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 | 664,332,700 | 12,770 | qoega | 1) Current error is thrown after query parsing and it is not trivial to see which part of the query is wrong.
2) No need to add explicit conversions or comparison.
Also see the same behavior in HAVING:
```sql
CREATE TABLE t (`c0` Int32, `c1` UInt8) ENGINE = Memory()
INSERT INTO t values (1,2),(3,4);
:) SELECT... | 2020-07-27T11:09:57 | 2020-07-27T11:09:57 | {} | MEMBER |
ClickHouse | ClickHouse | 664,342,609 | 12,767 | alexey-milovidov | Type `ninja clickhouse` instead of `ninja`. | 2020-07-27T11:32:46 | 2020-07-27T11:32:46 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 664,429,127 | 12,840 | azat | >@azat I have already forgotten why do we limit threads here.
It had been added in #11114
Quote from the comment to the patch from that PR:
```
Before this patch pipeline.setMaxThreads() was called with 0 (since
streams.size() == 0, while pipes.size() == 1), which leads that
pipeline.getMaxThreads() returns m... | 2020-07-27T14:26:33 | 2020-07-27T14:26:33 | {} | MEMBER |
ClickHouse | ClickHouse | 663,563,840 | 12,704 | azat | And #12522 had been merged after CI passed on the #12418 (and that's why tests passed with out_of_range exception) | 2020-07-24T14:22:53 | 2020-07-24T14:22:53 | {} | MEMBER |
ClickHouse | ClickHouse | 663,597,296 | 12,749 | den-crane | All engines collapse rows during merges. Engines like SummingMT AggregatingMT should be queried with sum/sumState (aggregation functions) to finalized pre-aggregated data stored in a table.
Example:
```sql
create table SM(A Int64, F Float64) Engine=SummingMergeTree order by A;
insert into SM values (1, 0.1);
i... | 2020-07-24T15:32:22 | 2020-07-24T16:02:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,809,211 | 12,436 | olegkv | Yes, as far as I understand, using different partition expressions for different volumes kinda solves that problem too (number and frequency of background merges, write amplification, etc), because new data comes to hot storage only where partitions are granular enough, and cold storage receives data only as already or... | 2020-07-25T04:53:24 | 2020-07-25T04:57:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,189,710 | 12,661 | filimonov | > Maybe kafka_skip_broken_messages may help?
> allowSyncAfterError needs to be added to the input format
Error processing in Kafka should happen in a completely different way that it is now.
Now we parse each message separately, and can really skip the broken messages separately (not the broken rows somewhere ... | 2020-07-27T08:09:09 | 2020-07-27T08:11:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,607,680 | 12,750 | den-crane | >SELECT (VisitorCode, groupArrayState(VisitDuration))
because of ( )
try `SELECT VisitorCode, groupArrayState(VisitDuration)`
( ) -- makes a special type - Tuple, syntax sugar for a function -- tuple
```
select 1, 2, tuple(1,2), (1,2)
┌─1─┬─2─┬─tuple(1, 2)─┬─tuple(1, 2)─┐
│ 1 │ 2 │ (1,2) │ (1,2) ... | 2020-07-24T15:55:21 | 2020-07-24T18:01:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,854,898 | 12,759 | siradjev | shall probably use this query as workaround to list partitions to be optimized.
```
select distinct table, partition
from system.parts
where active and
(
level=0 OR (table, partition) IN (
select table, partition
from system.parts
where active = 1
group by table, partition
having coun... | 2020-07-25T13:20:28 | 2020-07-25T13:20:28 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 663,881,461 | 12,761 | kssenii | I was reading some related articles and realized that I made a logical mistake in insert query part, so thought I should fix it.
But then I thought why not also add some improvements... | 2020-07-25T17:40:07 | 2020-07-25T17:40:07 | {} | MEMBER |
ClickHouse | ClickHouse | 664,213,784 | 12,417 | filimonov | Duplicate of #12086 | 2020-07-27T08:57:28 | 2020-07-27T08:57:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,248,579 | 12,615 | npetrusic | Yes, I'm using length-delimited messages from the beginning, the problem still persists. We've since worked around to put integer fields first, but it really shouldn't matter, especially since that bug isn't present when sending a protobuf into CH via HTTP. | 2020-07-27T09:51:30 | 2020-07-27T09:51:30 | {} | NONE |
ClickHouse | ClickHouse | 664,317,534 | 10,177 | PhantomPhreak | v20.5, also have thin problem after running
```
alter table <table> MODIFY TTL <column> + interval 6 month;
``` | 2020-07-27T10:52:59 | 2020-07-27T10:52:59 | {} | NONE |
ClickHouse | ClickHouse | 664,330,583 | 12,770 | alexey-milovidov | Works in MySQL:
http://sqlfiddle.com/#!9/0a28f18/494
Does not work in PostgreSQL:
http://sqlfiddle.com/#!17/0a28f/319
If you make a bold assumption that PostgreSQL is better than MySQL, we should close this task. | 2020-07-27T11:05:16 | 2020-07-27T11:05:16 | {} | MEMBER |
ClickHouse | ClickHouse | 664,420,767 | 6,682 | alesapin | It works
| 2020-07-27T14:12:00 | 2020-07-27T14:12:00 | {
"rocket": 1
} | MEMBER |
ClickHouse | ClickHouse | 664,420,906 | 12,815 | alexey-milovidov | Don't pick because of significant changes in code. | 2020-07-27T14:12:15 | 2020-07-27T14:12:15 | {} | MEMBER |
ClickHouse | ClickHouse | 664,421,809 | 12,835 | alexey-milovidov | Don't pick because of significant changes in code. | 2020-07-27T14:13:46 | 2020-07-27T14:13:46 | {} | MEMBER |
ClickHouse | ClickHouse | 663,689,118 | 12,750 | RonanMorgan | Thank you very much ! | 2020-07-24T19:07:41 | 2020-07-24T19:07:41 | {} | NONE |
ClickHouse | ClickHouse | 663,702,698 | 12,749 | RonanMorgan | > Engines like SummingMT AggregatingMT should be queried with sum/sumState
=> ok that's what I thought , the VIsitorCode is not aggregated because it's not used with a merging function, but I was not sure since it was in the group by
=> about the arrayStartingDate, I just wanted to accumulate in an array many va... | 2020-07-24T19:43:11 | 2020-07-24T19:46:47 | {} | NONE |
ClickHouse | ClickHouse | 663,865,064 | 12,757 | den-crane | duplicate https://github.com/ClickHouse/ClickHouse/issues/7187 | 2020-07-25T15:00:38 | 2020-07-25T15:00:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,142,263 | 12,489 | natalia-hw | @den-crane I understand, but from our point of view, this looks like rather as a critical bug, because the query is a run in the wrong database. Especially when some kind of ORM is in usage is really inconvenient to replace the usual table name with table and database name. Are you planning to change this behavior? If... | 2020-07-27T06:15:51 | 2020-07-27T06:15:51 | {} | NONE |
ClickHouse | ClickHouse | 664,216,186 | 11,749 | filimonov | Loosely relates #658 | 2020-07-27T09:02:02 | 2020-07-27T09:02:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,423,971 | 12,804 | alexey-milovidov | Not significant. | 2020-07-27T14:17:29 | 2020-07-27T14:17:29 | {} | MEMBER |
ClickHouse | ClickHouse | 663,765,350 | 12,754 | bobrik | This is also useful for multiple indices, consider:
```
Index `idx_tags` has dropped 1 / 1 granules.
Index `idx_duration` has dropped 0 / 0 granules.
``` | 2020-07-24T22:57:18 | 2020-07-24T22:57:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,011,859 | 12,763 | excitoon | Progress on #12188 | 2020-07-26T16:45:13 | 2020-07-26T16:45:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,194,975 | 12,765 | filimonov | Does S3 uses persistent connections (HTTP Keep-Alive) / some connection pool? | 2020-07-27T08:20:12 | 2020-07-27T08:20:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,276,788 | 12,770 | 4ertus2 | There're lots of cases where we allow UInt8 as Bool and disallow other integers. It's not clear if we should change it and in which direction. I.e. we could have a pseudo-type Bool and force user to CAST to it according to some setting: 'autocast none', 'autocast UInt8' (default), 'autocast all numbers / dates / enums'... | 2020-07-27T10:23:54 | 2020-07-27T10:23:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,411,377 | 12,840 | alexey-milovidov | Perf test? (reading small amount of data from a table with large amount of parts) | 2020-07-27T13:55:17 | 2020-07-27T13:55:17 | {} | MEMBER |
ClickHouse | ClickHouse | 664,426,695 | 12,844 | alexey-milovidov | Not significant. | 2020-07-27T14:22:17 | 2020-07-27T14:22:17 | {} | MEMBER |
ClickHouse | ClickHouse | 663,601,619 | 12,748 | den-crane | try (for test and our understanding the cause of the issue) to query it with disabled features
set experimental_use_processors =0
select ...
set optimize_read_in_order =0
select ...
set max_threads=1
select ...
can you also add _part column (it's a virtual column) to your select query and show the outpu... | 2020-07-24T15:41:46 | 2020-07-24T17:13:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,717,343 | 12,436 | olegkv | Yes, I think it could work.
If I understand correctly, the idea is that partition expression would be specified per volume in storage policy? If so, one volume would contain daily partitions and other volume would contain monthly partitions, for example.
'move_factor' in storage policy could be specified not only as... | 2020-07-24T20:19:58 | 2020-07-24T20:24:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,409,675 | 12,840 | KochetovNicolai | @azat I have already forgotten why do we limit threads here.
I suppose that this is not needed already, because the number of threads is limited by pipeline itself (it shouldn't be more than max streams for every step). | 2020-07-27T13:52:13 | 2020-07-27T13:52:13 | {} | MEMBER |
ClickHouse | ClickHouse | 664,419,182 | 12,823 | alexey-milovidov | Don't pick because of significant changes in code. | 2020-07-27T14:09:07 | 2020-07-27T14:09:07 | {} | MEMBER |
ClickHouse | ClickHouse | 664,419,775 | 12,820 | alexey-milovidov | Not significant. | 2020-07-27T14:10:13 | 2020-07-27T14:10:13 | {} | MEMBER |
ClickHouse | ClickHouse | 664,423,358 | 12,811 | alexey-milovidov | Don't pick because of significant changes in code. | 2020-07-27T14:16:24 | 2020-07-27T14:16:24 | {} | MEMBER |
ClickHouse | ClickHouse | 664,423,637 | 12,807 | alexey-milovidov | Non significant. | 2020-07-27T14:16:57 | 2020-07-27T14:16:57 | {} | MEMBER |
ClickHouse | ClickHouse | 664,424,204 | 12,806 | alexey-milovidov | Not significant. | 2020-07-27T14:17:50 | 2020-07-27T14:17:50 | {} | MEMBER |
ClickHouse | ClickHouse | 664,439,625 | 12,602 | 4ertus2 | Smaller sample
```
SELECT * FROM (
SELECT value
FROM (SELECT 0 as key, 1 as value) as common
INNER JOIN (SELECT 0 as key, 2 as name) trees
ON (common.key = trees.key)
)
UNION ALL
(
SELECT value
FROM (SELECT 1 as key, 3 as value) as common
INNER JOIN (SELECT 1 as key) flowers
... | 2020-07-27T14:43:35 | 2020-07-27T15:06:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,472,452 | 11,648 | qoega | We should mention it in docs with examples | 2020-07-27T15:40:46 | 2020-07-27T15:40:46 | {} | MEMBER |
ClickHouse | ClickHouse | 663,720,752 | 12,749 | den-crane | >I really don't understand why, according to the documentation it can behave like an ANY only if the max size is set
it does not properly documented but Summing and Aggregating behave like ANY for columns which unknown how to calculate if this column is not a part of order KEY (collapser).
Also table Engines have... | 2020-07-24T20:29:00 | 2020-07-24T20:32:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,738,562 | 12,744 | den-crane | ```
create table XC(name String, city String, age Int64) Engine=Memory;
cat xx.csv
name$$city$$age
ss$$hangzhou$$28
dd$$shanghai$$30
cat xx.csv |clickhouse-client -q "insert into XC FORMAT CustomSeparated SETTINGS format_custom_escaping_rule = 'CSV', format_custom_field_delimiter = '\$\$', format_custom_res... | 2020-07-24T21:18:21 | 2020-07-24T21:18:21 | {
"+1": 6,
"rocket": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,795,062 | 12,542 | doseeing | You are right, hard to set default state for column. Any chance to provide like NULL value for column, so maybe users can use function to set default value for their custom requirement?
```
select * from t2 right outer join t3 on t2.value=t3.value;
┌─value─┬─uv_state─┬─uid─┬─t3.value─┐
│ 0 │ NULL │ 1... | 2020-07-25T02:16:00 | 2020-07-25T02:16:00 | {} | NONE |
ClickHouse | ClickHouse | 663,971,963 | 10,495 | alexey-milovidov | @xiashuijun Position of arguments in command line should not matter. | 2020-07-26T10:42:49 | 2020-07-26T10:42:49 | {} | MEMBER |
ClickHouse | ClickHouse | 664,273,298 | 12,062 | filimonov | Let's try to backport it since it's a regression when updating from pre 19.1 | 2020-07-27T10:21:35 | 2020-07-27T10:21:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,421,668 | 12,548 | Akazz | Do not merge just yet! | 2020-07-27T14:13:30 | 2020-07-27T14:13:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,423,104 | 12,814 | alexey-milovidov | Not significant. | 2020-07-27T14:15:55 | 2020-07-27T14:15:55 | {} | MEMBER |
ClickHouse | ClickHouse | 664,424,491 | 12,810 | alexey-milovidov | Not significant. | 2020-07-27T14:18:19 | 2020-07-27T14:18:19 | {} | MEMBER |
ClickHouse | ClickHouse | 663,594,563 | 12,550 | myrrc | Also there is a linked PR #12751 that may result in a CI false negative test | 2020-07-24T15:26:43 | 2020-07-24T15:26:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,422,631 | 12,854 | alexey-milovidov | Don't pick because of significant changes in code. | 2020-07-27T14:15:08 | 2020-07-27T14:15:08 | {} | MEMBER |
ClickHouse | ClickHouse | 664,472,627 | 11,647 | qoega | We should mention it in docs with examples | 2020-07-27T15:41:08 | 2020-07-27T15:41:08 | {} | MEMBER |
ClickHouse | ClickHouse | 664,323,184 | 12,770 | alexey-milovidov | @qoega Why do you think it will improve usability? | 2020-07-27T10:56:53 | 2020-07-27T10:56:53 | {} | MEMBER |
ClickHouse | ClickHouse | 664,388,506 | 11,883 | alesapin | Closed in favor of https://github.com/ClickHouse/ClickHouse/pull/12670 | 2020-07-27T13:14:00 | 2020-07-27T13:14:00 | {} | MEMBER |
ClickHouse | ClickHouse | 664,394,389 | 9,846 | alesapin | Try to update both instances to the latest 20.3 version. If the error persists, just reopen this issue. | 2020-07-27T13:24:31 | 2020-07-27T13:24:31 | {} | MEMBER |
ClickHouse | ClickHouse | 664,426,167 | 12,832 | alexey-milovidov | Not pick because of significant changes in code. | 2020-07-27T14:21:20 | 2020-07-27T14:21:20 | {} | MEMBER |
ClickHouse | ClickHouse | 664,208,819 | 12,748 | filimonov | * Do you insert both minus and plus row in a single insert?
* Do you insert in a single replica? | 2020-07-27T08:47:34 | 2020-07-27T08:47:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,259,636 | 11,672 | quba | Hello everyone. Is there any update regarding this issue? Just curious, since we cannot update CH because of that. Thanks and continue the great work on Clickhouse :) | 2020-07-27T10:11:36 | 2020-07-27T10:11:36 | {} | NONE |
ClickHouse | ClickHouse | 664,416,557 | 12,839 | alexey-milovidov | It's a feature, not a bugfix. | 2020-07-27T14:04:23 | 2020-07-27T14:04:23 | {} | MEMBER |
ClickHouse | ClickHouse | 664,417,852 | 12,828 | alexey-milovidov | This is rare race condition. | 2020-07-27T14:06:42 | 2020-07-27T14:06:42 | {} | MEMBER |
ClickHouse | ClickHouse | 664,446,572 | 12,768 | abyss7 | Please, describe properly what is your cluster description? What are the "copies" - is it replicas?
If they are replicas of some ReplicatedMergeTree, then the ALTER TABLE query is executed only on the leader replica - all other replicas gets changes using another mechanism. For more details you can read this descrip... | 2020-07-27T14:55:43 | 2020-07-27T14:55:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,646,793 | 12,748 | dsigitov | > try (for test and our understanding the cause of the issue) to query it with disabled features
Unfortunately, none of these flags did help. I also tried all 8 possible combinations of them (0|1 for boolean flags and 1|8 for max_threads).
> can you also add _part column (it's a virtual column) to your select que... | 2020-07-24T17:24:34 | 2020-07-24T17:24:34 | {} | NONE |
ClickHouse | ClickHouse | 663,749,264 | 12,589 | bobrik | I checked with v20.7.1.4140-testing and it has logging in place, so it must be something about thread pools that hides logs. | 2020-07-24T21:53:24 | 2020-07-24T21:53:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,804,958 | 12,756 | sundy-li | fixes #12614 | 2020-07-25T04:06:11 | 2020-07-25T05:29:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,814,791 | 12,436 | olegkv | @hodgesrm something for your consideration, unrelated to storage policies, looks like supporting general partition expressions like:
PARTITION BY Expr1 LIMIT N1, Expr2 LIMIT N2, Expr3 LIMIT N3 ...
where new data always comes to partitions controlled by first expression and then merged into next level only when partit... | 2020-07-25T05:58:25 | 2020-07-25T05:58:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,830,053 | 12,757 | vmihailenco | As a workaround now I rewrite this query:
```sql
SELECT data FROM test ORDER BY time LIMIT 10
```
to these 2 queries
```sql
SELECT project_id, time, id FROM test ORDER BY time LIMIT 10
SELECT data FROM test WHERE project_id = ? AND time = ? AND id = ?
```
and get significant performance boost - from 5 ... | 2020-07-25T08:47:38 | 2020-07-25T08:59:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,150,831 | 12,659 | zhang2014 | cc @alexey-milovidov | 2020-07-27T06:41:02 | 2020-07-27T06:41:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,305,498 | 12,327 | nikitamikhaylov | > @nikitamikhaylov
> I finally pass tests and build checks.
> Docs check doesn't seem to related this PR.
> Please review again. Thank you in advance.
>
> Cc. @alexey-milovidov
Hello, sorry for not answered. I was on vacation that week. | 2020-07-27T10:44:15 | 2020-07-27T10:44:15 | {} | MEMBER |
ClickHouse | ClickHouse | 664,415,106 | 12,849 | alexey-milovidov | Non significant. | 2020-07-27T14:02:10 | 2020-07-27T14:02:10 | {} | MEMBER |
ClickHouse | ClickHouse | 664,464,879 | 12,642 | vzakaznikov | @vitlibar, any chance to get this merged this week? We have other PR's almost ready to be pushed. | 2020-07-27T15:27:13 | 2020-07-27T15:27:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,562,150 | 12,704 | azat | Egh, sorry, `std::out_of_range` inherited from `std::logic_error`, @alexey-milovidov can you revert this please?
gcc10:
```
../src/Server/TCPHandler.cpp: In member function ‘void DB::TCPHandler::runImpl()’:
../src/Server/TCPHandler.cpp:327:9: warning: exception of type ‘std::out_of_range’ will be caught ... | 2020-07-24T14:19:19 | 2020-07-24T14:23:25 | {} | MEMBER |
ClickHouse | ClickHouse | 663,994,774 | 12,760 | alesapin | - [x] Need to check that hard links branch still works. | 2020-07-26T14:28:22 | 2020-07-26T16:02:13 | {} | MEMBER |
ClickHouse | ClickHouse | 664,418,147 | 12,827 | alexey-milovidov | This is not a bug. | 2020-07-27T14:07:13 | 2020-07-27T14:07:13 | {} | MEMBER |
ClickHouse | ClickHouse | 664,424,366 | 12,809 | alexey-milovidov | Not a bug. | 2020-07-27T14:18:07 | 2020-07-27T14:18:07 | {} | MEMBER |
ClickHouse | ClickHouse | 664,493,213 | 6,913 | alesapin | Now ALTER CLEAR COLUMN is disabled for all primary key columns. I think this issue can be moved to the feature request category. https://github.com/ClickHouse/ClickHouse/pull/12951 | 2020-07-27T16:14:05 | 2020-07-27T16:17:51 | {} | MEMBER |
ClickHouse | ClickHouse | 664,496,646 | 12,589 | KochetovNicolai | @bobrik I've build this branch and I can see messages like
```
Used generic exclusion search over index for part 128_3_3_0 with 4353 steps
Index `a` has dropped 152 granules
```
in log. Code is still here:
https://github.com/ClickHouse/ClickHouse/pull/12589/files#diff-d5c8199c63973fc53a48faa0f94942a8R1525
So,... | 2020-07-27T16:20:22 | 2020-07-27T16:20:22 | {} | MEMBER |
ClickHouse | ClickHouse | 663,736,324 | 12,436 | hodgesrm | That's the general idea.
The other part of this is that as you move between volumes you want to control number of parts as well as background merges, on them, which is very closely related to what you are doing. Slower storage should merge less of not at all, as this is expensive and performs poorly, whereas read... | 2020-07-24T21:12:34 | 2020-07-24T21:12:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,769,340 | 12,745 | azat | PVS check does related | 2020-07-24T23:14:38 | 2020-07-24T23:14:38 | {} | MEMBER |
ClickHouse | ClickHouse | 663,913,866 | 12,661 | oandrew | Some ideas:
- To fix CH ingestion:
Set up schema-registry api proxy that dynamically replaces bad namespace in the response body and configure CH to use that instead. The only URL needed is `http://schema-registry/schemas/ids/{id}`. You could use http://nginx.org/en/docs/http/ngx_http_sub_module.html or just write ... | 2020-07-25T22:50:28 | 2020-07-25T22:50:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,001,042 | 12,758 | den-crane | ```
CREATE TABLE nested_test
(
`id` UInt8,
`Attrs` Nested( Key String, Value UInt8) CODEC(Gorilla)
)
ENGINE = MergeTree()
ORDER BY id
DESCRIBE TABLE nested_test
┌─name────────┬─type──────────┬─default_type─┬─default_expression─┬─comment─┬─codec_expression─┬─ttl_expression─┐
│ id ... | 2020-07-26T15:18:19 | 2020-07-26T15:19:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,034,148 | 7,462 | alexey-milovidov | Still relevant in 20.7. | 2020-07-26T20:08:24 | 2020-07-26T20:08:24 | {} | MEMBER |
ClickHouse | ClickHouse | 664,419,473 | 12,821 | alexey-milovidov | Don't pick because of significant changes in code. | 2020-07-27T14:09:41 | 2020-07-27T14:09:41 | {} | MEMBER |
ClickHouse | ClickHouse | 664,419,869 | 12,819 | alexey-milovidov | Not significant. | 2020-07-27T14:10:23 | 2020-07-27T14:10:23 | {} | MEMBER |
ClickHouse | ClickHouse | 664,422,221 | 12,838 | alexey-milovidov | Don't pick because of significant changes in code. | 2020-07-27T14:14:26 | 2020-07-27T14:14:26 | {} | MEMBER |
ClickHouse | ClickHouse | 664,497,570 | 12,589 | KochetovNicolai | Tried build from ci, and there are messages in log. | 2020-07-27T16:22:06 | 2020-07-27T16:22:06 | {} | MEMBER |
ClickHouse | ClickHouse | 664,426,953 | 12,846 | alexey-milovidov | Not pick because of significant changes in code. | 2020-07-27T14:22:43 | 2020-07-27T14:22:43 | {} | MEMBER |
ClickHouse | ClickHouse | 664,210,199 | 12,748 | dsigitov | > Do you insert both minus and plus row in a single insert?
Yes
> Do you insert in a single replica?
Yes, it's a standalone deployment. | 2020-07-27T08:50:15 | 2020-07-27T08:50:15 | {} | NONE |
ClickHouse | ClickHouse | 664,377,554 | 12,770 | filimonov | Questionable. Typing smth like `HAVING SUM(c1) > 0` or `SELECT 1 WHERE (SELECT toUInt16(1)) <> 0` is not a big complication, but make the intents 100% clear and query more expressive. | 2020-07-27T12:53:00 | 2020-07-27T12:53:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,392,750 | 10,083 | alesapin | I think we can close it. If someone encounters the error with more fresh versions (starting from 20.3) just reopen this issue. | 2020-07-27T13:21:30 | 2020-07-27T13:21:30 | {} | MEMBER |
ClickHouse | ClickHouse | 664,418,458 | 12,825 | alexey-milovidov | Not a bug. | 2020-07-27T14:07:45 | 2020-07-27T14:07:45 | {} | MEMBER |
ClickHouse | ClickHouse | 664,424,067 | 12,805 | alexey-milovidov | Not significant. | 2020-07-27T14:17:37 | 2020-07-27T14:17:37 | {} | MEMBER |
ClickHouse | ClickHouse | 664,462,063 | 11,856 | qoega | Better to add examples to docs about UUID | 2020-07-27T15:22:16 | 2020-07-27T15:22:16 | {} | MEMBER |
ClickHouse | ClickHouse | 663,654,629 | 12,753 | millin | Really not a bug, expected behavior. Closed | 2020-07-24T17:43:22 | 2020-07-24T17:43:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,665,798 | 12,707 | ifedorov | Thanks, got it | 2020-07-24T18:09:06 | 2020-07-24T18:09:06 | {} | NONE |
ClickHouse | ClickHouse | 663,971,940 | 10,495 | alexey-milovidov | Check server logs. The server can start up, print error message to log and shut down.
Check access rights to the logs directory.
(See `<logger>` in config). | 2020-07-26T10:42:28 | 2020-07-26T10:42:28 | {} | MEMBER |
ClickHouse | ClickHouse | 664,202,670 | 12,615 | filimonov | Do you use length prefix in messages?
See https://clickhouse.tech/docs/en/interfaces/formats/#protobuf
> ClickHouse inputs and outputs protobuf messages in the length-delimited format.
> It means before every message should be written its length as a varint.
> See also how to read/write length-delimited protobuf ... | 2020-07-27T08:35:15 | 2020-07-27T08:35:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,279,540 | 12,745 | akuzm | Why is the enum to string conversion is the only one that needs this?
E.g. `cast(toNullable('1'), 'Nullable(UInt8)')` or `cast(toNullable('null'), 'Nullable(UInt8)')` work w/o additional flags.
Btw, `cast(toNullable('null'), 'Nullable(Enum(\'a\' = 1))')` isn't going to work with this patch, is it? Not sure it should.... | 2020-07-27T10:26:04 | 2020-07-27T10:26:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,382,305 | 12,769 | tetafro | Is there a way to get the error from CI check? I don't see anything in `docs_output.txt`. | 2020-07-27T13:02:33 | 2020-07-27T13:02:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 664,386,581 | 12,739 | alesapin | > Bug: missing whitespace in the report:
> 
| 2020-07-27T13:10:37 | 2020-07-27T13:10:37 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 664,424,611 | 12,840 | KochetovNicolai | Sure | 2020-07-27T14:18:33 | 2020-07-27T14:18:33 | {} | MEMBER |
ClickHouse | ClickHouse | 664,576,998 | 12,949 | alexey-milovidov | Docs check - broken by @blinkov | 2020-07-27T18:55:00 | 2020-07-27T18:55:00 | {} | MEMBER |
ClickHouse | ClickHouse | 664,621,172 | 12,896 | alexey-milovidov | Not picked due to significant changes in code. | 2020-07-27T20:28:31 | 2020-07-27T20:28:31 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.