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 | 729,546,032 | 235 | develar | No `darwin` artifacts anymore on "ClickHouse build check", and link on page https://clickhouse.tech/docs/en/getting-started/install/#from-binaries-non-linux for macOS is broken. | 2020-11-18T09:17:24 | 2020-11-18T09:17:24 | {} | NONE |
ClickHouse | ClickHouse | 729,723,373 | 16,984 | den-crane | Check with the latest version of 20.3 LTS -- 20.3.21.2
Do you have reproducible example? | 2020-11-18T14:40:59 | 2020-11-18T14:40:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,956,806 | 17,070 | Akazz | Looking good to me. Let's wait for you to make CI checks happy and merge | 2020-11-17T14:16:37 | 2020-11-17T14:16:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,140,484 | 16,680 | Avogar | Failing checks are not related to this PR. | 2020-11-17T19:10:00 | 2020-11-17T19:10:00 | {} | MEMBER |
ClickHouse | ClickHouse | 729,145,726 | 17,055 | alexey-milovidov | @filimonov It's not related to the storage subsystem.
What query does is just a cross-product of 100 000 x 100 000 rows. | 2020-11-17T19:18:40 | 2020-11-17T19:18:40 | {} | MEMBER |
ClickHouse | ClickHouse | 729,141,727 | 15,419 | myrrc | Something strange is happening with the performance tests in https://github.com/ClickHouse/ClickHouse/pull/17100 :
They show about 8x times slowdown on the patched ClickHouse server, which I can't reproduce.
Attaching my local perf.tests results (before Akuzm patches).
[clean_test.txt](https://github.com/ClickHouse/... | 2020-11-17T19:12:24 | 2020-11-17T19:24:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,261,332 | 3,607 | Felixoid | Thank you, but it looks still more or less the same to me.
We use puppet though, and it looks more native. Here's an example https://github.com/innogames/puppet-clickhouse/blob/master/manifests/server/config.pp#L18 | 2020-11-17T22:50:45 | 2020-11-17T22:50:45 | {} | MEMBER |
ClickHouse | ClickHouse | 729,633,559 | 17,179 | sundy-li | cc @alesapin how to reproduce https://gist.github.com/alesapin/251a45e90e8fb87f26ec532b479badbf ? | 2020-11-18T12:01:43 | 2020-11-18T12:01:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,696,554 | 17,159 | shawel | Thanks @den-crane & @amosbird | 2020-11-18T14:01:39 | 2020-11-18T14:01:39 | {} | NONE |
ClickHouse | ClickHouse | 729,721,136 | 17,178 | den-crane | For CSV case you should use a String representation '2019-01-01 01:00:00.000'. | 2020-11-18T14:37:17 | 2020-11-18T14:37:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,825,118 | 12,489 | alexey-milovidov | It's non obvious how to solve it because there is no guarantee that the database exists on the remote server. | 2020-11-18T17:17:15 | 2020-11-18T17:17:15 | {} | MEMBER |
ClickHouse | ClickHouse | 729,057,191 | 16,636 | hombit | Sorry for the delay, finally I could test it locally and I hope it works now | 2020-11-17T16:50:08 | 2020-11-17T16:50:08 | {
"hooray": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,146,868 | 17,055 | alexey-milovidov | Check the number of cores and CPU frequency.
The number of (used) cores can be validated simply by running `top` while query is executed.
Divide the percentage value by 100, e.g. when you see `800` is means that 8 cores are used.
The number of threads can be tuned with the `max_threads` setting. | 2020-11-17T19:20:54 | 2020-11-17T19:21:28 | {} | MEMBER |
ClickHouse | ClickHouse | 729,326,558 | 17,159 | amosbird | > This works. the only way to identify b is if it is included in where clause as a set.
That's because the introduced names are rewritten like this
```
WITH
b AS
(
SELECT [1, 2, 3]
)
SELECT
number,
b AS c
FROM system.numbers
WHERE ([1, 2, 3] IN (( SELECT [1, 2, 3] ) AS b))
... | 2020-11-18T02:00:01 | 2020-11-18T02:00:01 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 729,370,893 | 16,979 | lddlww | thanks for reply very much | 2020-11-18T03:36:42 | 2020-11-18T03:36:42 | {} | NONE |
ClickHouse | ClickHouse | 729,393,938 | 17,161 | vzakaznikov | TestFlows check is green. Ready for merge.
| 2020-11-18T04:15:03 | 2020-11-18T04:15:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,405,080 | 17,055 | alexey-milovidov | Also check the number of actually used CPU cores during query execution.
The real CPU frequency may vary constantly.
You can measure it with the `turbostat` tool.
Desktop/Laptop CPUs usually have higher frequency than server CPUs (they have high number of cores).
For example, there is a server with two AMD EPYC... | 2020-11-18T04:51:22 | 2020-11-18T04:51:22 | {} | MEMBER |
ClickHouse | ClickHouse | 729,712,921 | 17,162 | trollhe | > Yes. You can.
>
> https://clickhouse.tech/docs/en/development/developer-instruction/
that`s great,thank you. | 2020-11-18T14:25:40 | 2020-11-18T14:25:40 | {} | NONE |
ClickHouse | ClickHouse | 729,797,811 | 17,173 | den-crane | Can you reproduce the issue with newer Clickhouse 20.11.3.3? | 2020-11-18T16:33:40 | 2020-11-18T16:33:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,227,223 | 17,150 | alexey-milovidov | Possible solutions:
1. Don't throw before inserted block if it's not the first block in INSERT query. It will allow INSERT SELECT to finish, maybe with creating too many parts. Motivation: small INSERT can be easily retried, but INSERT SELECT is too cumbersome to repeat.
2. When there are too many parts, wait for... | 2020-11-17T21:33:37 | 2020-11-17T21:34:42 | {
"heart": 3
} | MEMBER |
ClickHouse | ClickHouse | 729,259,253 | 3,607 | sergiustheblack | We do such thing with saltstack formula. [check out](https://github.com/sergiustheblack/clickhouse-saltstack-formula).
Definitely not what you are asking for, but... It just generates xml config from yaml pillar. | 2020-11-17T22:45:30 | 2020-11-17T22:45:30 | {} | NONE |
ClickHouse | ClickHouse | 729,481,661 | 17,055 | l1t1 | > > @filimonov It's not related to the storage subsystem.
> > What query does is just a cross-product of 100 000 x 100 000 rows.
>
> In the VMs there is a way to mount a folder from a host, in that case some weird filesystem is used, and it can perform very bad even in trivial `stat` operations.
>
> @l1t1 can yo... | 2020-11-18T07:07:20 | 2020-11-18T07:07:20 | {} | NONE |
ClickHouse | ClickHouse | 729,084,434 | 17,091 | zhang2014 | Integration tests (thread) — Timeout :( - unrelated
Stress test (thread) — Sanitizer assert (in stderr.log) - unrelated
Performance — 1 errors, 1 too long, 4 faster, 86 unstable - unrelated | 2020-11-17T17:32:30 | 2020-11-18T09:46:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,837,851 | 16,977 | vzakaznikov | @vitlibar, Hi Vitaliy, this PR is ready for your review and merge. We have more tests coming and so we would like to have these merged as soon as possible if possible. Thanks! | 2020-11-18T17:38:17 | 2020-11-18T17:38:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,059,232 | 17,122 | amosbird | Heh. It also indicates another issue: `key DateTime column and Date value`. Both are fixed in https://github.com/ClickHouse/ClickHouse/pull/17145 | 2020-11-17T16:53:17 | 2020-11-17T16:53:17 | {
"heart": 1
} | COLLABORATOR |
ClickHouse | ClickHouse | 729,078,226 | 16,532 | atercattus | @den-crane
It doesn't help either.
```
:) show create table TABLENAME
CREATE TABLE default.TABLENAME
(
`date` Date DEFAULT toDate(time),
`time` DateTime,
...
)
ENGINE = ReplicatedAggregatingMergeTree('/clickhouse/tables/{shard}/TABLENAME', '{replica}')
PARTITION BY toYYYYMM(date)
ORDER BY (..... | 2020-11-17T17:22:06 | 2020-11-17T19:27:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,454,796 | 17,055 | filimonov | > @filimonov It's not related to the storage subsystem.
> What query does is just a cross-product of 100 000 x 100 000 rows.
In the VMs there is a way to mount a folder from a host, in that case some weird filesystem is used, and it can perform very bad even in trivial `stat` operations.
@l1t1 can you also sen... | 2020-11-18T06:11:39 | 2020-11-18T06:11:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,566,989 | 16,835 | zhang2014 | It's in my TODO list, and I'm trying to reproduce it in my environment : ) | 2020-11-18T09:55:06 | 2020-11-18T09:55:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,637,528 | 17,149 | nikitamikhaylov | > @nikitamikhaylov the newest version has modified SOURCES in `CMakeList.txt`, we may need to keep the change update in `rocksdb-cmake/CMakeList.txt`
Yes, I forgot about our own CMake files for contrib/. | 2020-11-18T12:10:17 | 2020-11-18T12:10:17 | {} | MEMBER |
ClickHouse | ClickHouse | 729,857,457 | 17,178 | AKarbas | @den-crane Thanks for the rather quick response.
In that case:
Is there a way to insert millisecond-granular unix timestamps from CSVs? And is this requirement present if importing via the HTTP API?
Also, IMO, this requirement about different representations required in different input formats could be added to ... | 2020-11-18T18:04:00 | 2020-11-18T18:04:41 | {} | NONE |
ClickHouse | ClickHouse | 728,904,401 | 10,280 | belveyf | I just get into the same situation. It seems that bitmapToArray didn't function well with Uint64.
And bitmapCardinality seams have the same issue ( sql below returnd 37, which is expected to return 38).
SELECT bitmapCardinality(bitmapBuild([8391058,695788478213916450,693475830172484980,693566989590856286,693439459... | 2020-11-17T12:46:07 | 2020-11-17T12:55:50 | {} | NONE |
ClickHouse | ClickHouse | 729,175,402 | 17,111 | nikitamikhaylov | Accidentally deleted a branch | 2020-11-17T20:14:42 | 2020-11-17T20:14:42 | {} | MEMBER |
ClickHouse | ClickHouse | 729,285,479 | 17,160 | l1t1 | I used to use oracle:)
```
Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> show parameter max
NAME TYPE VALUE
------------------------------------ ----------- ... | 2020-11-17T23:56:07 | 2020-11-17T23:56:07 | {} | NONE |
ClickHouse | ClickHouse | 729,480,926 | 17,115 | god1dog | Tell me how long to wait for a solution to this problem? | 2020-11-18T07:05:59 | 2020-11-18T07:05:59 | {} | NONE |
ClickHouse | ClickHouse | 729,488,817 | 17,089 | alesapin | StorageRocksDB Failed one more time https://gist.github.com/alesapin/251a45e90e8fb87f26ec532b479badbf | 2020-11-18T07:21:32 | 2020-11-18T07:21:32 | {} | MEMBER |
ClickHouse | ClickHouse | 729,546,703 | 16,835 | rjfh1980s | > Can you try to execute `SHOW MASTER LOGS`? And paste the result set structure into this issue?
Hello, is there any solution to the above question?
Do I need to upgrade ClickHouse or change any configuration? | 2020-11-18T09:18:35 | 2020-11-18T09:18:35 | {} | NONE |
ClickHouse | ClickHouse | 729,573,790 | 16,532 | atercattus | @den-crane I can write to him on Telegram.
But I definitely can't share this partition as is (it contains 106G of our sensitive NDA data :) ). Also, we have a problem with this whole table, not only with one partition.
But I'm ready to do something for testing purposes. | 2020-11-18T10:06:41 | 2020-11-18T10:06:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,636,543 | 17,179 | nikitamikhaylov | @sundy-li Btw, why you have no diff in submodule itself? I thought that changing only `.gitmodules` file will make no sense. Correct me if I'm wrong. | 2020-11-18T12:08:10 | 2020-11-18T12:08:10 | {} | MEMBER |
ClickHouse | ClickHouse | 729,716,713 | 17,178 | den-crane | You should use
```
SELECT fromUnixTimestamp64Milli(toInt64(1546300800000))
┌─fromUnixTimestamp64Milli(toInt64(1546300800000))─┐
│ 2019-01-01 01:00:00.000 │
└──────────────────────────────────────────────────┘
```
DateTime64 is based on Decimal64 type and behaves as Decimal64. It do... | 2020-11-18T14:30:54 | 2020-11-18T14:30:54 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,956,584 | 17,088 | amosbird | This should fix it https://github.com/ClickHouse/ClickHouse/pull/17135 | 2020-11-17T14:16:18 | 2020-11-17T14:16:18 | {
"heart": 1
} | COLLABORATOR |
ClickHouse | ClickHouse | 728,957,602 | 16,922 | yard | Hey, sorry for the late reply!
The above works properly (empty set) whether I change suggested settings or not. I also tried changing the column to `String` (since I noticed the aggregation is using `key64` method with your sample), but the issue does not reproduce.
I have a suspicion the bug might be related to... | 2020-11-17T14:17:52 | 2020-11-17T14:17:52 | {} | NONE |
ClickHouse | ClickHouse | 728,973,853 | 6,093 | alexey-milovidov | Still relevant. The current behaviour is below the desired usability standard. | 2020-11-17T14:43:28 | 2020-11-17T14:43:28 | {
"+1": 3
} | MEMBER |
ClickHouse | ClickHouse | 729,094,403 | 12,615 | filimonov | No info, no reproducable test case. | 2020-11-17T17:49:23 | 2020-11-17T17:49:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,290,886 | 17,055 | l1t1 | AMD VM
```
│ max_threads │ 'auto(8)' │
[root@VM_0_52_centos tmp]# cat /proc/cpuinfo |grep "process"
processor : 0
processor : 1
processor : 2
processor : 3
processor : 4
processor : 5
processor : 6
processor : 7
processo... | 2020-11-18T00:11:47 | 2020-11-18T02:34:47 | {} | NONE |
ClickHouse | ClickHouse | 729,555,369 | 14,162 | ghost | @alexey-milovidov
Of course, Why not?
Thank you! | 2020-11-18T09:34:46 | 2020-11-18T09:34:46 | {} | NONE |
ClickHouse | ClickHouse | 729,835,201 | 17,180 | vklimenkov | Благодарю за ответ! Да, дедупликацию мы не используем. Так что будем пробовать Buffer. | 2020-11-18T17:33:59 | 2020-11-18T17:33:59 | {} | NONE |
ClickHouse | ClickHouse | 728,971,909 | 16,769 | nikitamikhaylov | Will wait only for `Yandex Synchonization Check`. | 2020-11-17T14:40:26 | 2020-11-17T14:40:26 | {} | MEMBER |
ClickHouse | ClickHouse | 729,144,405 | 17,112 | alexey-milovidov | Yes, looks like we only have to put every name into IHints and then ask for possible matches. | 2020-11-17T19:16:16 | 2020-11-17T19:16:16 | {} | MEMBER |
ClickHouse | ClickHouse | 729,172,556 | 17,111 | alexey-milovidov | Yes, glibc's memcpy is slightly better.
We cannot merge this PR as is due to `Compatibility check — glibc check failed`. | 2020-11-17T20:08:48 | 2020-11-17T20:08:48 | {} | MEMBER |
ClickHouse | ClickHouse | 729,288,591 | 17,136 | l1t1 | but this syntax works in oracle,
```
SQL> WITH x(n) AS (SELECT 1 from dual) SELECT n from x;
N
----------
1
```
doesn't work in ch
```
VM_0_52_centos :) WITH x(n) AS (SELECT 1 ) SELECT n from x;
Syntax error: failed at position 14 ('('):
WITH x(n) AS (SELECT 1 ) SELECT n from x;
E... | 2020-11-18T00:04:50 | 2020-11-18T00:04:50 | {} | NONE |
ClickHouse | ClickHouse | 729,345,042 | 17,070 | fastio | > @fastio FYI, failed "Stress test (memory)" is irrelevant to this PR and is a known issue with RocksDB lib in our contrib.
>
> Received signal Aborted (6)
Thanks for the reminder. | 2020-11-18T03:00:08 | 2020-11-18T03:00:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,433,386 | 17,055 | l1t1 | AMD VM
```
[root@VM_0_52_centos ~]# turbostat
CPU TSC_MHz
- 2602
0 2606
1 2606
2 2606
3 2606
4 2607
5 2606
6 2604
7 2604
8 2603
9 2602
10 2601
11 2599
12 2599
... | 2020-11-18T05:15:37 | 2020-11-18T05:15:37 | {} | NONE |
ClickHouse | ClickHouse | 729,557,282 | 17,152 | akuzm | ```
src/AggregateFunctions/CMakeFiles/clickhouse_aggregate_functions.dir/AggregateFunctionMinMaxAny.cpp.o -c ../src/AggregateFunctions/AggregateFunctionMinMaxAny.cpp
g++-10: fatal error: Killed signal terminated program cc1plus
compilation terminated.
```
The build failure is not related to the changes, it is a sp... | 2020-11-18T09:38:05 | 2020-11-18T09:38:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,638,434 | 17,179 | sundy-li | > @sundy-li Btw, why you have no diff in submodule itself? I thought that changing only `.gitmodules` file will make no sense. Correct me if I'm wrong.
Aha, you are right :) | 2020-11-18T12:12:08 | 2020-11-18T12:12:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,704,799 | 17,168 | den-crane | Worked before:
```
19.13.7.57, 20.3.21.2, 20.4.9.110
SELECT max(c1.value),c0.value FROM (select 1 AS insid,2 AS ts,3 AS value) c1
LEFT JOIN (SELECT 1 AS insid,2 AS ts,3 AS value) c0 ON c1.insid=c0.insid AND c1.ts=c0.ts
GROUP BY c0.value
┌─max(value)─┬─c0.value─┐
│ 3 │ 3 │
└────────────┴──... | 2020-11-18T14:13:59 | 2020-11-18T16:37:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,993,326 | 17,136 | den-crane | PG
```
WITH x AS (SELECT 1) SELECT x;
ERROR: column "x" does not exist
LINE 1: WITH x AS (SELECT 1) SELECT x;
^
WITH x AS (SELECT 1) SELECT * from x;
?column?
----------
1
(1 row)
``` | 2020-11-17T15:12:35 | 2020-11-17T15:13:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,073,465 | 17,115 | filimonov | > @filimonov It's correct, as it is designed, as in docs.
> It means: break after the specified number of rows is received in the result.
> Result is formed in blocks. We don't cut blocks.
That's OK. | 2020-11-17T17:14:27 | 2020-11-17T17:14:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,287,476 | 17,136 | l1t1 | you should add a col name for const value
```
WITH x AS
(
SELECT 1 AS n
)
SELECT n
FROM x
Query id: 8c1a7afc-6f91-4715-aabf-ca1057025813
┌─n─┐
│ 1 │
└───┘
1 rows in set. Elapsed: 0.036 sec.
``` | 2020-11-18T00:01:54 | 2020-11-18T00:01:54 | {} | NONE |
ClickHouse | ClickHouse | 727,270,243 | 16,384 | den-crane | I able to reproduce it with official.
```
:) insert into array values([1:+1, 2], 'aaa');
INSERT INTO array VALUES
Query id: fa09aee8-3f4b-41f8-aae6-0272095bbb12
Error on processing query: insert into array values([1:+1, 2], 'aaa');
Code: 62, e.displayText() = DB::Exception: Cannot parse expression of type... | 2020-11-14T21:52:11 | 2020-11-18T18:27:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,933,681 | 17,112 | fastio | It looks that the column name should be registered in the IHints.
Please assign it to me.
| 2020-11-17T13:40:32 | 2020-11-17T13:40:32 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,971,461 | 16,932 | egor-moysevich | Attaching subset of logs, if it helps.
```
Partition ('2015-06-01','<partition_name>') copied
Partition ('2015-06-01','<partition_name>') copied
Partition ('2015-06-01','<partition_name>') is copied successfully
It took 8.10 seconds to copy partition ('2015-06-01','<partition_name>'): 192.04 KB uncompressed byte... | 2020-11-17T14:39:42 | 2020-11-17T14:39:42 | {} | NONE |
ClickHouse | ClickHouse | 729,143,818 | 17,113 | alexey-milovidov | Another possibility is to allow to create VIEWs with parametrized queries. SELECT from VIEW will work only if parameters are defined. | 2020-11-17T19:15:05 | 2020-11-17T19:15:05 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 729,148,334 | 17,050 | alexey-milovidov | It is expected that when you run `clickhouse-test` with the most fresh test suite over the old version of ClickHouse server, the server will crash. Because we have fixed the old crashes and there are test cases to reproduce them.
Basically you check that tests are doing the right thing. | 2020-11-17T19:23:34 | 2020-11-17T19:24:06 | {} | MEMBER |
ClickHouse | ClickHouse | 729,155,983 | 17,054 | den-crane | ```sql
SELECT
name,
value
FROM system.settings
WHERE name IN ('max_memory_usage', 'max_bytes_before_external_group_by')
┌─name───────────────────────────────┬─value────────┐
│ max_bytes_before_external_group_by │ 76027960320 │
│ max_memory_usage │ 152055920640 │
└──────────────────... | 2020-11-17T19:37:55 | 2020-11-17T19:37:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,710,462 | 17,163 | trollhe | > 2020 11 71 15 37 00 .000
>
> What is 71 ?
>
> ```
> SELECT parseDateTimeBestEffort('20201117153700.000')
>
> ┌─parseDateTimeBestEffort('20201117153700.000')─┐
> │ 2020-11-17 15:37:00 │
> └───────────────────────────────────────────────┘
>
>
>
> SELECT parseDateTime64BestEf... | 2020-11-18T14:21:53 | 2020-11-18T14:21:53 | {} | NONE |
ClickHouse | ClickHouse | 729,796,373 | 17,180 | den-crane | Дедупликация инсертов перестанет работать. Вы повторяете идентичный инсерт, а engine Buffer нарушает идентичность потому порежет его на куски и склеит с другими кусками других инсертов.
Возможно вы не используете дедупликацию инсертов, тогда вам это нерелевантно.
https://clickhouse.tech/docs/ru/engines/table-engi... | 2020-11-18T16:31:28 | 2020-11-18T16:31:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,926,954 | 17,115 | alexey-milovidov | @filimonov It's correct, as it is designed, as in docs.
It means: break after the specified number of rows is received in the result.
Result is formed in blocks. We don't cut blocks. | 2020-11-17T13:28:56 | 2020-11-17T13:28:56 | {} | MEMBER |
ClickHouse | ClickHouse | 729,220,025 | 17,111 | alexey-milovidov | Ok, I don't need it :) | 2020-11-17T21:18:48 | 2020-11-17T21:18:48 | {} | MEMBER |
ClickHouse | ClickHouse | 729,409,444 | 14,162 | alexey-milovidov | @achimbab Thank you! I can read almost 100% info from the slides :)
Maybe we can add Kakao to the page: https://clickhouse.tech/docs/en/introduction/adopters/ ? | 2020-11-18T04:57:52 | 2020-11-18T04:58:09 | {} | MEMBER |
ClickHouse | ClickHouse | 729,529,767 | 17,149 | sundy-li | @nikitamikhaylov the newest version has modified SOURCES in `CMakeList.txt`, we may need to keep the change update in `rocksdb-cmake/CMakeList.txt` | 2020-11-18T08:47:14 | 2020-11-18T08:47:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,826,782 | 17,175 | akuzm | The `00377_shard_group_uniq_array_of_string_array` test failure looks spurious and has no diagnostics, I'll restart just in case. | 2020-11-18T17:19:58 | 2020-11-18T17:19:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,853,904 | 17,168 | CurtizJ | Will be fixed by #16951. | 2020-11-18T17:57:45 | 2020-11-18T17:58:29 | {} | MEMBER |
ClickHouse | ClickHouse | 729,080,154 | 16,751 | amosbird | > Integration tests - test_storage_kafka - failure is related: https://gist.github.com/tavplubix/26022e8f3fbdb9451068ed5c98aeeb75
Yeah, it's related. | 2020-11-17T17:25:26 | 2020-11-17T17:25:26 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 729,345,855 | 14,162 | ghost | @alexey-milovidov
Hello.
From my company, developer's online conference has just begun. It was supposed to take place offline but because of pandemic, it is going on online. You can access https://if.kakao.com/session/117
I decided not to use `performance test result`, but introduced features of ClickHouse t... | 2020-11-18T03:02:36 | 2020-11-18T03:02:36 | {} | NONE |
ClickHouse | ClickHouse | 729,635,114 | 17,120 | alesapin | Quite big changes for a very rare problem. So don't backport. | 2020-11-18T12:05:07 | 2020-11-18T12:05:07 | {} | MEMBER |
ClickHouse | ClickHouse | 729,698,421 | 16,984 | K4liber | I moved my conditions from PREWHERE to WHERE and use SAMPLE/OFFSET, and aggregate the results after all, but this is only a workaround not a solution. | 2020-11-18T14:04:20 | 2020-11-18T14:04:20 | {} | NONE |
ClickHouse | ClickHouse | 729,780,961 | 17,120 | alesapin | ```
2020.11.18 17:02:07.429869 [ 1073 ] {78c0f24a-f484-4082-83fb-8980891033eb} <Error> executeQuery: Code: 57, e.displayText() = DB::Exception: Cannot attach table with UUID 061e98ed-d127-4ae4-a606-d5727ca4d5c5, because it was detached but still used by some query. Retry later. (version 20.12.1.5209) (from [::1]:36242... | 2020-11-18T16:06:47 | 2020-11-18T16:06:47 | {} | MEMBER |
ClickHouse | ClickHouse | 729,826,760 | 12,489 | alexey-milovidov | Maybe we can always substitute the current database when sending a query to remote server?
It will break some queries but will look more logical. | 2020-11-18T17:19:56 | 2020-11-18T17:19:56 | {} | MEMBER |
ClickHouse | ClickHouse | 729,296,595 | 17,159 | den-crane | 2 different types of with. CTE and scalars. `WITH xxx as select` VS `WITH select as yyy`
```
WITH
a AS
(
SELECT arrayJoin([1, 2, 3])
),
(SELECT [1, 2, 3] y ) as b
SELECT
number,
b AS c
FROM system.numbers
WHERE toUInt8(number) IN (a)
LIMIT 10
┌─number─┬─c───────┐
│ ... | 2020-11-18T00:26:49 | 2020-11-18T00:26:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,642,597 | 6,359 | haiertashu | hi @alexey-milovidov
`if some data is used by queries, it will be automatically cached (in page cache) by OS.`
Does this mean ch use BIO to read and write data? and why not DIO? | 2020-11-18T12:20:47 | 2020-11-18T12:29:02 | {} | NONE |
ClickHouse | ClickHouse | 729,774,480 | 17,052 | den-crane | Does it reproduce with non-replicated tables?
| 2020-11-18T15:57:36 | 2020-11-18T15:57:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,981,019 | 17,136 | amosbird | It's by design... Those identifiers can appear in places where table objects are allowed. It's like we do `select table_1 from system.numbers` | 2020-11-17T14:54:07 | 2020-11-17T14:54:07 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 729,012,805 | 17,070 | Akazz | @fastio FYI, failed "Stress test (memory)" is irrelevant to this PR and is a known issue with RocksDB lib in our contrib.
<details><summary>Received signal Aborted (6)</summary>
```
2020.11.17 17:34:55.551210 [ 255 ] {} <Fatal> BaseDaemon: (version 20.12.1.5203, build id: 557EDE393D39EB80) (from thread 27547) Te... | 2020-11-17T15:40:13 | 2020-11-17T15:40:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,141,079 | 17,136 | alexey-milovidov | Ok. | 2020-11-17T19:11:09 | 2020-11-17T19:11:09 | {} | MEMBER |
ClickHouse | ClickHouse | 729,142,571 | 17,113 | alexey-milovidov | Also we can make them available as table functions to allow SELECT from these handlers or INSERT INTO TABLE FUNCTION.
```
SELECT * FROM handler(param1 = 'hello', param2 = 'world')
``` | 2020-11-17T19:13:47 | 2020-11-17T19:13:47 | {} | MEMBER |
ClickHouse | ClickHouse | 729,154,711 | 17,151 | filimonov | DROP PARTITION - the same. | 2020-11-17T19:35:22 | 2020-11-17T19:35:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,255,814 | 17,160 | alexey-milovidov | `SHOW SETTINGS LIKE '...'` would sound more obvious. | 2020-11-17T22:37:08 | 2020-11-17T22:37:08 | {} | MEMBER |
ClickHouse | ClickHouse | 729,486,939 | 17,089 | alesapin | No relevant errors. | 2020-11-18T07:17:57 | 2020-11-18T07:17:57 | {} | MEMBER |
ClickHouse | ClickHouse | 728,997,007 | 16,787 | lemire | Version 0.2.0 of fast_float is out:
https://github.com/lemire/fast_float/releases/tag/v0.2.0
Release notes:
- Support for Big Endian systems
- Build with pedantic flags under GNU GCC without warnings
- More standard compliant handling of problematic strings (issue19) | 2020-11-17T15:18:05 | 2020-11-17T15:18:05 | {} | NONE |
ClickHouse | ClickHouse | 729,097,259 | 16,979 | filimonov | After few adjustements I was able to make it work, and indeed I see that length of
`metrics.timeseries.labels.name` and `metrics.timeseries.labels.value` differ, while according to the schema they should have the same length.
[protobuf_issue.tar.gz](https://github.com/ClickHouse/ClickHouse/files/5555235/protobuf_... | 2020-11-17T17:52:26 | 2020-11-17T17:52:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,484,807 | 17,082 | alesapin | > @alesapin I think I can take a stab at this - can I have this assigned to me?
Of course! | 2020-11-18T07:13:27 | 2020-11-18T07:13:27 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 729,639,632 | 17,091 | tavplubix | Everything else LGTM | 2020-11-18T12:14:41 | 2020-11-18T12:14:41 | {} | MEMBER |
ClickHouse | ClickHouse | 729,707,677 | 17,163 | den-crane | 2020 11 71 15 37 00 .000
What is 71 ?
```
SELECT parseDateTimeBestEffort('20201117153700.000')
┌─parseDateTimeBestEffort('20201117153700.000')─┐
│ 2020-11-17 15:37:00 │
└───────────────────────────────────────────────┘
SELECT parseDateTime64BestEffortOrNull('2020111715... | 2020-11-18T14:17:58 | 2020-11-18T14:20:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,711,151 | 17,162 | den-crane | Yes. You can.
https://clickhouse.tech/docs/en/development/developer-instruction/ | 2020-11-18T14:22:55 | 2020-11-18T14:22:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,782,073 | 17,120 | alesapin | Test failures seem unrelated but suspicious. | 2020-11-18T16:08:31 | 2020-11-18T16:08:31 | {} | MEMBER |
ClickHouse | ClickHouse | 728,977,521 | 16,532 | den-crane | @atercattus probably you can simply re-merge that partition. Optimize table .. partition id ... final | 2020-11-17T14:49:10 | 2020-11-17T14:49:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,402,831 | 17,136 | alexey-milovidov | Yes, it is the syntax for column names for CTE, it's not supported. | 2020-11-18T04:44:25 | 2020-11-18T04:44:25 | {} | MEMBER |
ClickHouse | ClickHouse | 726,813,367 | 16,384 | hexiaoting | @alexey-milovidov I'm sorry for I forget to remove my own debug info "parse failed 0" before paste the error stack to this issue.
But I reproduce this exception on the latest master branch:
```
show tables;
SHOW TABLES
Query id: 33ef7187-e01c-46c4-a968-53d9fbc1e691
┌─name─────────────────┐
│ columns_tra... | 2020-11-13T15:02:10 | 2020-11-13T15:02:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 726,818,418 | 16,925 | den-crane | Is it reproducible if you query shard table directly ?
```
SELECT max(timestamp)
FROM shard.requests
WHERE (timestamp >= '2020-11-09 23:00:00') AND (timestamp <= '2020-11-09 23:59:59')
``` | 2020-11-13T15:11:11 | 2020-11-13T15:11:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 726,825,546 | 16,937 | satanson | @amosbird fixed | 2020-11-13T15:24:00 | 2020-11-13T15:24:00 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.