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 | 733,251,804 | 17,376 | alexey-milovidov | Ok.
It's better to make a function
toUnixTimestamp(date, [timezone])
that will first convert to datetime according to the timezone and then return a unix timestamp.
Translating date to unix timestamp is a legitimate operation (if timezone is given). | 2020-11-24T21:46:53 | 2020-11-24T21:47:20 | {} | MEMBER |
ClickHouse | ClickHouse | 733,485,742 | 15,111 | vitlibar | `test_storage_rabbitmq/test.py::test_rabbitmq_csv_with_delimiter` will be fixed in https://github.com/ClickHouse/ClickHouse/pull/17311 | 2020-11-25T06:00:21 | 2020-11-25T06:00:21 | {} | MEMBER |
ClickHouse | ClickHouse | 733,570,725 | 17,378 | filimonov | Minimal example
```
WITH it AS ( SELECT * FROM numbers(1) )
SELECT i.number
FROM it AS i
``` | 2020-11-25T09:08:40 | 2020-11-25T09:08:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,592,952 | 17,300 | filimonov | > It looks like buffer size is not enough.
Buffer tries to grow to fit your fat column automatically. But during that attempt it reach the memory limitation. | 2020-11-25T09:43:50 | 2020-11-25T09:44:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,610,302 | 16,421 | filimonov | > same error if i run clickhouse-server --config-file=/etc/clickhouse-server/config.xml
It's completely different...
```
2020.10.28 07:44:48.085210 [ 25804 ] {} <Error> Application: DB::Exception: Effective user of the process (root) does not match the owner of the data (clickhouse). Run under 'sudo -u clickhouse... | 2020-11-25T10:14:14 | 2020-11-25T10:14:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,640,398 | 16,535 | akuzm | Test failures are not related. | 2020-11-25T11:09:57 | 2020-11-25T11:09:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,529,106 | 17,199 | taylor131415 | > 检查/var/log/clickhouse-server1/stderr.log中的错误
There's no error log produced in this path. Seems like the service exited as soon as it starts. | 2020-11-24T01:32:57 | 2020-11-24T01:32:57 | {} | NONE |
ClickHouse | ClickHouse | 732,541,954 | 17,282 | vladimirteddy | Thank you @filimonov . | 2020-11-24T02:15:17 | 2020-11-24T02:15:17 | {} | NONE |
ClickHouse | ClickHouse | 732,678,210 | 17,317 | filimonov | Maybe different block sizes (or where vs prewhere) + https://github.com/ClickHouse/ClickHouse/issues/5105#issuecomment-486454768 ? | 2020-11-24T06:16:02 | 2020-11-24T06:17:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,745,582 | 17,346 | ralphM78 | Thanks for your answer. ATTACH on the wrong node could be possible. Would it work to delete the duplicate parts in zookeeper and on the nodes disk ? I don't really have a test environment to check.
Ralph | 2020-11-24T08:41:55 | 2020-11-24T08:41:55 | {} | NONE |
ClickHouse | ClickHouse | 732,962,710 | 17,255 | akuzm | By the way, both materialized_view_parallel_insert and parallel_inserts tests are now broken (take too long).
Those red cells and error counts are not just for show... | 2020-11-24T13:08:25 | 2020-11-24T13:08:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,106,747 | 17,309 | nikitamikhaylov | Yandex third-party checks failure is related. Maybe need to add overloaded totalBytes() without Context reference as a parameter. | 2020-11-24T16:54:59 | 2020-11-24T16:54:59 | {} | MEMBER |
ClickHouse | ClickHouse | 733,125,155 | 17,354 | den-crane | > What is the reasoning behind this?
performance | 2020-11-24T17:26:48 | 2020-11-24T17:26:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,560,786 | 17,379 | filimonov | Try clang.
https://clickhouse.tech/docs/en/development/build-osx/ | 2020-11-25T08:51:13 | 2020-11-25T08:51:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,581,192 | 17,349 | filimonov | `git submodule update --init --recursive`? | 2020-11-25T09:26:27 | 2020-11-25T09:26:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,585,364 | 16,621 | ilejn | @filimonov could you give your opinion, please? | 2020-11-25T09:32:46 | 2020-11-25T09:32:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,748,769 | 16,804 | vdimir | On @alexey-milovidov advice I've tried to use Eytzinger Binary Search https://algorithmica.org/en/eytzinger
It isn't merged to current PR, and I put changes to PR in my fork repository:
https://github.com/vdimir/ClickHouse/pull/1
It performs better on some cases, but case with random generated IPv6 suffers.
A... | 2020-11-24T08:48:03 | 2020-11-24T08:51:07 | {} | MEMBER |
ClickHouse | ClickHouse | 733,018,585 | 17,363 | nikitamikhaylov | Will look through the whole code of Copier to find more uninitialized variables. | 2020-11-24T14:43:51 | 2020-11-24T14:43:51 | {} | MEMBER |
ClickHouse | ClickHouse | 733,589,208 | 17,055 | filimonov | > `select * from n1 limit 3;`
...Is it your problematic query? We were talking about something completely different...
Are you sure you are doing fair A/B testing? Do you run exactly the same query on both nodes? Can you repeat the result then you repeat your query many times?
| 2020-11-25T09:37:26 | 2020-11-25T09:37:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,673,171 | 16,664 | nikitamikhaylov | Intergration tests failed for unknown reason. Try to rebase on master to rerun them. | 2020-11-25T12:18:53 | 2020-11-25T12:18:53 | {} | MEMBER |
ClickHouse | ClickHouse | 732,353,886 | 17,327 | alexey-milovidov | Experimental feature, no need to backport. | 2020-11-23T18:46:09 | 2020-11-23T18:46:09 | {} | MEMBER |
ClickHouse | ClickHouse | 732,773,698 | 17,317 | UnamedRus | Query duration is almost the same.
And if we set both settings:
```
set max_threads=1;
SELECT DISTINCT toUInt16OrNull(arrayFirst((v, k) -> (k = '15Id'),arr[2],arr[1])) FROM test_extract;
2020.11.24 12:20:35.999961 [ 13607 ] {df5a2bad-cc9a-495d-ac24-78dfaa461096} <Information> executeQuery: Read 2000000 rows,... | 2020-11-24T09:32:29 | 2020-11-24T09:32:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,848,083 | 14,896 | UnamedRus | Can you check clickhouse version 20.10+?
I have similar issue and it was fixed in that version.
https://github.com/ClickHouse/ClickHouse/issues/13862 | 2020-11-24T11:17:17 | 2020-11-24T11:17:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,053,927 | 7,902 | vamirio-chan | Short: incorrect users for one of my databases (i don't have a clue how it could happen because -ls lh displayed 'clickhouse' but I'm using WSL so it might be the reason).
I haven't checked what config clickhouse was using, but apparently not /etc/clickhouse-server/config.xml. So I used this command:
`$ sudo -u cl... | 2020-11-24T15:36:50 | 2020-11-24T15:41:19 | {
"+1": 6
} | NONE |
ClickHouse | ClickHouse | 733,111,549 | 17,354 | den-crane | https://clickhouse.tech/docs/en/interfaces/formats/#data-formatting
During a read operation, incorrect dates and dates with times can be parsed with natural overflow or as null dates and times, without an error message. | 2020-11-24T17:03:12 | 2020-11-24T17:03:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,154,541 | 17,206 | akuzm | Test failures are not related. | 2020-11-24T18:23:13 | 2020-11-24T18:23:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,474,508 | 17,366 | zhang2014 | CC: @tavplubix | 2020-11-25T05:27:31 | 2020-11-25T05:27:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,531,908 | 17,300 | liujiawinds | @den-crane It seems doesn't work, the same error message occured after I inserted the configuration item into `users.xml` and restart clickhouse-server process.
| 2020-11-24T01:41:37 | 2020-11-24T01:41:37 | {} | NONE |
ClickHouse | ClickHouse | 733,130,235 | 17,344 | nikitamikhaylov | Noticed, that there is performance degradation. | 2020-11-24T17:36:14 | 2020-11-24T17:36:14 | {} | MEMBER |
ClickHouse | ClickHouse | 733,565,546 | 17,230 | wutchzone | @den-crane @alexey-milovidov
Thank you both for help. Firstly I was very confused when I used the MergeeTree engine, because I tried to dump the data too early. Propably many parts were not merged and it consumed a lot of RAM when used with ORDER BY. When I try to dump after hour when they were inserted, it does not ... | 2020-11-25T08:59:52 | 2020-11-25T08:59:52 | {} | NONE |
ClickHouse | ClickHouse | 733,691,738 | 17,387 | filimonov | 1) select timezone();
2) http client vs tcp client? | 2020-11-25T12:59:27 | 2020-11-25T12:59:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,389,687 | 17,312 | god1dog | Thank you! | 2020-11-23T19:52:47 | 2020-11-23T19:52:47 | {} | NONE |
ClickHouse | ClickHouse | 732,534,055 | 17,199 | taylor131415 | > show the output of
>
> sudo -u clickhouse clickhouse-server --pid-file=/var/run/clickhouse-server1/clickhouse-server.pid --config-file=/etc/clickhouse-server1/config.xml
Required info as below:
[root@localhost ~]# sudo -u clickhouse clickhouse-server --pid-file=/var/run/clickhouse-server/clickhouse-server-1.pi... | 2020-11-24T01:49:38 | 2020-11-24T01:49:38 | {} | NONE |
ClickHouse | ClickHouse | 733,150,129 | 17,344 | alexey-milovidov | If the error is bogus, you can leave the code if it's compiled with MSan. | 2020-11-24T18:14:02 | 2020-11-24T18:14:02 | {} | MEMBER |
ClickHouse | ClickHouse | 733,578,029 | 7,659 | filimonov | May be we can just list them inside columns? With a special column type? In that case we will also hale parts_columns 'for free'. | 2020-11-25T09:21:07 | 2020-11-25T09:21:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,659,635 | 16,976 | PhantomPhreak | Updated to v20.11.3.3, can't reproduce anymore.
Thanks for suggestions, i'll close this issue | 2020-11-25T11:49:39 | 2020-11-25T11:49:39 | {} | NONE |
ClickHouse | ClickHouse | 733,682,974 | 5,182 | akuzm | We added some experimental support for OpenTelemetry: we accept trace context headers over HTTP and Native protocols, add a few spans for some stages of query execution, store them in the `system.opentelemetry_span_log` table, and propagate the trace context to remote ClickHouse servers and to URL engine.
https://cl... | 2020-11-25T12:40:53 | 2020-11-25T12:40:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,388,459 | 17,109 | azat | >Also as a long-term solution it worth looking at implementing them via
>RCU (to move the extra overhead out from the write code path into read
>side).
Does not worth it (and why would, since the write-path is hot, not the read):
```
-----------------------------------------------------------------------------... | 2020-11-23T19:50:21 | 2020-11-23T19:52:41 | {} | MEMBER |
ClickHouse | ClickHouse | 732,522,938 | 17,199 | den-crane | check errors in /var/log/clickhouse-server1/stderr.log | 2020-11-24T01:27:28 | 2020-11-24T01:27:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,109,099 | 17,142 | nikitamikhaylov | Original bugfix seems to be based on some changes in `wide_integer_impl.h` which are not present in current version. | 2020-11-24T16:59:08 | 2020-11-24T16:59:08 | {} | MEMBER |
ClickHouse | ClickHouse | 733,123,430 | 17,354 | veqryn | > https://clickhouse.tech/docs/en/interfaces/formats/#data-formatting
>
> During a read operation, incorrect dates and dates with times can be parsed with natural overflow or as null dates and times, without an error message.
This does not explain why everything is overflowing at different values.
Or why overflo... | 2020-11-24T17:23:31 | 2020-11-24T17:23:31 | {} | NONE |
ClickHouse | ClickHouse | 733,564,455 | 17,377 | filimonov | I guess the question is about `WITH FILL` feature.
No way for that in common case (how should i 'fill' the 'gap' between 'Alaska' and '襄樊市' ?)
In your case you can split your 'x1' into string part and numeric part, and do `WITH FILL` against numeric part.
```
SELECT concat('x', toString(x))
FROM
(
S... | 2020-11-25T08:57:45 | 2020-11-25T09:00:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,570,127 | 17,271 | azat | This looks pretty trivial and the CI failures should not be related, can someone take a look please? (otherwise you can't use cmake 3.19 right now) | 2020-11-25T09:07:43 | 2020-11-25T09:07:43 | {} | MEMBER |
ClickHouse | ClickHouse | 733,578,582 | 17,357 | filimonov | Duplicate of #7659
Also #8158 | 2020-11-25T09:22:05 | 2020-11-25T09:22:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,598,694 | 17,300 | liujiawinds | @den-crane That not work too.
I found a work around that use binary stream api solve this problem.
```
long[] longs = termList.stream().flatMap(Collection::stream).mapToLong(v -> Long.valueOf(v.toString())).toArray();
connection.createStatement().sendRowBinaryStream(
"insert into bloom_... | 2020-11-25T09:53:31 | 2020-11-25T09:53:31 | {} | NONE |
ClickHouse | ClickHouse | 733,614,861 | 5,590 | YoannBuch | Hello @hczhcz, very nice addition :)
How would one use `-Resample` in combination of `quantileTimingWeighted`?
This works (no defined `level`):
```
quantileTimingWeightedResample(1, 10, 1)(duration, count, t)
```
But how do we specify the `level`, e.g. 0.5 or 0.9?
Edit: this way, e.g. p90:
```
quantil... | 2020-11-25T10:22:25 | 2020-11-25T11:12:27 | {} | NONE |
ClickHouse | ClickHouse | 732,506,836 | 14,885 | xjewer | closing in favor of https://github.com/ClickHouse/ClickHouse/pull/17348 | 2020-11-24T00:34:30 | 2020-11-25T12:22:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,540,139 | 17,199 | den-crane | ```
chown -R clickhouse.clickhouse /etc/clickhouse-server1/
```
and again
sudo -u clickhouse clickhouse-server --pid-file=/var/run/clickhouse-server/clickhouse-server-1.pid --config-file=/etc/clickhouse-server1/config1.xml | 2020-11-24T02:09:21 | 2020-11-24T02:09:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,932,097 | 10,333 | shacuros | newest version still is affected by this, as i have had to notice in some usecases lately.
There are not only missing results like in the Example by denny, but also totally different ones as well:
CREATE TABLE neighbor_test (
rowNr UInt8,
val_string String,
val_low LowCardinality(String)
) ENGINE=Merge... | 2020-11-24T12:10:51 | 2020-11-24T12:13:27 | {} | NONE |
ClickHouse | ClickHouse | 733,107,444 | 17,362 | den-crane | just mention column names ((A,B,C,D)) in insert `insert into table(A,B,C,D) format CSV`
or use CSV with columns names.
| 2020-11-24T16:56:11 | 2020-11-24T16:56:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,603,423 | 17,278 | filimonov | BTW: in sentry / collecting anonymous stats context: it can be problematic to use even such anonymous ids without clear user opt-in in some countries (GDPR and similar laws) - using any kind of uniques identifiers / device ids etc. in 'anonimous' statistics is regulated, and should not happen without clear user opt-it.... | 2020-11-25T10:01:44 | 2020-11-25T10:03:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,353,910 | 17,325 | alexey-milovidov | Experimental feature, no need to backport. | 2020-11-23T18:46:12 | 2020-11-23T18:46:12 | {} | MEMBER |
ClickHouse | ClickHouse | 732,539,401 | 17,300 | den-crane | try 0
```
<input_format_parallel_parsing>0</input_format_parallel_parsing>
```
restart is not needed. | 2020-11-24T02:06:55 | 2020-11-24T02:06:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,760,315 | 17,220 | alesapin | `test_storage_rabbitmq/test.py::test_rabbitmq_csv_with_delimiter` also fixed in #17311 | 2020-11-24T09:09:00 | 2020-11-24T09:09:00 | {} | MEMBER |
ClickHouse | ClickHouse | 732,892,171 | 15,419 | myrrc | So, the general query slowdown was because of perf.py bug.
The clickhouse_driver was resetting the ClickHouse session on restarts, so the old server (unpatched) started processing the queries with many threads, not 1. | 2020-11-24T11:45:35 | 2020-11-24T12:13:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,109,707 | 17,356 | den-crane | https://github.com/ClickHouse/ClickHouse/issues/14582 | 2020-11-24T17:00:10 | 2020-11-24T17:00:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,152,466 | 17,356 | maqroll | I got to ATTACH FROM eventually. Just curious about the "different semantics" that @amosbird mentions in #14582
Thanks! | 2020-11-24T18:18:56 | 2020-11-24T18:18:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,953,455 | 16,883 | nikitamikhaylov | There was a bug in mannWhitney - we have to use std::abs(z) only when alternative is 'two-sided'. Also I noticed that there is significant rounding error in cases of quite big samples (65556), so I reduced the precision on this cases. | 2020-11-24T12:49:34 | 2020-11-24T12:49:34 | {} | MEMBER |
ClickHouse | ClickHouse | 733,068,957 | 16,867 | nikitamikhaylov | > hi @nikitamikhaylov,
> did you find something related to the first problem?
Currently no, but you can create a PR with failing test. (They are in tests/integration/test_cluster_copier) | 2020-11-24T16:01:21 | 2020-11-24T16:01:21 | {} | MEMBER |
ClickHouse | ClickHouse | 733,105,408 | 17,199 | den-crane | >when I use command：systemctl start clickhouse-server1,
>the service still couldn't load, info still shows: Init script is already running.
You need to stop using " service clickhouse stop / start " , " /etc/init.d/cl ...." and switch to systemctl commands only.
make a copy of `/etc/systemd/system/clickhouse-s... | 2020-11-24T16:52:45 | 2020-11-24T16:52:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,472,882 | 17,278 | sundy-li | why not `FQDN ` with `tcp_port` ? | 2020-11-25T05:22:48 | 2020-11-25T05:22:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,666,254 | 16,867 | BhavyaRajSharma | hi @nikitamikhaylov,
did you find something related to the first problem? | 2020-11-24T05:40:11 | 2020-11-24T05:42:30 | {} | NONE |
ClickHouse | ClickHouse | 732,708,743 | 17,220 | alesapin | `01035_lc_empty_part_bug` Probably fixed in https://github.com/ClickHouse/ClickHouse/pull/17311 | 2020-11-24T07:25:27 | 2020-11-24T07:25:27 | {} | MEMBER |
ClickHouse | ClickHouse | 732,727,462 | 17,236 | xinstein | I encountered the exact same problem, how have you resolved this? @AndrewNik | 2020-11-24T08:06:16 | 2020-11-24T08:06:16 | {} | NONE |
ClickHouse | ClickHouse | 732,760,686 | 17,220 | alesapin | ` Stress test (memory) — Server failed to start ` -- not related to changes | 2020-11-24T09:09:45 | 2020-11-24T09:09:45 | {} | MEMBER |
ClickHouse | ClickHouse | 733,597,934 | 15,419 | myrrc | @akuzm Looks like all of the tests (excluding the typo in avg_weighted_long) have successfully passed. | 2020-11-25T09:52:09 | 2020-11-25T09:52:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,479,968 | 17,346 | filimonov | > after restoring a replicated table,
Most probably you did backup recovery from the same backup on 2 replicas (some ATTACH / part / partition manipulation commands). You need to recover single replica from backup, on other replicas you need only to recreate the table and it will sync the data from the other one.
... | 2020-11-23T23:12:51 | 2020-11-23T23:12:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,109,256 | 17,141 | nikitamikhaylov | Original bugfix seems to be based on some changes in `wide_integer_impl.h` which are not present in current version. | 2020-11-24T16:59:23 | 2020-11-24T16:59:23 | {} | MEMBER |
ClickHouse | ClickHouse | 733,270,046 | 17,376 | den-crane | issue: https://github.com/ClickHouse/ClickHouse/issues/7479 | 2020-11-24T22:29:12 | 2020-11-24T22:29:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,615,804 | 17,055 | l1t1 | select * from n1 limit 3 isn't the same question sql, i I run it because the sql scans little rows. I will run th e same question sql tomorrow. | 2020-11-25T10:24:01 | 2020-11-25T10:24:01 | {} | NONE |
ClickHouse | ClickHouse | 733,678,016 | 13,444 | akuzm | We now have `system.opentelemetry_span_log`, accept trace context headers over HTTP and Native protocols, add a few spans for some stages of query execution, and propagate trace context to remote ClickHouse servers and to URL engine.
https://clickhouse.tech/docs/en/operations/opentelemetry/
#14195
#16535 | 2020-11-25T12:29:35 | 2020-11-25T12:29:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,481,607 | 17,366 | zhang2014 | Integration tests (release) — fail: 11, passed: 937, error: 2 - unrelated
Integration tests (thread) — fail: 1, passed: 943, error: 0 - unrelated
Performance — 4 errors, 4 too long, 3 faster, 65 unstable - unrelated | 2020-11-25T05:48:56 | 2020-11-25T05:48:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,676,147 | 10,510 | akuzm | We added some OpenTelemetry support based on this prototype, so this PR can be closed now.
https://github.com/ClickHouse/ClickHouse/pull/14195
https://github.com/ClickHouse/ClickHouse/pull/16535 | 2020-11-25T12:25:26 | 2020-11-25T12:25:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,644,144 | 17,243 | l1t1 | > i switch to root by run sudo -i, twp rpms installed, but cannot install client, and cannot start server as service
> (all those are done in the VirtualBox included in docker toolbox )
> 
> ![20... | 2020-11-24T04:24:07 | 2020-11-24T04:24:07 | {} | NONE |
ClickHouse | ClickHouse | 732,665,981 | 16,867 | BhavyaRajSharma | @nvartolomei thanks for the solution to the second problem. | 2020-11-24T05:39:15 | 2020-11-24T05:39:15 | {} | NONE |
ClickHouse | ClickHouse | 732,995,727 | 17,344 | nikitamikhaylov | ```
==255==WARNING: MemorySanitizer: use-of-uninitialized-value
#0 0x10dcb857 in DB::parseIPv6(char const*, unsigned char*) (/usr/bin/clickhouse+0x10dcb857)
#1 0x10dc9aa3 in DB::FunctionIPv6StringToNum::executeImpl(std::__1::vector<DB::ColumnWithTypeAndName, std::__1::allocator<DB::ColumnWithTypeAndName> > c... | 2020-11-24T14:07:46 | 2020-11-24T14:07:46 | {} | MEMBER |
ClickHouse | ClickHouse | 733,061,740 | 17,311 | alesapin | ```
2020.11.24 18:48:18.605507 [ 36 ] {9a8d5322-25e1-449f-8642-4d69cf33733d} <Error> TCPHandler: Code: 27, e.displayText() = DB::Exception: Cannot parse input: expected '' before: ', 0\n': (at row 1)
Row 1:
Column 0, name: key, type: UInt64, ERROR: text ", 0<LINE FEED>" is not like UInt64
: While executing ... | 2020-11-24T15:49:28 | 2020-11-24T15:49:28 | {} | MEMBER |
ClickHouse | ClickHouse | 733,147,660 | 17,248 | nikitamikhaylov | Maybe you can add some test case? See `tests/integration/test_cluster_copier` | 2020-11-24T18:08:51 | 2020-11-24T18:08:51 | {} | MEMBER |
ClickHouse | ClickHouse | 733,471,643 | 17,199 | taylor131415 | > > 当我使用命令：systemctl start clickhouse-server1时，
> > 该服务仍然无法加载，信息仍然显示：初始化脚本已在运行。
>
> 您需要停止使用“ service clickhouse stop / start”，“ /etc/init.d/cl ....”，并仅切换到systemctl命令。
>
> 复制`/etc/systemd/system/clickhouse-server.service `
> `cp /etc/systemd/system/clickhouse-server.service /etc/systemd/system/clickhouse-server1... | 2020-11-25T05:19:34 | 2020-11-25T05:19:34 | {} | NONE |
ClickHouse | ClickHouse | 732,414,326 | 17,311 | alesapin | Seems like test_odbc_interaction/test.py::test_sqlite_odbc_hashed_dictionary really broken :( | 2020-11-23T20:41:01 | 2020-11-23T20:41:01 | {} | MEMBER |
ClickHouse | ClickHouse | 733,452,873 | 16,806 | TCeason | > @tavplubix Integration tests (asan) — Timeout :( Should I re-push a commit to rerun the test?
@robot-clickhouse
Integration tests (asan) — Timeout :(
Details
@robot-clickhouse
Integration tests (thread) — Timeout :(
timeout again.. | 2020-11-25T04:16:25 | 2020-11-25T04:16:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,513,035 | 17,311 | alesapin | `Stress test (undefined) — Killed by signal (in clickhouse-server.log) ` Will try to investigate (26.2Gb log) | 2020-11-25T07:09:20 | 2020-11-25T07:09:20 | {} | MEMBER |
ClickHouse | ClickHouse | 733,578,241 | 8,158 | filimonov | Duplicate of #7659 | 2020-11-25T09:21:29 | 2020-11-25T09:21:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,594,040 | 17,278 | filimonov | > why not `FQDN ` with `tcp_port` ?
to **anonymously** identify (also fqdn / port can change due some network reconfiguration).
| 2020-11-25T09:45:32 | 2020-11-25T09:46:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,530,750 | 17,199 | den-crane | show the output of
sudo -u clickhouse clickhouse-server --pid-file=/var/run/clickhouse-server1/clickhouse-server.pid --config-file=/etc/clickhouse-server1/config.xml | 2020-11-24T01:38:00 | 2020-11-24T01:38:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 732,668,894 | 17,199 | taylor131415 | > ```
> chown -R clickhouse.clickhouse /etc/clickhouse-server1/
> ```
>
> and again
>
> sudo -u clickhouse clickhouse-server --pid-file=/var/run/clickhouse-server/clickhouse-server-1.pid --config-file=/etc/clickhouse-server1/config1.xml
OMG, thank you so so so much, this time after executing this command：'s... | 2020-11-24T05:48:36 | 2020-11-24T05:48:36 | {} | NONE |
ClickHouse | ClickHouse | 733,066,337 | 17,274 | tavplubix | Need backport to 20.12 | 2020-11-24T15:57:06 | 2020-11-24T15:57:06 | {} | MEMBER |
ClickHouse | ClickHouse | 733,207,638 | 16,787 | lemire | We just published version 0.3.0:
- Better support for 32-bit systems
- Removed iostream dependency
- The algorithmic fast paths are more common
- Adopted slightly faster SWAR code
The repository moved to https://github.com/fastfloat/fast_float | 2020-11-24T20:11:43 | 2020-11-24T20:11:43 | {} | NONE |
ClickHouse | ClickHouse | 733,252,594 | 17,376 | alexey-milovidov | The original behaviour that we are removing was intentional, but the name was not appropriate (better would be to name it `toUnixDay` but `toUInt16` is enough). | 2020-11-24T21:48:40 | 2020-11-24T21:49:01 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 733,511,481 | 17,311 | alesapin | Release tests unlucky again.... | 2020-11-25T07:05:38 | 2020-11-25T07:05:38 | {} | MEMBER |
ClickHouse | ClickHouse | 733,515,169 | 17,378 | JIANCHUJUN | What I did is that I rename the table 'item' to be 'it' first, then I rename 'it' to be 'i' in the second part of the sql. | 2020-11-25T07:14:35 | 2020-11-25T07:14:35 | {} | NONE |
ClickHouse | ClickHouse | 733,518,011 | 16,837 | alesapin | Tests Ok. | 2020-11-25T07:21:26 | 2020-11-25T07:21:26 | {} | MEMBER |
ClickHouse | ClickHouse | 733,569,637 | 17,376 | azat | @alexey-milovidov maybe it worth marking it as "Backward Incompatible Change" instead of "Improvement"? | 2020-11-25T09:06:56 | 2020-11-25T09:06:56 | {} | MEMBER |
ClickHouse | ClickHouse | 733,582,239 | 17,346 | filimonov | > Would it work to delete the duplicate parts in zookeeper and on the nodes disk ?
It is not safe (it's hard to do it w/o braking some invariants), beside that i guess lot of those parts were merged together already. | 2020-11-25T09:28:15 | 2020-11-25T09:28:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 733,662,672 | 11,428 | PhantomPhreak | stil present in v20.11.3.3
```
# clickhouse-client -mn -u <user> --ask-password
ClickHouse client version 20.11.3.3.
Password for user (<user>): ... | 2020-11-25T11:56:16 | 2020-11-25T11:56:16 | {} | NONE |
ClickHouse | ClickHouse | 733,669,651 | 15,889 | nikitamikhaylov | Another test case:
```sql
WITH addressToSymbol(arrayJoin(trace)) AS symbol
SELECT count() > 0
FROM system.trace_log AS t
WHERE greaterOrEquals(event_date, ignore(ignore(ignore(NULL, '')), 256), yesterday()) AND (trace_type = 'Memory') AND (query_id =
(
SELECT
ignore(ignore(ignore(ignore(65536)), i... | 2020-11-25T12:11:41 | 2020-11-25T12:11:41 | {} | MEMBER |
ClickHouse | ClickHouse | 733,669,956 | 17,363 | nikitamikhaylov | AST Fizzer failure #15889 | 2020-11-25T12:12:21 | 2020-11-25T12:12:21 | {} | MEMBER |
ClickHouse | ClickHouse | 709,096,452 | 8,802 | mateng0915 | Does this join predicate pushdown feature is under implementation ? | 2020-10-15T10:20:23 | 2020-10-15T10:20:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 709,303,329 | 15,199 | alexey-milovidov | @filimonov Check server log for exception with logical error. | 2020-10-15T12:52:35 | 2020-10-15T12:52:35 | {} | MEMBER |
ClickHouse | ClickHouse | 709,322,469 | 15,989 | den-crane | Because they contain different number of rows. Check the **rows** column.
_0_1607 -- means that it's a zero level part -- just inserted. You inserted 65k rows by an insert.
Later this parts will be merged into bigger parts. _1 _2 _3 level 1,2,3
https://clickhouse.tech/docs/en/development/architecture/#merge-tr... | 2020-10-15T13:24:50 | 2020-10-15T13:28:02 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.