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 | 640,944,652 | 11,505 | den-crane | Can you please describe your environment? As I can you have issues with select from hadoop?
Is it immediate error or you query? works for some time before the error? | 2020-06-08T23:44:32 | 2020-06-08T23:44:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,089,736 | 11,484 | alexey-milovidov | system.distribution_queue https://clickhouse-test-reports.s3.yandex.net/11536/4a09c822d9305ef78d7fac939b96bf316ce97cdf/functional_stateless_tests_(ubsan).html#fail1 @azat | 2020-06-09T07:28:09 | 2020-06-09T07:28:09 | {} | MEMBER |
ClickHouse | ClickHouse | 641,364,659 | 11,540 | alesapin | Steps to reproduce:
1. Try to read something with AIO: `is_aio = true`, `is_pending_data = true`, `is_eof=false`
2. Seek to EOF: `is_pending_data = false`, `is_aio = true`, `is_eof = false`
3. Read something: `is_pending_data = false`, `is_aio = true`, `is_eof = true`
3. Seek backward: `is_pending_data = false`, `i... | 2020-06-09T15:17:28 | 2020-06-09T15:18:04 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 640,863,792 | 11,497 | azat | >It does not work:
This can be fixed by installing `<access_management>1</access_management>` in `users.xml`, not the desired way? (cc @vitlibar ) | 2020-06-08T20:19:14 | 2020-06-08T20:19:14 | {} | MEMBER |
ClickHouse | ClickHouse | 640,864,484 | 11,497 | azat | Another option is to execute this query separately and ignore access denied error | 2020-06-08T20:20:38 | 2020-06-08T20:20:38 | {} | MEMBER |
ClickHouse | ClickHouse | 641,108,969 | 1,697 | alexey-milovidov | @nikitamikhaylov Maybe you can look at this? | 2020-06-09T08:04:29 | 2020-06-09T08:04:29 | {} | MEMBER |
ClickHouse | ClickHouse | 641,124,449 | 1,697 | alexey-milovidov | @millin `This issue is still relevant.` is not a question :) | 2020-06-09T08:33:03 | 2020-06-09T08:33:03 | {
"laugh": 1
} | MEMBER |
ClickHouse | ClickHouse | 640,760,907 | 11,473 | gbakthavatchalam | Oh I see okay. Yeah I am trying to query `2250 (table1) + 2250 (table2) = 4500` columns while joining `table1 and table2` using a `single` column | 2020-06-08T17:15:58 | 2020-06-08T17:15:58 | {} | NONE |
ClickHouse | ClickHouse | 640,930,922 | 9,108 | 4ertus2 | There was an additional fix here #11380 | 2020-06-08T23:00:31 | 2020-06-08T23:00:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,976,802 | 1,017 | alexey-milovidov | #10058 | 2020-06-09T01:39:04 | 2020-06-09T01:39:04 | {} | MEMBER |
ClickHouse | ClickHouse | 641,116,750 | 11,484 | azat | >system.distribution_queue https://clickhouse-test-reports.s3.yandex.net/11536/4a09c822d9305ef78d7fac939b96bf316ce97cdf/functional_stateless_tests_(ubsan).html#fail1 @azat
#11543 | 2020-06-09T08:19:46 | 2020-06-09T08:19:46 | {} | MEMBER |
ClickHouse | ClickHouse | 641,246,122 | 11,551 | YiuRULE | Hum slighty different but maybe related. Ensure as I don't know too much the workflow of ClickHouse with zookeeper during a table creation. the difference between both are on this one, we were able to create the table successfully but lost the metadatas in a way or an another, while on the first issue, we wasn't able t... | 2020-06-09T12:06:10 | 2020-06-09T12:12:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,018,417 | 11,541 | bobrik | Maybe there's a better way to achieve what I want, I'm totally open to that. | 2020-06-09T04:19:31 | 2020-06-09T04:19:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,185,878 | 11,162 | KochetovNicolai | As I see from tests, it works for 20.4 and master.
Probably, some refactoring wasn't backported for other branches. | 2020-06-09T10:08:37 | 2020-06-09T10:08:37 | {} | MEMBER |
ClickHouse | ClickHouse | 641,241,782 | 11,551 | qoega | Duplicate #11432? | 2020-06-09T11:56:04 | 2020-06-09T11:56:04 | {} | MEMBER |
ClickHouse | ClickHouse | 641,302,724 | 11,550 | istarkov | Instead of forcing protocol `prefer-family` seems like fixing all issues
see log what if address family not exists
```
Spider mode enabled. Check if remote file exists.
--2020-06-09 13:43:54-- (try: 3) http://localhost:1/ping
Connecting to localhost (localhost)|::1|:1... failed: Address family not supported by p... | 2020-06-09T13:37:37 | 2020-06-09T13:48:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,804,743 | 11,494 | Turum | Yes, you are right, set socket_timeout in jdbc settings in datagrip for my CH db and my select was executed without an error. Don't you know if such parameter exists for MSSQL odbc?
Thanks. | 2020-06-08T18:44:42 | 2020-06-08T18:44:42 | {} | NONE |
ClickHouse | ClickHouse | 640,819,029 | 11,494 | den-crane | I think you need to set timeout in CH odbc driver URL
https://github.com/ClickHouse/clickhouse-odbc/issues/175#issuecomment-496914645 | 2020-06-08T19:14:00 | 2020-06-08T19:14:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,858,413 | 11,529 | alexey-milovidov | The same should be done for `anyLast`.
What will happen if we have `any(a + b)`? Would it be transformed to `any(a) + any(b)` and is it beneficial? | 2020-06-08T20:07:20 | 2020-06-08T20:07:20 | {} | MEMBER |
ClickHouse | ClickHouse | 640,866,271 | 11,497 | alexey-milovidov | @vitlibar Why don't make the `system.users` table accessible if there is no access_management?
It effectively means that everything is allowed, so it should be allowed to see the list of users. | 2020-06-08T20:24:18 | 2020-06-08T20:24:18 | {} | MEMBER |
ClickHouse | ClickHouse | 640,867,191 | 11,494 | Turum | Thanks for the clue, I think I found the place. It works now, slow but I think this is expectable, since the request goes through 2 odbc drivers, insert in CH should work faster in theory.
Thanks. | 2020-06-08T20:26:11 | 2020-06-08T20:26:35 | {} | NONE |
ClickHouse | ClickHouse | 640,945,814 | 11,501 | den-crane | You can try to use `ALTER TABLE ... DELETE` as part of your day-to-day flow.
Though it's not recommended because `ALTER TABLE ... DELETE` more administrative operation and it was designed for rare massive processing.
An implementation still have issues and a mutation may stuck forever or execute for a long time. | 2020-06-08T23:48:48 | 2020-06-08T23:48:48 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,235,372 | 6,757 | winter7 | How to avoid `Table a doesn't exist` error during `RENAME table a TO c, b TO a`? | 2020-06-09T11:40:51 | 2020-06-09T11:40:51 | {} | NONE |
ClickHouse | ClickHouse | 640,740,938 | 11,484 | alesapin | > > how we can get to this line if query from loop still not finished?
>
> It may happen if clickhouse-client was killed by timeout after clickhouse-server has received the query, so server still executes it after finishing of the loop in test.
Yes, fix is here https://github.com/ClickHouse/ClickHouse/pull/11523 | 2020-06-08T16:38:59 | 2020-06-08T16:38:59 | {} | MEMBER |
ClickHouse | ClickHouse | 640,780,497 | 11,525 | matwey | > Ok.
>
> PS. How did you find it?
> It does not reproduce in any build configuration in CI.
Accidentally. I've tried gcc 10.1, and it quite aggressively finds such kind of things. This morning, I've fixed another one project with the similar issue https://github.com/Xilinx/XRT/pull/3499 | 2020-06-08T17:55:03 | 2020-06-08T17:55:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,903,373 | 11,508 | nvartolomei | Related #1287? | 2020-06-08T21:43:14 | 2020-06-08T21:43:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,924,655 | 11,527 | akuzm | #10867 is not in any release branch yet, so we don't have to backport this one as well, right? I'll add the no-backport tag. | 2020-06-08T22:40:23 | 2020-06-08T22:40:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,069,879 | 11,501 | jeanatpi | @den-crane thanks for your answer. In my case we delete a small portion of the data each day and from what I've tested the `ALTER TABLE ... DELETE` is quick. But the test are not in a production situation and from what you wrote I don't want my `INSERT` to be blocked by the mutation.
Using the ReplacingMergeTree I'v... | 2020-06-09T06:47:39 | 2020-06-09T06:47:39 | {} | NONE |
ClickHouse | ClickHouse | 641,075,095 | 1,697 | millin | I don't know, I have not used this for a long time | 2020-06-09T06:58:58 | 2020-06-09T06:58:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,088,735 | 9,621 | ralphM78 | @den-crane I have tried 20.4.2.9 and now everything looks good. :thumbsup:
Hopefully you are able to find the problem in 2.4.4.18.
Ralph | 2020-06-09T07:26:21 | 2020-06-09T07:26:21 | {} | NONE |
ClickHouse | ClickHouse | 641,174,465 | 11,543 | alexey-milovidov | sleep(1) does not help to avoid false positive test result. | 2020-06-09T09:49:10 | 2020-06-09T09:49:10 | {} | MEMBER |
ClickHouse | ClickHouse | 640,795,494 | 11,507 | nikitamikhaylov | Hello!
Q1: I debug it and created an issue #11526
Q2: There is no option `copier`, it is a typo
Q3: `task-upload-force` is interpreted as a boolean in our codebase
You can contribute to our documentation by clicking the pencil in the top of the site :wink: | 2020-06-08T18:25:43 | 2020-06-08T18:25:43 | {} | MEMBER |
ClickHouse | ClickHouse | 640,860,881 | 11,497 | alexey-milovidov | It does not work:
```
$ clickhouse-client
ClickHouse client version 20.5.1.1.
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 20.5.1 revision 54435.
milovidov-desktop :) Cannot load data for command line suggestions: Code: 497, e.displayText() = DB::Exception: Received fr... | 2020-06-08T20:13:00 | 2020-06-08T20:13:00 | {} | MEMBER |
ClickHouse | ClickHouse | 640,861,359 | 11,497 | alexey-milovidov | I will add a test. | 2020-06-08T20:14:04 | 2020-06-08T20:14:04 | {} | MEMBER |
ClickHouse | ClickHouse | 640,976,385 | 1,697 | alexey-milovidov | This issue is still relevant. | 2020-06-09T01:37:30 | 2020-06-09T01:37:30 | {} | MEMBER |
ClickHouse | ClickHouse | 641,088,879 | 11,536 | alexey-milovidov | 01293_system_distribution_queue - reported @azat | 2020-06-09T07:26:38 | 2020-06-09T07:26:38 | {} | MEMBER |
ClickHouse | ClickHouse | 641,218,179 | 11,293 | alexey-milovidov | Yes, it will be Ok. | 2020-06-09T11:07:07 | 2020-06-09T11:07:07 | {} | MEMBER |
ClickHouse | ClickHouse | 641,323,928 | 11,548 | alesapin | func tests passed | 2020-06-09T14:11:41 | 2020-06-09T14:11:41 | {} | MEMBER |
ClickHouse | ClickHouse | 641,363,556 | 11,507 | Fanduzi | > Hello!
> Q1: I debug it and created an issue #11526
> Q2: There is no option `copier`, it is a typo
> Q3: `task-upload-force` is interpreted as a boolean in our codebase
>
> You can contribute to our documentation by clicking the pencil in the top of the site 😉
thanks! | 2020-06-09T15:15:34 | 2020-06-09T15:15:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,365,270 | 9,621 | den-crane | @ralphM78 thank you. Now we have some confidence that I reproduced your issue but not something else. :) | 2020-06-09T15:18:27 | 2020-06-09T15:19:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,721,642 | 11,484 | tavplubix | > how we can get to this line if query from loop still not finished?
It may happen if clickhouse-client was killed by timeout after clickhouse-server has received the query, so server still executes it after finishing of the loop in test. | 2020-06-08T16:03:03 | 2020-06-08T16:03:03 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 640,929,705 | 11,537 | alexey-milovidov | Now the error message is perfect:
```
Code: 62. DB::Exception: Received from localhost:9000. DB::Exception: Syntax error (columns declaration list): failed at position 1: (a UInt8). Expected one of: identifier, column declaration, column declaration list, list of elements.
``` | 2020-06-08T22:56:41 | 2020-06-08T22:56:41 | {} | MEMBER |
ClickHouse | ClickHouse | 640,971,245 | 10,405 | alexey-milovidov | `SET max_threads = 24` | 2020-06-09T01:18:59 | 2020-06-09T01:18:59 | {} | MEMBER |
ClickHouse | ClickHouse | 641,201,087 | 11,293 | azat | Maybe it is a good idea to make `SYSTEM FLUSH LOGS` query create tables even if the queue is empty? (this will be enough for most cases I guess) | 2020-06-09T10:35:29 | 2020-06-09T10:35:29 | {} | MEMBER |
ClickHouse | ClickHouse | 641,273,519 | 11,550 | istarkov | Have no idea, haven't checked this case | 2020-06-09T12:57:33 | 2020-06-09T12:57:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,301,496 | 11,529 | CurtizJ | Also, please, provide some test cases, that shows performance improvement from this optimization. | 2020-06-09T13:35:27 | 2020-06-09T13:35:27 | {} | MEMBER |
ClickHouse | ClickHouse | 640,821,813 | 11,479 | millin | In my opinion this still needs to be fixed for the LTS branch which is 20.3 | 2020-06-08T19:19:02 | 2020-06-08T19:19:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,912,719 | 11,475 | rtarasov-mg | Yes, of course.
MySQL 8
```
mysql> SHOW TABLE STATUS LIKE 'clients' \G;
*************************** 1. row ***************************
Name: clients
Engine: InnoDB
Version: 10
Row_format: Dynamic
Rows: 415283
Avg_row_length: 833
Data_length: 346013696
Max_d... | 2020-06-08T22:06:57 | 2020-06-08T22:06:57 | {} | NONE |
ClickHouse | ClickHouse | 641,089,125 | 11,536 | alexey-milovidov | Performance tests have failed for reason, cannot merge. | 2020-06-09T07:27:04 | 2020-06-09T07:27:04 | {} | MEMBER |
ClickHouse | ClickHouse | 641,360,400 | 11,555 | filimonov | @alesapin? Shouldn't it be atomic? | 2020-06-09T15:10:19 | 2020-06-09T15:10:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,362,634 | 11,535 | filimonov | Maybe related: https://github.com/ClickHouse/ClickHouse/issues/4432 | 2020-06-09T15:13:59 | 2020-06-09T15:13:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,750,730 | 11,473 | gbakthavatchalam | Hi @den-crane , No luck.
I did try the above approaches but ended up with same error .
Apart from the above, I also set `max_memory_usage = 0` and `max_rows_in_join = 100000`
Memory limit (total) exceeded: would use 28.27 GiB (attempt to allocate chunk of 4208640 bytes), maximum: 28.27 GiB
--
| 2020-06-08T16:58:02 | 2020-06-08T16:58:02 | {} | NONE |
ClickHouse | ClickHouse | 640,774,561 | 11,300 | alexey-milovidov | Ok. These tests are not flaky - definitely related. | 2020-06-08T17:43:07 | 2020-06-08T17:43:15 | {} | MEMBER |
ClickHouse | ClickHouse | 640,776,162 | 11,525 | alexey-milovidov | Ok.
PS. How did you find it?
It does not reproduce in any build configuration in CI. | 2020-06-08T17:46:26 | 2020-06-08T17:46:26 | {} | MEMBER |
ClickHouse | ClickHouse | 640,884,822 | 11,525 | alexey-milovidov | Build Ok. | 2020-06-08T21:02:18 | 2020-06-08T21:02:18 | {} | MEMBER |
ClickHouse | ClickHouse | 640,976,080 | 2,592 | alexey-milovidov | Not @Potya | 2020-06-09T01:36:22 | 2020-06-09T01:36:22 | {} | MEMBER |
ClickHouse | ClickHouse | 641,059,481 | 7,512 | tavplubix | > Can we "alter" a database to `Atomic` engine?
As for now it can be done only manually by modifying metadata files and moving data of tables. Also it's possible to `RENAME TABLE db_ordinary.table_name TO db_atomic.table_name`. But notice that `Atomic` engine is experimetal and backward compatibility may be broken i... | 2020-06-09T06:24:23 | 2020-06-09T06:24:23 | {} | MEMBER |
ClickHouse | ClickHouse | 641,172,004 | 11,162 | KochetovNicolai | Reverted fix from 20.1 and 20.3 because it cause exception `Column already exists` for `OPTIMIZE TABLE` queries:
https://clickhouse-test-reports.s3.yandex.net/9624/3f2cba6824fddf31c30bde8c6f4f860572f4f580/functional_stateless_tests_(release).html#fail1 | 2020-06-09T09:44:22 | 2020-06-09T09:44:22 | {} | MEMBER |
ClickHouse | ClickHouse | 641,314,974 | 11,401 | akuzm | Let's see what we have with muzzy decay enabled. [Perf test result](https://clickhouse-test-reports.s3.yandex.net/11401/bb925d26168ab900e29ce79c00bf0bdce4095659/performance_comparison/report.html#fail1)
No real performance or stability changes.
Call graph for madvise:
[master](https://clickhouse-test-reports.s3.... | 2020-06-09T13:58:10 | 2020-06-09T13:58:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,327,118 | 11,556 | KochetovNicolai | I have occasionally already backported it | 2020-06-09T14:17:09 | 2020-06-09T14:17:09 | {} | MEMBER |
ClickHouse | ClickHouse | 641,332,089 | 11,556 | filimonov | OK, you were faster :) | 2020-06-09T14:24:47 | 2020-06-09T14:24:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,346,711 | 11,546 | den-crane | @dmitryikh Check that all 4 records belong to 1 partition (table partition by). Partitions break insert to several blocks. | 2020-06-09T14:48:33 | 2020-06-09T14:48:33 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,757,677 | 11,496 | alexey-milovidov | Waiting for performance test... | 2020-06-08T17:09:30 | 2020-06-08T17:09:30 | {} | MEMBER |
ClickHouse | ClickHouse | 640,970,891 | 10,405 | alexey-milovidov | @CurtizJ @hombit It can be related to the choice of number of threads.
You can check on different machine with the same value of `max_threads` setting.
`SELECT * FROM system.settings WHERE name = 'max_threads'`
`SET max_threads = ...` | 2020-06-09T01:17:38 | 2020-06-09T01:17:38 | {} | MEMBER |
ClickHouse | ClickHouse | 640,973,713 | 11,484 | alexey-milovidov | @alesapin We have the same case with https://clickhouse-test-reports.s3.yandex.net/9278/54a746e24d1350497a42dee24abe146dcb274e1e/functional_stateless_tests_(debug).html#fail1
PS. And is it true that "00992 and 00993 race condition tests. Sometimes fails to remove directories" are also fixed? | 2020-06-09T01:27:35 | 2020-06-09T01:27:35 | {} | MEMBER |
ClickHouse | ClickHouse | 641,037,742 | 11,501 | andrea-s | It would be nice if ClickHouse would set a "to be deleted" flag on rows that are awaiting deletion, and those would be transparently skipped by all SELECT queries (time of execution for alter delete would not be instant, but rather the time needed to set the flag).
Do we know if this was ever considered? | 2020-06-09T05:28:42 | 2020-06-09T05:28:42 | {} | NONE |
ClickHouse | ClickHouse | 641,101,524 | 11,535 | selitsky | The same problem with sum(if()):
```
select a
from (
with a+1 as aa
-- , sumIf(c, b > 0) as aaif
, sum(if(b>0,c,0)) as aaif2
select a, aaif2
FROM
(select 1 as a, 2 as b, 3 as c)
GROUP BY a
) as V;
```
`Code: 47, e.displayText() = DB::Exception: Unknown column: if(... | 2020-06-09T07:50:28 | 2020-06-09T07:51:47 | {} | NONE |
ClickHouse | ClickHouse | 641,124,888 | 11,543 | azat | >Single sleep(1) invocation is questionable.
Ok (anyway this sleep() was to avoid false-positive on CI), then it should be fixed in a more reliable way:
https://github.com/ClickHouse/ClickHouse/pull/11543/files#diff-22ff8fc4563d14b4e6d84d1d2510d8f2R16 | 2020-06-09T08:33:53 | 2020-06-09T08:33:53 | {} | MEMBER |
ClickHouse | ClickHouse | 641,188,896 | 11,517 | CurtizJ | This feature is already in development #10930. It will be possible to apply `DISTINCT` for any aggregate function. | 2020-06-09T10:13:45 | 2020-06-09T10:15:08 | {
"hooray": 1
} | MEMBER |
ClickHouse | ClickHouse | 641,278,513 | 11,550 | istarkov | Will check | 2020-06-09T13:07:01 | 2020-06-09T13:07:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,100,767 | 11,513 | 4ertus2 | It looks like it's not possible to run PredicateExpressionsOptimizer before join rewriter if thre're several joins in view. | 2020-06-09T07:49:05 | 2020-06-09T07:49:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,129,250 | 11,544 | blinkov | cc @zhang2014 @sundy-li | 2020-06-09T08:41:13 | 2020-06-09T08:41:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,183,977 | 11,484 | alesapin | > > @alesapin We have the same case with https://clickhouse-test-reports.s3.yandex.net/9278/54a746e24d1350497a42dee24abe146dcb274e1e/functional_stateless_tests_(debug).html#fail1
> > PS. And is it true that "00992 and 00993 race condition tests. Sometimes fails to remove directories" are also fixed?
>
> No timeout ... | 2020-06-09T10:05:26 | 2020-06-09T10:05:26 | {} | MEMBER |
ClickHouse | ClickHouse | 640,966,191 | 11,539 | alexey-milovidov | Affected versions:
master
20.4
20.3
20.1.12
19.14 (difficult to reproduce, only after 52 queries)
19.1.16 (with another error message: "Can't adjust last granule because it has 8192rows, but try to subtract 49152 rows.")
18.16 (the same error message as 19.1) | 2020-06-09T01:02:02 | 2020-06-09T01:10:58 | {} | MEMBER |
ClickHouse | ClickHouse | 640,972,719 | 11,531 | alexey-milovidov | #11538 | 2020-06-09T01:24:07 | 2020-06-09T01:24:07 | {} | MEMBER |
ClickHouse | ClickHouse | 641,086,536 | 7,512 | winter7 | Thanks! I think I will wait for the official ALTER release and keep an eye on this engine. | 2020-06-09T07:21:59 | 2020-06-09T07:21:59 | {} | NONE |
ClickHouse | ClickHouse | 641,024,140 | 11,514 | filimonov | > Here, JSONAsString is applied to the full table. Is it possible to define a JSONEachRow format, yet define a field to be a JSON that should be parsed and set as a String (basically define field as JSONAsString where the JSON object will be serialized into a String).
That issue is about the bug. Your proposal is ac... | 2020-06-09T04:42:32 | 2020-06-09T14:46:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,722,131 | 11,509 | alesapin | Integration tests passed. | 2020-06-08T16:03:59 | 2020-06-08T16:03:59 | {} | MEMBER |
ClickHouse | ClickHouse | 640,858,780 | 11,529 | alexey-milovidov | PS. This pull request includes commits from other pull requests... it's better to rebase it on upstream master. | 2020-06-08T20:08:11 | 2020-06-08T20:08:11 | {} | MEMBER |
ClickHouse | ClickHouse | 640,974,433 | 11,028 | alexey-milovidov | @vitlibar It looks ready. Let's merge with master and then merge into master... | 2020-06-09T01:30:23 | 2020-06-09T01:30:33 | {} | MEMBER |
ClickHouse | ClickHouse | 641,173,919 | 11,162 | alexey-milovidov | @KochetovNicolai Should we also revert from 20.4 and master? | 2020-06-09T09:48:01 | 2020-06-09T09:48:01 | {} | MEMBER |
ClickHouse | ClickHouse | 641,225,629 | 11,546 | alexey-milovidov | Insertion of one block into table of MergeTree family is atomic. This is true in all ClickHouse versions.
Is it true that all records were sent in one block? | 2020-06-09T11:21:20 | 2020-06-09T11:21:20 | {} | MEMBER |
ClickHouse | ClickHouse | 641,233,478 | 11,546 | dmitryikh | I didn't dive into Java Clickhouse driver. But I suppose that this code should insert in one block:
```java
try(Connection connection = dataSource.getConnection()) {
// create batch insert statement
PreparedStatement statement = connection.prepareStatement(this.insertSql);
... | 2020-06-09T11:36:32 | 2020-06-09T11:38:50 | {} | NONE |
ClickHouse | ClickHouse | 640,763,048 | 11,300 | blinkov | > Integration tests of ODBC interaction are broken in this pull request.
> Most likely due to the fact that we check the scenario of segfault inside clickhouse-odbc-bridge.
> But as clickhouse-odbc-bridge is a special process that is expected to crash, we should disable sentry for it.
@alexey-milovidov it might be... | 2020-06-08T17:20:15 | 2020-06-08T17:20:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,865,707 | 11,497 | alexey-milovidov | Yes, it is the desired way, but it's still not enabled by default, and it will not be enabled for old servers,
so the new client will get an error communicating with old server. | 2020-06-08T20:23:11 | 2020-06-08T20:23:11 | {} | MEMBER |
ClickHouse | ClickHouse | 640,873,910 | 11,532 | alexey-milovidov | The bug exists at least since version 19.11. | 2020-06-08T20:39:10 | 2020-06-08T20:39:10 | {} | MEMBER |
ClickHouse | ClickHouse | 640,930,582 | 11,497 | vitlibar | > @vitlibar Why don't make the `system.users` table accessible if there is no access_management?
> It effectively means that everything is allowed, so it should be allowed to see the list of users.
I don't think we should allow every user to know other users' names.
Maybe `system.users` should show only the curren... | 2020-06-08T22:59:25 | 2020-06-08T22:59:25 | {} | MEMBER |
ClickHouse | ClickHouse | 640,967,599 | 4,897 | alexey-milovidov | Not fixed: #11539 | 2020-06-09T01:07:19 | 2020-06-09T01:07:19 | {} | MEMBER |
ClickHouse | ClickHouse | 640,974,791 | 8,301 | alexey-milovidov | @mayankkt9 This task is waiting for you! | 2020-06-09T01:31:38 | 2020-06-09T01:31:38 | {} | MEMBER |
ClickHouse | ClickHouse | 641,109,256 | 11,484 | alesapin | > @alesapin We have the same case with https://clickhouse-test-reports.s3.yandex.net/9278/54a746e24d1350497a42dee24abe146dcb274e1e/functional_stateless_tests_(debug).html#fail1
>
> PS. And is it true that "00992 and 00993 race condition tests. Sometimes fails to remove directories" are also fixed?
No timeout is n... | 2020-06-09T08:05:03 | 2020-06-09T08:05:14 | {} | MEMBER |
ClickHouse | ClickHouse | 641,198,109 | 11,517 | alexey-milovidov | @CurtizJ But the case for countIf should be special with more simple implementation. | 2020-06-09T10:29:43 | 2020-06-09T10:29:43 | {} | MEMBER |
ClickHouse | ClickHouse | 640,717,558 | 11,513 | 4ertus2 | Related #11127 #10481 #9424 (need tests after fix) | 2020-06-08T15:55:56 | 2020-06-08T15:55:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,855,790 | 11,459 | jakesylvestre | Yeah - we might just need to throw some more ram at this either way. `LiveView` is an abstraction over @den-crane's solution since they're both storing results in memory for live querying. Appreciate the response | 2020-06-08T20:01:29 | 2020-06-08T20:01:29 | {} | NONE |
ClickHouse | ClickHouse | 640,925,731 | 10,403 | alexey-milovidov | The latter issue is solved. | 2020-06-08T22:43:55 | 2020-06-08T22:43:55 | {} | MEMBER |
ClickHouse | ClickHouse | 640,928,452 | 9,108 | alexey-milovidov | Yes, it is fixed in #9601. | 2020-06-08T22:52:42 | 2020-06-08T22:52:42 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 640,931,397 | 11,497 | alexey-milovidov | > I don't think we should allow every user to know other users' names.
This is Ok when access management is turned off.
This info is already available in SHOW PROCESSLIST and query_log.
> Also we can provide another option show_access in users.xml which will enable readonly access to access management.
This c... | 2020-06-08T23:01:58 | 2020-06-08T23:01:58 | {} | MEMBER |
ClickHouse | ClickHouse | 641,208,104 | 11,544 | zhang2014 | LGTM | 2020-06-09T10:48:45 | 2020-06-09T10:48:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 641,248,695 | 11,550 | qoega | Do you know how docker sets up container on ipv6only server with `--network=host`? Probably it is easier to try both `-4`/`-6` calls? | 2020-06-09T12:12:10 | 2020-06-09T12:12:27 | {} | MEMBER |
ClickHouse | ClickHouse | 660,018,267 | 12,541 | bohutang | @KochetovNicolai
May i have your review?
Cc @alexey-milovidov | 2020-07-17T10:14:52 | 2020-07-17T10:14:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 660,047,721 | 12,480 | alexey-milovidov | Perf test: #12459 | 2020-07-17T11:11:00 | 2020-07-17T11:11:00 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.