catch same stack trace for 20.3 and 20.4 server randomly crashed when I try to insert into Buffer table with many LowCardinality(String) fields aroud 100 queries per second
and use INSERT INTO buffer VALUES
for 20.3 stacktrace
```
2020.07.14 01:45:36.897051 [ 1090... | 2020-07-15T15:04:48 | 2020-07-15T15:04:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,252,193 | 12,523 | alexey-milovidov | Error message is also cryptic: what is "continuation"? | 2020-07-16T08:37:24 | 2020-07-16T08:37:24 | {} | MEMBER |
ClickHouse | ClickHouse | 659,252,508 | 12,523 | alexey-milovidov | @4ertus2 You can simply forbid `arrayJoin` in ON or USING. | 2020-07-16T08:38:04 | 2020-07-16T08:38:04 | {} | MEMBER |
ClickHouse | ClickHouse | 659,330,423 | 12,452 | nikitamikhaylov | Will be continued in #12339 | 2020-07-16T10:44:20 | 2020-07-16T10:44:20 | {} | MEMBER |
ClickHouse | ClickHouse | 659,336,109 | 12,536 | alesapin | Related fix https://github.com/ClickHouse/ClickHouse/pull/12545 | 2020-07-16T10:57:09 | 2020-07-16T10:57:09 | {} | MEMBER |
ClickHouse | ClickHouse | 659,408,198 | 12,549 | alesapin | Actually test 01318 failed because someone stopped merges globally. It's already fixed in tests. But the new version is much more reliable. | 2020-07-16T13:22:40 | 2020-07-16T13:22:40 | {} | MEMBER |
ClickHouse | ClickHouse | 659,599,748 | 9,948 | vkingnew | it works.
>Clickhouse> alter table t_city rename column city_level TO cityLevel;
>ALTER TABLE t_city
RENAME COLUMN city_level TO cityLevel
Ok.
0 rows in set. Elapsed: 0.018 sec. | 2020-07-16T18:45:24 | 2020-07-16T18:45:24 | {} | NONE |
ClickHouse | ClickHouse | 658,855,094 | 4,456 | alesapin | I think this issue can be closed. | 2020-07-15T16:08:49 | 2020-07-15T16:08:49 | {} | MEMBER |
ClickHouse | ClickHouse | 659,242,593 | 12,508 | alesapin | `00955_test_final_mark` suspicious flap with database atomic https://clickhouse-test-reports.s3.yandex.net/12294/e35c49faa2ad63466ada9f29ea08a9fa3a022ba8/functional_stateless_tests_(release,_databaseatomic).html#fail1 | 2020-07-16T08:20:23 | 2020-07-16T08:20:23 | {} | MEMBER |
ClickHouse | ClickHouse | 659,439,657 | 12,547 | den-crane | ```
select department_id, employ_id, y.1.1 sal_day, y.1.2 prod, y.2.1 next_day, y.2.2 nex_prod from (
select department_id, employ_id,
arrayJoin(arrayMap(i-> (x[i],x[i+1]),arrayEnumerate(arraySort(j->j.1, groupArray((sal_day, product))) as x))) as y
from
(select 'a' as department_id, 1 as employ_id, 'milk1' a... | 2020-07-16T14:12:50 | 2020-07-16T14:14:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,529,815 | 12,512 | Deninc | @den-crane thanks for the information, but maybe this should be noted in the documentation? It took me hours to figure this out. So it might help others.
Either add [here](https://clickhouse.tech/docs/en/sql-reference/statements/create/dictionary/) or [here](https://clickhouse.tech/docs/en/sql-reference/dictionaries... | 2020-07-16T16:36:49 | 2020-07-16T16:36:49 | {} | NONE |
ClickHouse | ClickHouse | 659,591,101 | 12,546 | akuzm | Docs check passed, I fixed most of the comments except removing dashes, and checked how it looks on the test site -- will merge now. | 2020-07-16T18:28:10 | 2020-07-16T18:28:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,632,485 | 12,508 | alexey-milovidov | 00992_system_parts_race_condition_zookeeper:
https://clickhouse-test-reports.s3.yandex.net/12426/6df282e81397680c9f8fd5f40bf1108764d13c93/functional_stateless_tests_(debug).html#fail1 | 2020-07-16T19:49:44 | 2020-07-16T19:49:44 | {} | MEMBER |
ClickHouse | ClickHouse | 659,697,891 | 12,481 | oandrew | Could you provide a file sample? | 2020-07-16T21:59:40 | 2020-07-16T21:59:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,777,406 | 12,512 | den-crane | Database qualifier is mandatory for dictGet with dictionaries created by `CREATE DICTIONARY`.
It needed to distinguish old XML and new DDL dictionaries. | 2020-07-15T13:46:06 | 2020-07-15T13:46:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,780,761 | 12,516 | hydra-yl | > Most probable you have not deterministic result of the query. It's not correlated to `OUTER` keyword for sure. This word is just ignored at parser stage of query analisys.
> In future version it could force `ALL JOIN` to override `set join_default_strictness=ANY`. But for now parser reads and ignores `OUTER` keyword... | 2020-07-15T13:52:06 | 2020-07-15T13:52:06 | {} | NONE |
ClickHouse | ClickHouse | 658,815,136 | 12,524 | den-crane | probably it's supported only in 20.4+
```
20.6.1.4066
GRANT dictGet ON default.item_dict TO test0
Ok.
``` | 2020-07-15T14:50:47 | 2020-07-15T14:51:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,825,008 | 12,527 | alex-zaitsev | It works if configured this way, some dummy line is required:
```
<databases>
<default>
<table name="dummy"><filter>1</filter></table>
<table name='test.abc'><filter>user='test'</filter></table>
</default>
</databases>
```
| 2020-07-15T15:08:36 | 2020-07-15T15:08:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,868,014 | 12,464 | alexey-milovidov | > Yandex synchronization check (only for Yandex employees)
Broken in https://github.com/ClickHouse/ClickHouse/pull/12458 | 2020-07-15T16:33:11 | 2020-07-15T16:33:11 | {} | MEMBER |
ClickHouse | ClickHouse | 659,226,622 | 12,506 | alexey-milovidov | Now it should finish. | 2020-07-16T07:52:04 | 2020-07-16T07:52:04 | {} | MEMBER |
ClickHouse | ClickHouse | 659,402,784 | 12,546 | akuzm | Is there some rationale for avoiding em dashes? I like them how I wrote them. Not sure `--` is properly replaced by em dash in the final page, have to check. | 2020-07-16T13:14:09 | 2020-07-16T13:14:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,595,402 | 11,422 | vkingnew | syntax highligting to clickhouse-client using ReplXX，this function,cloud you give a demo to show it ? | 2020-07-16T18:36:42 | 2020-07-16T18:36:42 | {} | NONE |
ClickHouse | ClickHouse | 659,620,816 | 11,422 | alexey-milovidov | @vkingnew This PR is already merged and basic syntax highlighting is available since 20.5 release. | 2020-07-16T19:26:18 | 2020-07-16T19:26:18 | {} | MEMBER |
ClickHouse | ClickHouse | 659,939,955 | 12,543 | alesapin | no related test errors | 2020-07-17T07:56:08 | 2020-07-17T07:56:08 | {} | MEMBER |
ClickHouse | ClickHouse | 658,835,406 | 4,456 | alesapin | Yes, probably it's a problem with a too-small timeout. It's fixed in https://github.com/ClickHouse/ClickHouse/pull/10990. Also we had a lot of fixes in the dictionaries loader. | 2020-07-15T15:28:12 | 2020-07-15T15:29:10 | {} | MEMBER |
ClickHouse | ClickHouse | 658,848,038 | 10,913 | 4ertus2 | Need rechek. It could be fixed in PR with join with dictionary optimisation. | 2020-07-15T15:53:58 | 2020-07-15T15:53:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,305,611 | 12,544 | alex-zaitsev | CC @vitlibar | 2020-07-16T09:53:20 | 2020-07-16T09:53:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,903,343 | 12,480 | azat | @alexey-milovidov great, worth changing changelog category to bug fix? | 2020-07-17T06:53:32 | 2020-07-17T06:53:32 | {} | MEMBER |
ClickHouse | ClickHouse | 658,756,074 | 12,524 | cronzevid | Same here, using the exact syntax from [documentation](https://clickhouse.tech/docs/en/sql-reference/statements/grant/#grant-dictget).
Even though it is possible to use `GRANT ALL ON db.* to username`, this works only during runtime. If you have `access_management` enabled, this query would leave clickhouse-server in... | 2020-07-15T13:06:55 | 2020-07-15T13:06:55 | {} | NONE |
ClickHouse | ClickHouse | 658,788,790 | 12,327 | nikitamikhaylov | @achimbab Please, replace `std::vector` with `PODArray` from our codebase. | 2020-07-15T14:06:03 | 2020-07-15T14:06:03 | {} | MEMBER |
ClickHouse | ClickHouse | 659,062,492 | 12,438 | alexey-milovidov | Performance test has shown that QPS on concurrent benchmark is improved.
```
queries/s | 0.892 | 1.035 | 0.161 | x1.161
```
But I did not change anything for it.
And the full performance report did not build.
| 2020-07-15T23:15:16 | 2020-07-15T23:15:16 | {} | MEMBER |
ClickHouse | ClickHouse | 659,155,057 | 12,327 | ghost | @nikitamikhaylov
I will replace std::vector with PODArray.
Thank you very much. | 2020-07-16T04:44:31 | 2020-07-16T04:44:31 | {} | NONE |
ClickHouse | ClickHouse | 659,251,539 | 12,539 | alexey-milovidov | Duplicate? https://github.com/ClickHouse/ClickHouse/issues/8889 | 2020-07-16T08:36:08 | 2020-07-16T08:36:08 | {} | MEMBER |
ClickHouse | ClickHouse | 659,493,889 | 12,327 | nikitamikhaylov | > @nikitamikhaylov
> I will replace std::vector with PODArray.
> Thank you very much.
There are also some problems with style, but I think you will fix them by yourself :) Also I will add some review-comments later. | 2020-07-16T15:37:42 | 2020-07-16T15:38:29 | {} | MEMBER |
ClickHouse | ClickHouse | 659,632,890 | 12,426 | alexey-milovidov | Flaky test is reported.
> Yandex synchronization check (only for Yandex employees) Pending — error, check will be restarted
This check is unreliable (known issue @exprmntr).
| 2020-07-16T19:50:38 | 2020-07-16T19:50:38 | {} | MEMBER |
ClickHouse | ClickHouse | 659,964,851 | 12,545 | alesapin | `Logical error: 'Sort order of blocks violated for column ``, left: UInt64_255, right: UInt64_0.'.` | 2020-07-17T08:40:33 | 2020-07-17T08:40:33 | {} | MEMBER |
ClickHouse | ClickHouse | 659,144,144 | 12,539 | jackpgao | MySQL引擎在最近的版本里偶发报错，跟使用的MySQL版本也有一定的关系，我猜你用的是阿里的RDS？ | 2020-07-16T04:01:43 | 2020-07-16T04:01:43 | {} | NONE |
ClickHouse | ClickHouse | 659,229,558 | 12,423 | alexey-milovidov | Looks like we also have to edit
`yt/yt/server/clickhouse_server/functions/yson_extract.cpp`
in Arcadia. | 2020-07-16T07:57:44 | 2020-07-16T07:57:44 | {} | MEMBER |
ClickHouse | ClickHouse | 659,574,648 | 12,188 | alex-zaitsev | Another option is to introduce partition based mutations, similar to CLEAR COLUMN for example:
```
ALTER TABLE table_name UPDATE/DELETE expr IN PARTITION partition_expr
``` | 2020-07-16T17:57:12 | 2020-07-16T17:57:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,947,592 | 12,543 | alesapin | No related failures. `synchronization check` fixed in master. | 2020-07-17T08:09:35 | 2020-07-17T08:09:35 | {} | MEMBER |
ClickHouse | ClickHouse | 658,815,128 | 12,511 | Crazylearner30 | latest_fail_reason is empty in system.mutations table.
No exceptions found in the log.
here is the create table statement of the table
CREATE TABLE database.test_table (
`column1` UInt32,
`column2` UInt32,
`column3` String,
`column4` String,
`date` Date
) ENGINE = MergeTree
ORDER BY
column3 ... | 2020-07-15T14:50:47 | 2020-07-15T14:50:47 | {} | NONE |
ClickHouse | ClickHouse | 658,866,813 | 12,532 | den-crane | WA (no table aliases)
```
CREATE VIEW vu AS
SELECT
emp.empno,
emp.ename,
emp.job,
emp.sal,
dept.dname,
dept.loc,
dept.deptno
FROM emp
ALL INNER JOIN dept ON emp.deptno = dept.deptno
WHERE deptno = 30
```
duplicate: https://github.com/ClickHouse/ClickHouse/issues/11000 https://github.com/ClickHouse... | 2020-07-15T16:31:06 | 2020-07-15T16:31:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,899,278 | 12,532 | vkingnew | it works.BUT IF the table name too long,without alias,it's a hard work. Wish to support table's alias better.
ANSI SQL support table's alias where create view. | 2020-07-15T17:27:11 | 2020-07-15T17:33:27 | {} | NONE |
ClickHouse | ClickHouse | 657,695,148 | 11,572 | ksverdlov | Same exception on OPTIMIZE TABLE xxx DEDUPLICATE
```
2020.07.13 15:31:14.363796 [ 133 ] {76b755df-95c1-4771-820a-79bc0641c21d} <Error> executeQuery: Code: 365, e.displayText() = DB::Exception: Output of TreeExecutor is not sorted (version 20.5.2.7 (official build)) (from 127.0.0.1:41894) (in query: OPTIMIZE TABLE n... | 2020-07-13T17:38:03 | 2020-07-15T19:09:45 | {} | NONE |
ClickHouse | ClickHouse | 659,171,496 | 12,539 | spihiker | 是官方mysql. mysql5.6的版本。这个比较严重，直接导致CH的服务闪退，导致服务不可用。我认为正常的处理逻辑，应该是查询异常，最多把这个查询kill掉。 | 2020-07-16T05:40:27 | 2020-07-16T05:40:27 | {} | NONE |
ClickHouse | ClickHouse | 659,227,652 | 12,413 | alexey-milovidov | I will modify some unrelated file for this purpose... | 2020-07-16T07:54:01 | 2020-07-16T07:54:01 | {} | MEMBER |
ClickHouse | ClickHouse | 659,455,339 | 12,508 | alesapin | > `00955_test_final_mark` suspicious flap with database atomic https://clickhouse-test-reports.s3.yandex.net/12294/e35c49faa2ad63466ada9f29ea08a9fa3a022ba8/functional_stateless_tests_(release,_databaseatomic).html#fail1
Fixed in https://github.com/ClickHouse/ClickHouse/pull/12549/files | 2020-07-16T14:37:54 | 2020-07-16T14:37:54 | {} | MEMBER |
ClickHouse | ClickHouse | 659,501,143 | 10,019 | aadant | This is a painful bug. We notice a 25-30x performance impact for users having row level security as a result of prewhere optimization disabling. | 2020-07-16T15:49:26 | 2020-07-16T15:49:26 | {} | NONE |
ClickHouse | ClickHouse | 658,755,712 | 12,520 | akuzm | I think I understand it better now after looking at this image... Maybe it's more standard this way indeed (from [this article](https://riptutorial.com/sql/example/22934/join-terminology--inner--outer--semi--anti---)).
 | 2020-07-15T13:51:34 | 2020-07-15T13:51:34 | {} | MEMBER |
ClickHouse | ClickHouse | 658,849,974 | 12,294 | alesapin | Merge after some stateless tests will pass. | 2020-07-15T15:58:16 | 2020-07-15T15:58:16 | {} | MEMBER |
ClickHouse | ClickHouse | 659,050,543 | 12,538 | alexey-milovidov | Docs check Ok. | 2020-07-15T22:38:15 | 2020-07-15T22:38:15 | {} | MEMBER |
ClickHouse | ClickHouse | 659,209,686 | 12,509 | toby080508050805 | In log, our grafana still grep data from system.metrics. Also we have some other data insert to our buffer table via distributed table. | 2020-07-16T07:18:20 | 2020-07-16T07:18:20 | {} | NONE |
ClickHouse | ClickHouse | 659,612,271 | 8,331 | akuzm | AST-based query fuzzer found a minor problem:
```
CREATE WINDOW VIEW wv TO dst WATERMARK toIntervalSecond('2') AS SELECT count(a) AS count, TUMBLE_END(wid) AS w_end FROM mt GROUP BY TUMBLE(timestamp, toIntervalSecond(NULL)) AS wid
clickhouse: ../src/Core/Field.h:702: T &DB::Field::get() [T = std::__1::basic_stri... | 2020-07-16T19:08:51 | 2020-07-16T19:08:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,949,092 | 12,518 | akuzm | Yandex check is unrelated.
Will think about adding a CI config later.
Unresolved problems:
* unw_backtrace segfaults, try to add -fno-omit-framepointer
* have to comment out shared memory in Poco, + enumeration of network interfaces in Poco doesn't work. The latter is fixed in the last release, maybe we'll have t... | 2020-07-15T19:05:29 | 2020-07-15T19:05:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,120,061 | 12,486 | alexey-milovidov | @alesapin
I noticed that it's very easy to get
```
E error: [Errno 113] No route to host
```
message even when I run integration tests on my machine with runner script. | 2020-07-16T02:31:46 | 2020-07-16T02:31:46 | {} | MEMBER |
ClickHouse | ClickHouse | 659,329,774 | 12,524 | nuget1991 | Thanks, we updated to latest version of clickhouse and solved the issue. | 2020-07-16T10:42:53 | 2020-07-16T10:42:53 | {} | NONE |
ClickHouse | ClickHouse | 659,420,477 | 12,547 | den-crane | ```
select department_id, employ_id, y.1 sal_day, y.2 next_day from (
select department_id, employ_id,
arrayJoin(arrayMap(i-> (x[i],x[i+1]),arrayEnumerate(arraySort(groupArray(sal_day)) as x))) as y
from
(select 'a' as department_id, 1 as employ_id, toDate('2019-12-01') as sal_day
UNION ALL
select 'b' as de... | 2020-07-16T13:42:03 | 2020-07-16T13:42:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,852,535 | 12,549 | alexey-milovidov | Ok.
> Stress test (undefined) — Timeout :(
The reason unclear. | 2020-07-17T05:01:18 | 2020-07-17T05:01:18 | {} | MEMBER |
ClickHouse | ClickHouse | 659,937,893 | 12,558 | nvartolomei | Have you considered using a single partition? `PARTITION BY tuple()` or omit it completely. To optimize queries by store use it in order by clause. Ie `ORDER BY storeId`. | 2020-07-17T07:52:43 | 2020-07-17T07:54:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,950,629 | 12,518 | akuzm | If we have a CI build for this, we might as well try to package it as a Termux package. | 2020-07-15T19:06:57 | 2020-07-15T19:06:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,251,080 | 12,539 | alexey-milovidov | You should upgrade to the latest 20.3 patch release (20.3.13), it contains important bugfixes. | 2020-07-16T08:35:17 | 2020-07-16T08:35:17 | {} | MEMBER |
ClickHouse | ClickHouse | 659,309,499 | 12,536 | alesapin | Maybe we can do renames of files at server startup. | 2020-07-16T10:00:58 | 2020-07-16T10:01:05 | {} | MEMBER |
ClickHouse | ClickHouse | 658,773,481 | 12,520 | 4ertus2 | > I think I understand it better now after looking at this image...
Cool picture
| 2020-07-15T13:38:48 | 2020-07-15T13:38:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,806,776 | 12,525 | den-crane | All engines use `order by` table section to do sorting data on disk and collapsing rows during merges.
`primary key` section sets a set of columns which will be used in table primary index (it should be prefix of `order by`)
Basically it's impossible by any engine collapse rows by id and store them in order of ev... | 2020-07-15T14:36:23 | 2020-07-15T14:39:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,862,415 | 12,525 | den-crane | @gane5hvarma you don't have a choice. No engine to fulfill your need. | 2020-07-15T16:23:07 | 2020-07-15T16:23:07 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,898,202 | 12,444 | alexey-milovidov | Almost, because we are using timezones embedded in clickhouse binary. | 2020-07-15T17:25:15 | 2020-07-15T17:25:15 | {} | MEMBER |
ClickHouse | ClickHouse | 659,050,440 | 1,424 | traceon | LDAP authentication for preconfigured users ("Simple Bind" method) will be added with #11234 | 2020-07-15T22:37:55 | 2020-07-15T22:37:55 | {
"hooray": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,486,791 | 12,550 | myrrc | Known caveats (why this PR is still a draft):
The LC column we receive is const, so we cant' call `ReverseIndex::getInsertionPoint` as it modifies the index if it's not present.
So I added the[ const](https://github.com/ClickHouse/ClickHouse/pull/12550/files#diff-afa36e4dc3bd7bc4fea86a88ff11d1e3R524) method which thr... | 2020-07-16T15:26:45 | 2020-07-16T16:59:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,621,649 | 11,422 | alexey-milovidov | This PR implements syntax highlighting based on tokens (lexer).
I also have another PR (not merged) with advanced syntax highlighting. | 2020-07-16T19:27:50 | 2020-07-16T19:27:50 | {} | MEMBER |
ClickHouse | ClickHouse | 658,790,982 | 12,527 | alex-zaitsev | It does not seem to work properly. I tried this:
```
<databases>
<default>
<table name='test.abc'><filter>user='test'</filter></table>
</default>
</databases>
```
And I see in row_policies this:
```
SELECT *
FROM system.row_policies
WHERE name = 'tes... | 2020-07-15T14:09:45 | 2020-07-15T14:09:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,634,861 | 12,515 | alexey-milovidov | > Yandex synchronization check (only for Yandex employees) Pending — error, check will be restarted
This check is unreliable @exprmntr | 2020-07-16T19:54:09 | 2020-07-16T19:54:09 | {} | MEMBER |
ClickHouse | ClickHouse | 658,869,753 | 12,533 | den-crane | https://clickhouse.tech/docs/en/engines/table-engines/mergetree-family/mergetree/#choosing-a-primary-key-that-differs-from-the-sorting-key
https://clickhouse.tech/docs/en/engines/table-engines/mergetree-family/mergetree/#table_engine-mergetree-creating-a-table
```
CREATE TABLE PK
(
`A` Int64,
`B` Int6... | 2020-07-15T16:36:13 | 2020-07-15T16:36:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,007,650 | 12,488 | azat | @filimonov will you take a look? | 2020-07-15T21:01:03 | 2020-07-15T21:01:03 | {} | MEMBER |
ClickHouse | ClickHouse | 658,902,526 | 12,530 | den-crane | repro:
20.6.1.4066
```
HOST2:
CREATE TABLE test12345 (A Int64) ENGINE = MergeTree ORDER BY tuple();
CREATE MATERIALIZED VIEW test12345mv Engine=MergeTree partition by A order by tuple() AS SELECT A FROM test12345;
HOST1:
CREATE TABLE test12345 (A Int64) ENGINE = MergeTree ORDER BY tuple();
CREATE TABL... | 2020-07-15T17:33:23 | 2020-07-15T22:00:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,227,479 | 12,413 | alexey-milovidov | Unfortunately CI did not create a build for this PR
(the reason is clear and it's not an issue). | 2020-07-16T07:53:41 | 2020-07-16T07:56:12 | {} | MEMBER |
ClickHouse | ClickHouse | 659,233,915 | 12,508 | alexey-milovidov | @tavplubix Test "metadata_loading" is using real time for criteria of success - it's inherently flaky. What shall we do? | 2020-07-16T08:06:03 | 2020-07-16T08:06:03 | {} | MEMBER |
ClickHouse | ClickHouse | 659,259,903 | 12,539 | spihiker | 非常感谢！我先升级试试看。 | 2020-07-16T08:50:35 | 2020-07-16T08:50:35 | {} | NONE |
ClickHouse | ClickHouse | 659,299,003 | 10,252 | vvchistyakov | > @vvchistiakov, do you have any further questions?
Nope:) | 2020-07-16T09:40:08 | 2020-07-16T09:40:08 | {} | NONE |
ClickHouse | ClickHouse | 658,784,594 | 12,528 | UnamedRus | Duplicate of #12527 | 2020-07-15T13:58:50 | 2020-07-15T13:58:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,866,410 | 11,576 | alex-zaitsev | @jondeandres , this is scheduled for development already. Should be fixed fairly soon. For now just set merge_with_ttl_timeout to some smaller value, like 1800. | 2020-07-15T16:30:21 | 2020-07-15T16:30:21 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,066,746 | 12,536 | alexey-milovidov | Maybe as a workaround we can autodetect mark format from the file size? | 2020-07-15T23:28:05 | 2020-07-15T23:28:18 | {} | MEMBER |
ClickHouse | ClickHouse | 659,195,106 | 7,690 | Slach | @Enmk any news about this issue? | 2020-07-16T06:47:17 | 2020-07-16T06:47:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,286,283 | 12,536 | alesapin | > Maybe as a workaround we can autodetect mark format from the file size?
It's possible, but I think it can lead to confusing bugs, especially when the marks file is broken. Also in the code, we rely on invariant that if the extension is `.mrk` then the part has fixed marks if `.mrk2` then it has adaptive marks if `... | 2020-07-16T09:19:18 | 2020-07-16T09:19:18 | {} | MEMBER |
ClickHouse | ClickHouse | 659,294,838 | 11,940 | Slach | I tried to reproduce issue in docker-compose environment, but unsuccessful
https://gist.github.com/Slach/a70193200df3e3e8f738f411a836b84a | 2020-07-16T09:31:37 | 2020-07-16T09:31:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,434,575 | 12,547 | lzy305 | @den-crane thanks. another case .
select department_id, employ_id,product, sal_day,
neighbor(sal_day, 1) as next_day
from
(select 'a' as department_id, 1 as employ_id, 'milk1' as product, toDate('2019-12-01') as sal_day
UNION ALL
select 'b' as department_id, 2 as employ_id, 'milk2' as product, toDate('2... | 2020-07-16T14:04:32 | 2020-07-16T14:04:32 | {} | NONE |
ClickHouse | ClickHouse | 659,634,172 | 12,506 | alexey-milovidov | > Yandex synchronization check (only for Yandex employees) Pending — error, check will be restarted
This check is unreliable @exprmntr | 2020-07-16T19:53:16 | 2020-07-16T19:53:16 | {} | MEMBER |
ClickHouse | ClickHouse | 659,634,289 | 12,413 | alexey-milovidov | > Yandex synchronization check (only for Yandex employees) Pending — error, check will be restarted
This check is unreliable @exprmntr | 2020-07-16T19:53:30 | 2020-07-16T19:53:30 | {} | MEMBER |
ClickHouse | ClickHouse | 659,972,238 | 11,740 | jerome-leclercq-contentsquare | we have the same problem in 20.3.8.53 using non-replicated CollapsingMergeTree
however using 20.3.12 seems to work fine | 2020-07-17T08:53:24 | 2020-07-17T08:53:24 | {} | NONE |
ClickHouse | ClickHouse | 658,780,778 | 12,509 | den-crane | Do you observe any messages about incoming connection in clickhouserver.log ?
| 2020-07-15T13:52:09 | 2020-07-15T13:52:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 658,918,839 | 12,532 | den-crane | @vkingnew #11000 #10894 registered as bugs. They will be fixed eventually. Maybe. | 2020-07-15T18:06:55 | 2020-07-15T18:06:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 659,233,006 | 12,508 | alexey-milovidov | Live view fix: https://github.com/ClickHouse/ClickHouse/pull/12519 | 2020-07-16T08:04:18 | 2020-07-16T08:04:18 | {} | MEMBER |
ClickHouse | ClickHouse | 659,241,324 | 12,294 | alesapin | `00955_test_final_mark` not changed in this PR. Flap? | 2020-07-16T08:19:08 | 2020-07-16T08:19:08 | {} | MEMBER |
ClickHouse | ClickHouse | 663,629,574 | 12,753 | den-crane | https://clickhouse.tech/docs/en/sql-reference/functions/json-functions/#functions-for-working-with-json
.... **If there are multiple matching fields, the first occurrence is used.** ....
visitParam -- it's very fast set of functions which parse in manner expected in Yandex and they created to parse a special subs... | 2020-07-24T16:43:35 | 2020-07-24T16:47:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,679,229 | 12,436 | hodgesrm | This feature has been requested by a number of users. Would it work to tie the partition scheme to the underlying storage type? I'm a little concerned that adding unconnected partitioning rules as suggested above adds a lot of complexity to storage implementation, which in turn puts the burden on DBAs to make good dec... | 2020-07-24T18:42:21 | 2020-07-24T18:42:21 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.