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 | 656,495,146 | 12,317 | zhbdesign | > Network timeout happens before optimise finishes. Check timeout settings both on server and client
config.xml
<?xml version="1.0"?>
<!--
NOTE: User and query level settings are set up in "users.xml" file.
-->
<yandex>
<logger>
<!-- Possible levels: https://github.com/pocoproject/poco/blob/de... | 2020-07-10T05:44:58 | 2020-07-10T05:44:58 | {} | NONE |
ClickHouse | ClickHouse | 656,571,878 | 12,302 | filimonov | > This will not work, since it still can go to `nextImpl` with an error message (basically all you did, is guarded against errors on restart, i.e. when poll() returns the same batch and don't call `poll_batch`)
You can't. `nextImpl` is blocked till next poll (see `allowed` flag). | 2020-07-10T09:07:03 | 2020-07-10T09:07:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,614,694 | 12,373 | vitlibar | Row policies affect each user. Currently there is no way to avoid that. To allow the `default` user to access your table without any filters you have to define one more row policy like
```
CREATE ROW POLICY policy1 ON default.foo USING 1 TO default
```
or, for example,
```
CREATE ROW POLICY policy1 ON defau... | 2020-07-10T10:54:05 | 2020-07-10T10:54:05 | {} | MEMBER |
ClickHouse | ClickHouse | 656,636,940 | 12,081 | flipper86 | is it possible to have a JSONEachRow parser for deep json structure in the next release? e.g. able to parse nested json with following format {header: {key1: value, key2: value}}? :)
| 2020-07-10T11:56:20 | 2020-07-10T11:56:20 | {} | NONE |
ClickHouse | ClickHouse | 656,740,069 | 12,118 | nikitamikhaylov | @4ertus2 Please, look at perf test. Why can't we run test `SELECT COUNT() FROM ints l RIGHT JOIN ints r USING i64 WHERE i32 = 20042` on old server?
https://github.com/ClickHouse/ClickHouse/blob/master/docker/test/performance-comparison/README.md#partial-queries | 2020-07-10T15:37:30 | 2020-07-10T15:37:30 | {} | MEMBER |
ClickHouse | ClickHouse | 656,824,551 | 3,036 | alexey-milovidov | It's Ok if you experience this issue when replicas have different versions.
Replication will stop and it will continue to work when upgrade of the remaining servers will be done.
The system remains available for SELECT queries during rolling update and data remains in correct state.
So, rolling upgrade remains pos... | 2020-07-10T18:38:44 | 2020-07-10T18:39:23 | {} | MEMBER |
ClickHouse | ClickHouse | 656,833,588 | 10,961 | alexey-milovidov | We have figured out that the issue only happened in development version of ClickHouse with your modifications. | 2020-07-10T18:59:27 | 2020-07-10T18:59:27 | {} | MEMBER |
ClickHouse | ClickHouse | 656,462,429 | 12,314 | alexey-milovidov | 00993_system_parts_race_condition_drop_zookeeper - known failure, reported #11815 | 2020-07-10T03:34:04 | 2020-07-10T03:34:04 | {} | MEMBER |
ClickHouse | ClickHouse | 656,494,450 | 12,317 | filimonov | Network timeout happens before optimise finishes. Check timeout settings both on server and client | 2020-07-10T05:42:33 | 2020-07-10T05:42:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,617,281 | 12,373 | vitlibar | Related to https://github.com/ClickHouse/ClickHouse/issues/11400 | 2020-07-10T11:01:22 | 2020-07-10T11:01:22 | {} | MEMBER |
ClickHouse | ClickHouse | 656,795,912 | 12,408 | filimonov | May be related: https://github.com/ClickHouse/ClickHouse/issues/6943 | 2020-07-10T17:33:23 | 2020-07-10T17:33:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,482,971 | 12,293 | filimonov | I though about it and didnt find any obvious use cases when it can lead to problems. Do you know some? | 2020-07-10T04:59:38 | 2020-07-10T04:59:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,563,452 | 12,377 | hellflame | `1970-01-01 08:00:00` and `1970-01-01 08:00:01`
printed out to be 2 thousands years away, which is really unexpected =.= | 2020-07-10T08:48:01 | 2020-07-10T08:48:01 | {} | NONE |
ClickHouse | ClickHouse | 656,570,249 | 12,302 | azat | >check this filimonov@9d46775
This will not work, since it still can go to `nextImpl` with an error message (basically all you did, is guarded against errors on restart, i.e. when poll() returns the same batch and don't call `poll_batch`) | 2020-07-10T09:03:16 | 2020-07-10T09:03:16 | {} | MEMBER |
ClickHouse | ClickHouse | 656,729,256 | 12,179 | alesapin | Suspicious backport... | 2020-07-10T15:16:27 | 2020-07-10T15:16:27 | {} | MEMBER |
ClickHouse | ClickHouse | 656,783,071 | 12,397 | alexey-milovidov | Ok, but why it was needed? @tavplubix ? | 2020-07-10T17:05:01 | 2020-07-10T17:05:01 | {} | MEMBER |
ClickHouse | ClickHouse | 656,813,225 | 12,379 | alexey-milovidov | No need to backport. | 2020-07-10T18:12:54 | 2020-07-10T18:12:54 | {} | MEMBER |
ClickHouse | ClickHouse | 656,491,637 | 12,317 | zhbdesign | Below is the detailed error log, is it due to network reasons?
2020.07.10 13:16:04.270023 [ 1803 ] {0322b521-4c9d-4092-866e-c37ff8d2100b} <Error> executeQuery: Code: 209, e.displayText() = DB::NetException: Timeout: connect timed out: 192.168.10.168:9002 (hadoop-node6:9002) (version 19.16.12.49) (from [::ffff:192.16... | 2020-07-10T05:31:36 | 2020-07-10T05:31:36 | {} | NONE |
ClickHouse | ClickHouse | 656,503,274 | 12,331 | alexey-milovidov | This is by design.
```
/// Will not log operations on system tables (including part_log itself).
/// It doesn't make sense and not allow to destruct PartLog correctly due to infinite logging and flushing,
/// and also make troubles on startup.
if (part_database == DatabaseCatalog::SYSTEM_DATABASE... | 2020-07-10T06:13:22 | 2020-07-10T06:13:22 | {} | MEMBER |
ClickHouse | ClickHouse | 656,555,722 | 12,302 | filimonov | >> Here
> Don't see problem there.
Imagine the situation when parsing of empty message (i.e. error) will produce one row - right now there is no such format, but for Kafka someday may be desired to add be smth like Raw / Plain / Unprocesses / MessageAsString. | 2020-07-10T08:30:26 | 2020-07-10T08:31:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,575,676 | 12,302 | filimonov | > > You can't. nextImpl is blocked till next poll (see allowed flag).
>
> Ugh, tricky, too much flags
Can't disagree :D The problem is that responsibility is spread between Buffer & InputStream, so they need to expose some internals to each other, and that creates a lot of mess. The refactoring I mentioned should... | 2020-07-10T09:15:49 | 2020-07-10T09:15:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,579,703 | 10,897 | sneakythr0ws | Same with ALTER TABLE DROP PARTITION on 20.5.2.7
```
2020.07.10 09:43:14.382172 [ 31263 ] {} <Error> DynamicQueryHandler: Code: 473, e.displayText() = DB::Exception: READ locking attempt on "table" has timed out! (120000ms) Possible deadlock avoided. Client should retry.: While executing Tables, Stack trace (when cop... | 2020-07-10T09:24:45 | 2020-07-10T09:46:24 | {} | NONE |
ClickHouse | ClickHouse | 656,638,537 | 12,377 | filimonov | Duplicate of #9277 #7316 #3470 #1756 etc. | 2020-07-10T12:00:02 | 2020-07-10T12:00:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,766,027 | 12,409 | den-crane | joinGet only supports StorageJoin containing exactly one key.
| 2020-07-10T16:28:18 | 2020-07-10T16:28:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,809,805 | 12,408 | shawel | also maybe related to this #7187 pehraps ??? | 2020-07-10T18:05:18 | 2020-07-10T18:05:18 | {} | NONE |
ClickHouse | ClickHouse | 656,391,603 | 12,335 | CurtizJ | Test failures don't look related. | 2020-07-09T23:01:36 | 2020-07-09T23:04:31 | {} | MEMBER |
ClickHouse | ClickHouse | 656,440,723 | 7,433 | alexey-milovidov | I have tried to reproduce the issue on master but without success.
```
CREATE TABLE cdp_tags (tag_id String, mid_seqs AggregateFunction(groupBitmap, UInt32)) engine=MergeTree() ORDER BY (tag_id) SETTINGS index_granularity=1
INSERT INTO cdp_tags SELECT 'hello', groupBitmapState(rand()) FROM numbers(1000000)
cl... | 2020-07-10T02:08:54 | 2020-07-10T02:08:54 | {} | MEMBER |
ClickHouse | ClickHouse | 656,490,691 | 12,317 | filimonov | Check if problem happens in supported versions (for example 20.3) | 2020-07-10T05:28:03 | 2020-07-10T05:28:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,494,333 | 12,317 | zhbdesign | > Check if problem happens in supported versions (for example 20.3)
> Please create reproducable example, or provide full context (with stacktraces, log fragments etc)
>
> Otherwise we can't help you
Below is the detailed error log, is it due to network reasons?
2020.07.10 13:16:04.270023 [ 1803 ] {0322... | 2020-07-10T05:42:05 | 2020-07-10T05:42:05 | {} | NONE |
ClickHouse | ClickHouse | 656,789,922 | 12,376 | alexey-milovidov | I recommend @akuzm @KochetovNicolai @CurtizJ just to take a look - this is a really weird issue in managing of aggregation states. | 2020-07-10T17:20:02 | 2020-07-10T17:20:20 | {} | MEMBER |
ClickHouse | ClickHouse | 656,831,903 | 11,306 | alexey-milovidov | Something is changed in your configuration during upgrade.
Maybe you have configured ClickHouse with different ZooKeeper cluster that does not have metadata for your tables.
There is a chance that you have installed new ClickHouse version with new, clean ZooKeeper cluster.
Maybe you have removed metadata for your ... | 2020-07-10T18:55:28 | 2020-07-10T18:56:56 | {} | MEMBER |
ClickHouse | ClickHouse | 656,363,837 | 12,371 | den-crane | duplicate https://github.com/ClickHouse/ClickHouse/issues/10994
fixed https://github.com/ClickHouse/ClickHouse/pull/12116 | 2020-07-09T21:38:22 | 2020-07-09T21:38:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,495,390 | 12,317 | zhbdesign | > Network timeout happens before optimise finishes. Check timeout settings both on server and client
user.xml
<?xml version="1.0"?>
<yandex>
<!-- Profiles of settings. -->
<profiles>
<!-- Default settings. -->
<default>
<!-- Maximum memory usage for processing single quer... | 2020-07-10T05:45:50 | 2020-07-10T05:45:50 | {} | NONE |
ClickHouse | ClickHouse | 656,518,074 | 12,264 | alexey-milovidov | It's a side effect of `enable_optimize_predicate_expression`.
If you turn the setting off, the query works as expected.
CC @zhang2014 | 2020-07-10T06:59:28 | 2020-07-10T06:59:35 | {} | MEMBER |
ClickHouse | ClickHouse | 656,742,352 | 12,341 | akuzm | The server in perf test Just Died without any info in log and without any indications of oom killer in dmesg. It was on arithmetic test, which is the most memory-heavy one. Not sure what to do, let's restart and see if it persists... | 2020-07-10T15:41:59 | 2020-07-10T15:41:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,757,918 | 12,402 | qoega | Same problem triggered by GROUP BY query
```
SELECT t0.c1, t0.c0 FROM t0 GROUP BY t0.c1, t0.c0
Caused by: ru.yandex.clickhouse.except.ClickHouseException: ClickHouse exception, code: 49, host: localhost, port: 8123; Code: 49, e.displayText() = DB::Exception: Invalid number of rows in Chunk column String: expected ... | 2020-07-10T16:11:44 | 2020-07-10T16:11:44 | {} | MEMBER |
ClickHouse | ClickHouse | 656,817,758 | 12,397 | tavplubix | However, it's probably better to quote `update_field` if it's not an expression. | 2020-07-10T18:22:52 | 2020-07-10T18:22:52 | {} | MEMBER |
ClickHouse | ClickHouse | 656,489,894 | 12,324 | filimonov | Update to 20.4 or pad/cast the strings manually | 2020-07-10T05:24:59 | 2020-07-10T05:24:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,504,130 | 12,377 | alexey-milovidov | The unix timespamp for 1970-01-01 08:00:00 in Asia/Shanghai is zero.
And we print zero value as all-zeros: `0000-00-00 00:00:00`. | 2020-07-10T06:16:12 | 2020-07-10T06:16:12 | {} | MEMBER |
ClickHouse | ClickHouse | 656,555,386 | 12,307 | alexey-milovidov | Superseded in #12386 | 2020-07-10T08:29:42 | 2020-07-10T08:29:42 | {} | MEMBER |
ClickHouse | ClickHouse | 656,617,421 | 2,261 | sushiMix | is cgroup supported ? It seems some code has been merged from #2325 but it is not clear for me. | 2020-07-10T11:01:45 | 2020-07-10T11:01:45 | {} | NONE |
ClickHouse | ClickHouse | 656,715,985 | 12,405 | Slach | @CurtizJ could you look to this issue? | 2020-07-10T14:49:57 | 2020-07-10T14:49:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,791,279 | 12,294 | alesapin | slitghly better
| 2020-07-10T17:23:14 | 2020-07-10T17:23:14 | {} | MEMBER |
ClickHouse | ClickHouse | 656,825,151 | 5,893 | alexey-milovidov | #11737 | 2020-07-10T18:40:05 | 2020-07-10T18:40:05 | {} | MEMBER |
ClickHouse | ClickHouse | 656,365,388 | 12,342 | genzgd | A couple of other points:
The only tables that display this issue are aggregate tables that are populated with materialized views
One other column appeared to have an issue -- a FixedString(16) column used as a key
The parts created that showed this issue were basically corrupted and unusable. This is a critical bu... | 2020-07-09T21:42:38 | 2020-07-09T21:43:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,441,479 | 7,433 | alexey-milovidov | Also it is strange that our source of Roaring Bitmaps is different from upstream.
Shouldn't we fix / investigate it first? | 2020-07-10T02:12:09 | 2020-07-10T02:12:09 | {} | MEMBER |
ClickHouse | ClickHouse | 656,497,191 | 12,317 | zhbdesign | > Network timeout happens before optimise finishes. Check timeout settings both on server and client
Client used cc.blynk.clickhous The driver is configured as follows:
clickHouseDataSource = new BasicDataSource();
clickHouseDataSource.setDriverClassName (" cc.blynk.clickhouse .ClickHouseDriver");
clickHous... | 2020-07-10T05:52:01 | 2020-07-10T05:52:01 | {} | NONE |
ClickHouse | ClickHouse | 656,543,785 | 12,302 | filimonov | > > and will not allow buffer to read non existing message
>
> Reading is done from nextImpl,
Why should we even TRY to read/parse the message which is not a message but an error?
Give me a min, i'll send a proposal. | 2020-07-10T08:03:15 | 2020-07-10T08:03:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,562,082 | 12,302 | filimonov | Yeah, your changed approach should be ok too :+1: (but it requires one extra iteration on messages, and vector manipulations :P) | 2020-07-10T08:44:47 | 2020-07-10T08:48:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,622,130 | 12,342 | genzgd | Downgrading fixed the issue in the sense that new parts created by the materialized views were not corrupted and merged cleanly. However, the ttms_avg columns that were created under v20.5 were not readable under 20.3 and caused the same merge exceptions/stack traces as the example above. Hardware issues are unlikel... | 2020-07-10T11:14:51 | 2020-07-10T11:14:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,711,912 | 11,770 | 4ertus2 | Continuation is here #12404 | 2020-07-10T14:41:37 | 2020-07-10T14:41:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,749,964 | 12,118 | akuzm | > @4ertus2 Please, look at perf test. Why can't we run test `SELECT COUNT() FROM ints l RIGHT JOIN ints r USING i64 WHERE i32 = 20042` on old server?
> https://github.com/ClickHouse/ClickHouse/blob/master/docker/test/performance-comparison/README.md#partial-queries
`left-server-log.log` says:
```
2020.07.10 08:... | 2020-07-10T15:56:25 | 2020-07-10T15:56:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,804,291 | 12,341 | alexey-milovidov | > without any indications of oom killer in dmesg
AFAIK it just does not work in containers. | 2020-07-10T17:52:37 | 2020-07-10T17:52:37 | {} | MEMBER |
ClickHouse | ClickHouse | 656,400,432 | 12,333 | alexey-milovidov | Accepted. Clear column is not like drop column, we should not argue. | 2020-07-09T23:32:30 | 2020-07-09T23:32:30 | {} | MEMBER |
ClickHouse | ClickHouse | 656,462,155 | 11,815 | alexey-milovidov | It is already mentioned as
> Logical error, can be reproduced with 00993_system_parts_race_condition_drop_zookeeper and https://clickhouse-test-reports.s3.yandex.net/0/90b5e5d2f9edbcd4f01bd596ec919a32368fd716/functional_stateless_tests_(debug).html#fail1 | 2020-07-10T03:33:01 | 2020-07-10T03:33:01 | {} | MEMBER |
ClickHouse | ClickHouse | 656,535,690 | 12,302 | azat | >Here
Don't see problem there.
>or here
This, is a problem, yes (will try get currentTopic and fail)
>If would be really better if poll would return bad flag
Another new flag?
I don't see any reason in this, either it can fail if there is at least on message with an error or silently skip errors
> an... | 2020-07-10T07:44:41 | 2020-07-10T07:44:41 | {} | MEMBER |
ClickHouse | ClickHouse | 656,657,858 | 12,153 | alesapin | backported to 20.3, but doesn't work for some reason. | 2020-07-10T12:46:30 | 2020-07-10T12:46:30 | {} | MEMBER |
ClickHouse | ClickHouse | 656,732,124 | 12,406 | filimonov | The issue reproduces. | 2020-07-10T15:21:57 | 2020-07-10T15:21:57 | {
"+1": 4
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,787,398 | 12,390 | alexey-milovidov | Also let's do one of the following:
- make WHERE clause not optional for DELETE;
- introduce a setting that will forbid omission of WHERE clause in DELETE by default.
`DELETE FROM table` without WHERE is a common source of disasters. | 2020-07-10T17:14:31 | 2020-07-10T17:15:02 | {} | MEMBER |
ClickHouse | ClickHouse | 656,369,572 | 12,373 | MyroTk | The `CREATE ROW POLICY` doesn't need to be assigned to user or role to cause this issue. Using `AS` and `USING` together is enough on their own. | 2020-07-09T21:53:37 | 2020-07-09T21:53:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,503,365 | 12,331 | alexey-milovidov | Please make some clarifications in documentation. | 2020-07-10T06:13:39 | 2020-07-10T06:13:39 | {} | MEMBER |
ClickHouse | ClickHouse | 656,512,563 | 12,262 | alexey-milovidov | The query
```
SELECT 1 != NOT 1;
```
does not parse due to operator priorities.
It is normal.
BTW, it also does not parse in MySQL. Occasionally, the query `SELECT TRUE != NOT TRUE` works in Postgres.
If we add parens:
```
SELECT 1 != NOT (1);
```
The `NOT` will be parsed as a function, not as op... | 2020-07-10T06:43:34 | 2020-07-10T06:43:34 | {} | MEMBER |
ClickHouse | ClickHouse | 656,552,941 | 12,294 | alesapin | 
Not so fast :( | 2020-07-10T08:24:20 | 2020-07-10T08:24:20 | {} | MEMBER |
ClickHouse | ClickHouse | 656,791,412 | 12,408 | shawel | I see! is there a way to speed it up? idk if it is slow on the write side or the read side. it seems to me like it it might be reading
to many unnecessary rows initially. just a hunch though. My machine is 16 cores and 125 GB memory and 2TB EBS
This query climbs to using about 30GB from 2GB or so as the query goes... | 2020-07-10T17:23:34 | 2020-07-10T17:30:59 | {} | NONE |
ClickHouse | ClickHouse | 656,808,731 | 12,397 | tavplubix | > Ok, but why it was needed? @tavplubix ?
Column names in `WHERE` clause were not quoted, so I replaced `writeString` with `writeQuoted` in several places. This line was replaced by mistake. | 2020-07-10T18:02:58 | 2020-07-10T18:02:58 | {} | MEMBER |
ClickHouse | ClickHouse | 656,398,792 | 11,177 | excitoon | Shall be continued in NanoBjorn#1 | 2020-07-09T23:27:05 | 2020-07-09T23:27:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,539,284 | 11,815 | alexey-milovidov | Sporadic build failure: https://clickhouse-builds.s3.yandex.net/12378/a4b35a8a6fe524793e36a09d234ee8a42437f3ca/clickhouse_build_check/build_log_720579080_1594364683.txt | 2020-07-10T07:53:05 | 2020-07-10T07:53:05 | {} | MEMBER |
ClickHouse | ClickHouse | 656,550,218 | 4,567 | 4ertus2 | One more case we should throw an error
```
mysql> select k from (select 1 as k union all select 2 as k) s1 right join (select 1 as k union all
select 0 as k) s2 on s1.k = s2.k;
ERROR 1052 (23000): Column 'k' in field list is ambiguous
``` | 2020-07-10T08:18:06 | 2020-07-10T08:18:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,553,685 | 12,054 | alexey-milovidov | @KochetovNicolai is on vacation, unassigned.
@tavplubix Maybe you can take it? | 2020-07-10T08:26:00 | 2020-07-10T08:26:00 | {} | MEMBER |
ClickHouse | ClickHouse | 656,572,773 | 12,302 | azat | >You can't. nextImpl is blocked till next poll (see allowed flag).
Ugh, tricky, too much flags | 2020-07-10T09:09:06 | 2020-07-10T09:09:06 | {} | MEMBER |
ClickHouse | ClickHouse | 656,697,094 | 12,342 | genzgd | A somewhat related issue -- we only upgrade a single shard on a multi shard cluster, and a distributed query generated the following error. Does that mean all shards must be upgraded to version 20.5 simultaneously?
```
2020.07.09 17:39:28.440083 [ 398558 ] {} <Error> DynamicQueryHandler: Code: 271, e.displayText()... | 2020-07-10T14:11:01 | 2020-07-10T14:11:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,818,499 | 12,402 | alexey-milovidov | Looks like the documented behaviour of Log table engines: https://clickhouse.tech/docs/en/engines/table-engines/log-family/
It's reasonable to rewrite SQLancer scripts to use MergeTree tables.
For small tables, Memory is also Ok.
I will think on how to avoid this issue nevertheless. | 2020-07-10T18:24:33 | 2020-07-10T18:24:33 | {} | MEMBER |
ClickHouse | ClickHouse | 656,829,063 | 11,135 | alexey-milovidov | Ok. I have successfully reproduced this issue on 20.3.8 and checked that it is fixed in 20.3.12 and more recent versions. | 2020-07-10T18:48:57 | 2020-07-10T18:48:57 | {} | MEMBER |
ClickHouse | ClickHouse | 656,419,698 | 11,815 | bohutang | Functional stateless tests (debug)
https://clickhouse-test-reports.s3.yandex.net/12314/260bcb9d795e2ea5607fd6a50481b4c2698c0249/functional_stateless_tests_(debug).html#fail1 | 2020-07-10T00:43:49 | 2020-07-10T00:44:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,492,955 | 12,317 | zhbdesign | Below is metrika.xml.
```
<yandex>
<clickhouse_remote_servers>
<srt_cluster_3shards_2replicas>
<shard>
<internal_replication>true</internal_replication>
<replica>
<host>192.168.10.168</host>
<port>9002</port>
</replica>
<replica>
<host>192.168.10.178</host>
<port>9... | 2020-07-10T05:36:46 | 2020-07-10T05:40:13 | {} | NONE |
ClickHouse | ClickHouse | 656,536,392 | 12,386 | fastio | @alexey-milovidov Thanks for your review for #12307
Follow your suggestion：
* new function name;
* some flags for URI parsing.
I open a new RP, and close the old one #12307
| 2020-07-10T07:46:20 | 2020-07-10T07:46:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,615,954 | 12,036 | 4ertus2 | You may enable `set multiple_joins_rewriter_version = 2` to fix it. We just need some more success stories for second version of rewriter to enable it by default. | 2020-07-10T10:57:39 | 2020-07-10T10:57:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,752,344 | 12,118 | 4ertus2 | This PR adds RIGHT JOIN support for partial merge join so it's normal that the version without it cannot run the query with this JOIN. | 2020-07-10T16:01:01 | 2020-07-10T16:01:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,396,492 | 12,278 | alexey-milovidov | 6 slower performance tests don't look good. | 2020-07-09T23:19:18 | 2020-07-09T23:19:18 | {} | MEMBER |
ClickHouse | ClickHouse | 656,491,881 | 12,342 | filimonov | We need to exclude the probability that it happens due to some data corruption or hw issues. Can you try to create some reproducable example? Do check table return 1? Does downgrade helps with the issue? | 2020-07-10T05:32:35 | 2020-07-10T05:33:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,776,000 | 12,393 | den-crane | https://clickhouse.tech/docs/en/sql-reference/functions/rounding-functions/#rounding_functions-round
round(x[, N])
Rounds a value to a specified number of decimal places.
The function returns the nearest number of the specified order. In case when given number has equal distance to surrounding numbers, the fu... | 2020-07-10T16:49:52 | 2020-07-10T16:54:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,456,582 | 12,376 | alexey-milovidov | This also wrong:
```
SELECT uniqStateForEach([1, NULL]) WITH TOTALS
``` | 2020-07-10T03:13:43 | 2020-07-10T03:13:43 | {} | MEMBER |
ClickHouse | ClickHouse | 656,489,731 | 12,282 | 01egen | The answer turned out to be very simple and was found elsewhere
(https://stackoverflow.com/questions/62814185/clickhouse-matrix-entrywise-addition-how-to-sum-2-dimensional-arrays)
`SELECT sumForEachForEach(m) FROM t1`
| 2020-07-10T05:24:21 | 2020-07-10T05:24:21 | {} | NONE |
ClickHouse | ClickHouse | 656,508,956 | 9,336 | bohutang | Comments:
Add `SELECT DATABASE()` #12314 | 2020-07-10T06:32:23 | 2020-07-10T06:32:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,522,192 | 12,317 | zhbdesign | > Network timeout happens before optimise finishes. Check timeout settings both on server and client
I have set maxwait to 3000 seconds, or will there be a timeout? Which configuration parameter can be modified to avoid this problem? | 2020-07-10T07:10:02 | 2020-07-10T07:10:02 | {} | NONE |
ClickHouse | ClickHouse | 656,549,337 | 12,303 | alesapin | This is expected behavior, but the error message is improved in #12335. | 2020-07-10T08:15:58 | 2020-07-10T08:16:06 | {} | MEMBER |
ClickHouse | ClickHouse | 656,670,067 | 12,387 | filimonov | Yes.
```
CREATE TABLE default.x
(
`a` UInt64
)
ENGINE = MergeTree
PARTITION BY a
ORDER BY tuple()
SETTINGS index_granularity = 8192;
SELECT * FROM x ORDER BY a ASC FORMAT PrettyCompactMonoBlock;
┌─a─┐
│ 0 │
│ 1 │
│ 2 │
│ 3 │
│ 4 │
│ 5 │
│ 6 │
│ 7 │
│ 8 │
│ 9 │
└───┘
ALTER TABLE x
... | 2020-07-10T13:13:16 | 2020-07-10T13:15:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,389,943 | 12,277 | CurtizJ | Yes, I meant that. | 2020-07-09T22:55:53 | 2020-07-09T22:55:53 | {} | MEMBER |
ClickHouse | ClickHouse | 656,442,709 | 7,433 | alexey-milovidov | If there is an issue, it is most likely not in aggregate function but somewhere in `FunctionsBitmap`. | 2020-07-10T02:17:04 | 2020-07-10T02:17:04 | {} | MEMBER |
ClickHouse | ClickHouse | 656,494,619 | 12,280 | alexey-milovidov | Superseded by #12376 | 2020-07-10T05:43:11 | 2020-07-10T05:43:11 | {} | MEMBER |
ClickHouse | ClickHouse | 656,518,278 | 12,264 | alexey-milovidov | `EXPLAIN SYNTAX SELECT a, sum(b) FROM (SELECT 1 AS a, 1 AS b, 0 AS c) GROUP BY a HAVING c` | 2020-07-10T07:00:04 | 2020-07-10T07:00:04 | {} | MEMBER |
ClickHouse | ClickHouse | 656,546,678 | 12,291 | Slach | @CurtizJ thanks a lot | 2020-07-10T08:09:38 | 2020-07-10T08:09:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,560,969 | 12,302 | filimonov | check this https://github.com/filimonov/ClickHouse/commit/9d46775a6f43bb809c2bf7c8cc24531ae7486501 | 2020-07-10T08:42:16 | 2020-07-10T08:42:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,748,771 | 12,408 | den-crane | Please provide test_1 DDL.
It's expected that such insert is slow. Because it need to create 4000 files (bin+mrk).
But 139.411 sec is too slow
Also such insert will use 4GB memory (2MB*2000 (1MB compress buffer + 1MB write buffer)) | 2020-07-10T15:54:05 | 2020-07-10T15:54:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,832,814 | 11,306 | alexey-milovidov | https://clickhouse.tech/docs/en/engines/table-engines/mergetree-family/replication/#recovery-when-metadata-in-the-zookeeper-cluster-is-lost-or-damaged | 2020-07-10T18:57:33 | 2020-07-10T18:57:33 | {} | MEMBER |
ClickHouse | ClickHouse | 656,485,032 | 12,293 | filimonov | Also may be we should have a possibility to push down predicates somehow in other cases which are currently forbidden (final, limit etc).
Either by special setting (push the responsibility to the user) either by implementing smth like prewhere for views, either by distinguishing which column are real and not affecte... | 2020-07-10T05:06:42 | 2020-07-10T05:07:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 656,789,005 | 12,378 | alexey-milovidov | ClickHouse build check:
> Step 11/12 : COPY build.sh /
COPY failed: stat /place/sandbox-data/tasks/0/8/720579080/docker_root/tmp/docker-builder025526973/build.sh: no such file or directory
Traceback (most recent call last):
File "./packager", line 221, in <module>
build_image(image_with_version, dockerfile... | 2020-07-10T17:17:58 | 2020-07-10T17:17:58 | {} | MEMBER |
ClickHouse | ClickHouse | 656,816,597 | 12,409 | alexey-milovidov | CC @amosbird | 2020-07-10T18:20:18 | 2020-07-10T18:20:18 | {} | MEMBER |
ClickHouse | ClickHouse | 640,757,888 | 11,473 | den-crane | Probably it's just impossible to query 2250 columns with joining 1125 of them.
| 2020-06-08T17:09:59 | 2020-06-08T17:09:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,814,027 | 11,496 | alexey-milovidov | Performance test Ok.
There is some slowdown but only in "unstable" queries - it has no statistical significance. | 2020-06-08T19:03:42 | 2020-06-08T19:03:42 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.