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 | 777,219,706 | 20,174 | amosbird | Why are there so many flaky check failures...
| 2021-02-11T05:57:09 | 2021-02-11T05:57:09 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 777,281,225 | 19,820 | qoega | @kitaisreal have you checked what was the problem? | 2021-02-11T08:41:12 | 2021-02-11T08:41:12 | {} | MEMBER |
ClickHouse | ClickHouse | 776,805,318 | 20,275 | filimonov | > Client sends all changed settings to the server and server accepts it if not readonly.
But the opposite way it doesn't work.
A sequence like that:
1) the user hit "Maximum parse depth (1000) exceeded. Consider rising max_parser_depth parameter."
2) he adjusts that setting in the default profile server side.
... | 2021-02-10T15:47:47 | 2021-02-10T15:47:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,242,391 | 16,991 | vdimir | > and `ENGINE = Join(ALL, INNER, serial2);` is ok
Thanks for your notice it really helps. Actually it depend on type of join, bug was connected with `ANY INNER` join (and some others `RIGHT`, `FULL`).
But I see another problem here, we cannot use `JOIN ON (test.serial == test_join.serial)` an should rename it to ... | 2021-02-11T07:05:16 | 2021-02-11T07:05:16 | {} | MEMBER |
ClickHouse | ClickHouse | 777,262,737 | 20,300 | alesapin | > What about Sign column?
I'll check and make separate PR. | 2021-02-11T08:02:53 | 2021-02-11T08:02:53 | {} | MEMBER |
ClickHouse | ClickHouse | 777,325,725 | 19,685 | vdimir | It's slightly related to #19856, not I'm trying to realize can it be solved independently or we should make changes common for both cases. | 2021-02-11T09:56:38 | 2021-02-11T09:56:38 | {} | MEMBER |
ClickHouse | ClickHouse | 777,414,117 | 19,516 | abyss7 | Fuzzer (debug): #20046 | 2021-02-11T12:21:26 | 2021-02-11T12:21:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,537,158 | 17,708 | hombit | @alexey-milovidov Sorry, I don't feel I have enough time to do it next few weeks | 2021-02-10T08:35:20 | 2021-02-10T08:35:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,547,973 | 20,231 | CurtizJ | Fuzzer error: #18793 | 2021-02-10T08:54:34 | 2021-02-10T08:54:49 | {} | MEMBER |
ClickHouse | ClickHouse | 776,552,551 | 20,087 | Felixoid | Hey @ervikrant06 .
If you want to "drop" the metrics behind some age, then there's no way to do it besides writing the custom SQL.
On the other hand, here's a description of the difference between whisper and GraphiteMergeTree policies https://github.com/lomik/graphite-clickhouse/wiki/ClickHouse-aggregation-VS-g... | 2021-02-10T09:02:21 | 2021-02-10T09:02:21 | {} | MEMBER |
ClickHouse | ClickHouse | 776,713,094 | 20,275 | alexey-milovidov | > the fact that parser settings (like max_parser_depth) should be adjusted BOTH on client & server
Client sends all changed settings to the server and server accepts it if not readonly.
BTW. I discourage changing parsing settings, we should improve something to make it unneeded. | 2021-02-10T13:41:39 | 2021-02-10T13:41:39 | {} | MEMBER |
ClickHouse | ClickHouse | 776,801,387 | 20,231 | CurtizJ | Leave it as is. In fuzz check all queries are being fuzzed and sometimes it finds random errors, which are not related to PR changes. | 2021-02-10T15:42:05 | 2021-02-10T15:42:05 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 777,131,271 | 3,852 | fzyzcjy | @azat Sorry I mean "apply only if not null", not "any first value". (Maybe I am wrong) | 2021-02-11T00:25:51 | 2021-02-11T00:25:51 | {} | NONE |
ClickHouse | ClickHouse | 777,351,152 | 19,719 | nikitamikhaylov | It reproduces easily
```
python3 -c "for i in range(10):print('{{\"a\":\"{}\", \"b\":\"{}\"'.format('clickhouse'* 1000000, 'dbms' * 1000000))" > big_json.json
python3 -c "for i in range(1000):print('{{\"a\":\"{}\", \"b\":\"{}\"}}'.format('clickhouse'* 1000000, 'dbms' * 1000000))" > big_json.json
```
```
↳ $ l... | 2021-02-11T10:38:11 | 2021-02-11T10:38:11 | {} | MEMBER |
ClickHouse | ClickHouse | 777,423,369 | 3,852 | azat | >An interesting question would be if it is possible to reverse that behaviour and also consider null (to blank out previously set column values).
There is no such setting or aggregate function, maybe workaround using `ifNull(null, some_default_value)` will help to do what you need.
And AFAIK this is SQL standard (a... | 2021-02-11T12:36:02 | 2021-02-11T12:36:02 | {} | MEMBER |
ClickHouse | ClickHouse | 776,704,457 | 20,253 | fuCtor | > > We can successful bind user and login to clickhouse with i.i.ivanov or s.v.petrov.
>
> Where is this specified? How do the records above associate to these user names?
In config:
```
<bind_dn>DOMAIN\{user_name}</bind_dn>
```
It's allow bind user. AD support this DN string and if I'm made request from cons... | 2021-02-10T13:26:26 | 2021-02-10T13:26:26 | {} | NONE |
ClickHouse | ClickHouse | 776,712,177 | 20,253 | traceon | What is the output of
```sh
dsquery user -name s.v.petrov
```
and
```sh
dsquery group -name users_l0.2
```
? | 2021-02-10T13:40:04 | 2021-02-10T13:40:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,219,761 | 20,174 | amosbird | It seems function name normalization will break old instances when doing distributed query | 2021-02-11T05:57:19 | 2021-02-11T05:57:19 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 777,256,104 | 18,439 | amosbird | Closed as lack of interests. | 2021-02-11T07:44:38 | 2021-02-11T07:44:38 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 777,317,892 | 20,139 | nvartolomei | @sonirico unfortunately I don't have merge permissions. Maybe @alexey-milovidov can take a quick look at this tiny change? | 2021-02-11T09:43:49 | 2021-02-11T09:43:49 | {
"eyes": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,479,264 | 20,308 | alexey-milovidov | CatBoost is another repository, you should create an issue at https://github.com/catboost/catboost | 2021-02-11T14:01:09 | 2021-02-11T14:01:09 | {} | MEMBER |
ClickHouse | ClickHouse | 776,570,409 | 20,251 | KochetovNicolai | > Is it possible to make a test?
I suppose it will be pretty difficult.
This bug was found pretty fast by our tests, so I suppose if it is not fixed, we will find it soon. | 2021-02-10T09:28:24 | 2021-02-10T09:28:24 | {} | MEMBER |
ClickHouse | ClickHouse | 776,682,209 | 8,646 | whoIsJackson | @litaoyx
大佬不介意的话，能不能把微信留一下，我想找你请教请教，哈哈哈哈。 | 2021-02-10T12:45:22 | 2021-02-10T12:45:22 | {} | NONE |
ClickHouse | ClickHouse | 776,757,305 | 20,097 | KochetovNicolai | > What is the main purpose of these changes? Seems like we want to get rid of manipulation with individual actions and do everything with DAG.
Exactly.
If something may be expressed in ActionsDAG, it most likely need to be expressed this way because:
* DAG may be build once, then executed on block stream. It shoul... | 2021-02-10T14:46:47 | 2021-02-10T14:46:47 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 776,847,717 | 20,277 | filimonov | +1.
I've put discussion - because it's not clear if it really needed and will increase the number of things to support.
Advantages of the snap are one-line installation, existence in the listing of snap applications in ubuntu store, sandbox (nice to try w/o bloating the system).
Disadvantages: one more distri... | 2021-02-10T16:44:30 | 2021-02-10T16:44:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,910,572 | 20,288 | den-crane | https://youtu.be/ckChUkC3Pns?t=10715
https://den-crane.github.io/Everything_you_should_know_about_materialized_views_commented.pdf | 2021-02-10T18:16:40 | 2021-02-10T18:16:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,971,658 | 19,580 | alesapin | Todo: simplify startup | 2021-02-10T19:50:16 | 2021-02-10T19:50:16 | {} | MEMBER |
ClickHouse | ClickHouse | 777,348,429 | 3,852 | gersmann | > AFAIR right now it should work this way, however I doubt that this is a good thing to relay on (maybe @alexey-milovidov will correct me here).
Hm, @fzyzcjy, according to Clickhouse documentation, null values are skipped in aggregation. So I think it should be possible to rely on this. An interesting question would... | 2021-02-11T10:33:47 | 2021-02-11T10:33:47 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 777,413,869 | 17,623 | filimonov | > I think I can look into this.
@bharatnc Please check the code around, and write down a task / short plan of implementation.
We need a bless from @vitlibar and may be some sync with @traceon. After that you can take the task. | 2021-02-11T12:21:01 | 2021-02-11T12:21:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,825,558 | 20,265 | nvartolomei | Upon more reading of the code it seems that this data race is benign as the `trace_log` is always pointing to a valid memory location. I can't find any code paths where loggers are deallocated. | 2021-02-10T16:15:29 | 2021-02-10T16:15:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,828,417 | 20,203 | SaltTan | 20.12 is affected as well
```
create table i20203_1 (a Int8) engine=ReplicatedMergeTree('/clickhouse/tables/i20203','r1') order by tuple();
create table i20203_2 (a Int8) engine=ReplicatedMergeTree('/clickhouse/tables/i20203','r2') order by tuple();
detach table i20203_2;
insert into i20203_1 values (2);
detach t... | 2021-02-10T16:19:15 | 2021-02-10T16:19:15 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,848,381 | 20,277 | filimonov | Let's keep it open and collect feedback. | 2021-02-10T16:45:27 | 2021-02-10T16:45:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,940,309 | 20,215 | tavplubix | Do I understand correctly, that we can safely execute `StorageDistributedDirectoryMonitor::shutdownAndDropAllData()` and `StorageDistributedDirectoryMonitor::flushAllData()` concurrently?
| 2021-02-10T19:04:54 | 2021-02-10T19:04:54 | {} | MEMBER |
ClickHouse | ClickHouse | 777,188,158 | 19,627 | Slach | @filimonov my 50 cents questions
* what should happen when we execute `ALTER TABLE ... FREEZE PARTITION` and after its runs, we run the `DELETE statement`?
* what should happen when we execute `UPDATE` on different nodes with different `SET` clauses but with the same conditions for WHERE clause which affect the sa... | 2021-02-11T03:55:02 | 2021-02-11T03:55:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,255,937 | 16,096 | amosbird | Closed as it's not useful. | 2021-02-11T07:44:10 | 2021-02-11T07:44:10 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 776,605,239 | 19,580 | alesapin | Todo: implement `sync` request:
https://github.com/python-zk/kazoo/blob/master/kazoo/protocol/serialization.py#L263-L271
https://github.com/apache/zookeeper/blob/master/zookeeper-jute/src/main/resources/zookeeper.jute#L179-L184
https://zookeeper.apache.org/doc/r3.4.2/api/org/apache/zookeeper/ZooKeeper.html#sync(java... | 2021-02-10T10:22:21 | 2021-02-10T10:22:21 | {} | MEMBER |
ClickHouse | ClickHouse | 776,627,719 | 20,046 | tavplubix | It can be reproduced with
```
:) CREATE TABLE t1 (x UInt32, y UInt64) engine = MergeTree ORDER BY (x,y);
:) CREATE TABLE t2 (x UInt32, y UInt64) engine = MergeTree ORDER BY (x,y);
:) SET join_algorithm = 'prefer_partial_merge';
:) SET any_join_distinct_right_table_keys = 1;
:) SET join_use_nulls = 1;
:) SELECT i... | 2021-02-10T10:59:56 | 2021-02-10T10:59:56 | {} | MEMBER |
ClickHouse | ClickHouse | 776,685,939 | 20,274 | alexey-milovidov | This PR is:
- for review and comments;
- not for merge, because final implementation will be different. | 2021-02-10T12:52:52 | 2021-02-10T12:52:52 | {} | MEMBER |
ClickHouse | ClickHouse | 776,706,289 | 20,280 | alexey-milovidov | As far as I remember, shell is set to `/bin/false` for security reasons to avoid the possibility that you accidentially allow someone to login as clickhouse. We also set home directory to `/nonexistent`.
If you set shell to `/bin/bash` but no one from outside can log in to your server as clickhouse, it is still secu... | 2021-02-10T13:29:35 | 2021-02-10T13:29:35 | {} | MEMBER |
ClickHouse | ClickHouse | 776,710,362 | 20,253 | traceon | OU is a "group", you usually put references there (the `member=CN=Петров Сергей Владимирович` attribute), to the actual DN with user definition. So I am asking, where that user definition is located?
| 2021-02-10T13:36:51 | 2021-02-10T13:36:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,915,390 | 20,288 | ilugid | Thanks!! That helped. | 2021-02-10T18:24:26 | 2021-02-10T18:24:26 | {} | NONE |
ClickHouse | ClickHouse | 777,252,623 | 19,721 | azat | >Stress test (thread) — Sanitizer assert (in stderr.log)
```
Stats: LargeMmapAllocator: allocated 377831 times, remains 4869 (3693112 K) max 6117 M; by size logs: 17:23222; 18:72316; 19:107149; 20:143857; 21:29088; 22:1313; 23:673; 24:147; 25:66;
==299==Launching Symbolizer process: /usr/bin/llvm-symbolizer-11 --... | 2021-02-11T07:34:04 | 2021-02-11T07:34:04 | {} | MEMBER |
ClickHouse | ClickHouse | 777,321,701 | 20,280 | filimonov | ```
sudo -u clickhouse HOME=/tmp bash
sudo su clickhouse -s /usr/bin/bash
```
etc. | 2021-02-11T09:50:00 | 2021-02-11T09:50:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,638,191 | 20,253 | fuCtor | ```
CN=Петров Сергей Владимирович,OU=users_l0.2,OU=users_l0,DC=domain,DC=tld objectClass=person,top,user
CN=Смирнов Дмитрий Витальевич,OU=users_l0.2,OU=users_l0,DC=domain,DC=tld objectClass=person,top,user
OU=users_l0.2.1,OU=users_l0.2,OU=users_l0,DC=domain,DC=tld objectClass=top,organizationalUnit
OU=users_l0.2.2... | 2021-02-10T11:18:34 | 2021-02-10T12:34:16 | {} | NONE |
ClickHouse | ClickHouse | 776,680,448 | 8,646 | whoIsJackson | @litaoyx hello，i am you fans in CSDN。Recently, I am studying the bitmap of Clickhouse. I want to consult the Variant type in the following article. Have you implemented the variant in Java?
https://blog.csdn.net/qq_27639777/article/details/111005838 | 2021-02-10T12:41:55 | 2021-02-10T12:41:55 | {} | NONE |
ClickHouse | ClickHouse | 776,839,900 | 20,079 | vdimir | @CurtizJ Yes, flags were stored inside hash table with row refs. So, it was shared between `HashJoin` instances pointing at the same `RightTableData`. To reset flags but do not rebuild whole hash table I split flags into separate array that have same size as hash table buffer and keep correspondence between it. | 2021-02-10T16:34:31 | 2021-02-10T16:34:31 | {} | MEMBER |
ClickHouse | ClickHouse | 776,902,127 | 20,017 | abyss7 | Another query: `SELECT item_id FROM (SELECT item_id FROM t GROUP BY item_id WITH TOTALS) AS l FULL OUTER JOIN (SELECT item_id FROM t GROUP BY ((NULL, '0.0000065537', '0', tuple(tuple(tuple(NULL))), -9223372036854775808, NULL), (NULL, '0.0001048577', (('100', NULL), tuple((256, NULL))), NULL), NULL), item_id WITH TOTALS... | 2021-02-10T18:02:31 | 2021-02-10T18:02:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,933,880 | 19,257 | nikitamikhaylov | I benchmarked a function for area calculation in comparison to Postgis. Hopefully, there will be some benchmarks for all of functions in the near future.
Dataset
https://datahub.io/core/geo-countries#readme
```python
from shapely.geometry import shape
import geojson
import wget
url = 'https://datahub.io/co... | 2021-02-10T18:54:26 | 2021-02-10T18:54:26 | {
"rocket": 1
} | MEMBER |
ClickHouse | ClickHouse | 776,941,294 | 20,273 | SaltTan | Any column from system.processes can be used in the WHERE section of KILL QUERY | 2021-02-10T19:06:33 | 2021-02-10T19:06:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,013,657 | 3,852 | azat | Reverse version of `anyLast` is `any`, you can use it
AFAIR right now it should work this way, however I doubt that this is a good thing to relay on (maybe @alexey-milovidov will correct me here). | 2021-02-10T20:36:06 | 2021-02-10T20:36:06 | {} | MEMBER |
ClickHouse | ClickHouse | 777,072,293 | 19,673 | azat | >Functional stateless tests (ANTLR debug) — fail: 1, passed: 2531, skipped: 231
#20305 | 2021-02-10T22:15:52 | 2021-02-10T22:15:52 | {} | MEMBER |
ClickHouse | ClickHouse | 777,230,688 | 3,852 | azat | @fzyzcjy It will apply any first non-null value, or NULL if there is no non-null value
```sql
create table t2 (x SimpleAggregateFunction(any, Nullable(UInt64))) engine=Memory;
insert into t2 values (null);
insert into t2 values (1);
insert into t2 values (2);
select any(x) from t2;
┌─any(x)─┐
│ 1 │
└───... | 2021-02-11T06:31:11 | 2021-02-11T06:31:11 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 777,327,057 | 3,852 | fzyzcjy | @azat Aha sounds interesting - Thanks! | 2021-02-11T09:58:54 | 2021-02-11T09:58:54 | {} | NONE |
ClickHouse | ClickHouse | 776,533,233 | 20,034 | YiuRULE | The error doesn't seem to be reproduced to `20.10.7`, we tried to update it using the latest `20.9.*` but no change on that side.
| 2021-02-10T08:27:28 | 2021-02-10T08:27:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,948,064 | 20,273 | den-crane | yeah, like
KILL QUERY WHERE elapsed > 600 | 2021-02-10T19:17:01 | 2021-02-10T19:17:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 775,690,648 | 20,089 | bharatnc | > The purpose is to uniquely identify installed ClickHouse instance across the network.
> It can be useful for ON CLUSTER queries, cross-replication, server discovery...
>
> PS. Maybe we can generate it in Install.cpp and only read then.
Thank you for the review, I think I get the gist. I will make the necessary... | 2021-02-09T05:59:42 | 2021-02-11T02:30:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,212,563 | 20,307 | towelenee | Sorry wrong repo | 2021-02-11T05:32:59 | 2021-02-11T05:32:59 | {} | NONE |
ClickHouse | ClickHouse | 777,266,659 | 20,224 | alesapin | Something wrong with compose in integration tests:
```
$ find test_user_ip_restrictions/_instances/ -name 'clickhouse-server.log'
$ find test_user_ip_restrictions/_instances/ -name 'stderr.log'
$ cat test_user_ip_restrictions/_instances/docker.log
Attaching to
```
But no related failures. | 2021-02-11T08:11:25 | 2021-02-11T08:11:25 | {} | MEMBER |
ClickHouse | ClickHouse | 777,322,465 | 19,580 | alesapin | > Todo: simplify startup
done | 2021-02-11T09:51:17 | 2021-02-11T09:51:17 | {} | MEMBER |
ClickHouse | ClickHouse | 776,720,079 | 20,227 | akuzm | The time has not yet come. The test is bad, and we're getting the same timeout in parsing of complex queries. We need a principled way to handle this:
```
Alexander Kuzmenkov, [10.02.21 13:57]
У нас есть функция, которую можно вызывать в произвольных местах, чтобы проверить, не случился ли таймаут выполнения запро... | 2021-02-10T13:53:30 | 2021-02-10T13:53:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,065,288 | 2,241 | corpix | Corresponding to https://clickhouse.tech/docs/en/operations/server-configuration-parameters/settings/ this is not working anymore
UPD
Hacky workaround:
```xml
<log>/proc/self/fd/1</log>
<errorlog>/proc/self/fd/2</errorlog>
```
But it would be better if logging to stdout&stderr will be simpler than this. | 2021-02-10T22:01:54 | 2021-02-10T22:08:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,125,459 | 20,304 | alexey-milovidov | Yes, it's possible to skip INSERT completely if the block is empty. | 2021-02-11T00:10:30 | 2021-02-11T00:10:30 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 776,514,347 | 19,580 | alesapin | Ok, now the ugliest part is the startup process. NuRaft requires one server (future leader) to add other servers (future followers) to the RAFT quorum. Other servers just wait until they receive a request to join. Now I just choose one of the servers (based on priority or lexicographical order) and other servers just w... | 2021-02-10T07:49:14 | 2021-02-10T07:49:14 | {} | MEMBER |
ClickHouse | ClickHouse | 776,654,156 | 3,123 | crakjie | @den-crane
Ok, my bad | 2021-02-10T11:49:59 | 2021-02-10T11:50:16 | {} | NONE |
ClickHouse | ClickHouse | 776,709,782 | 20,277 | alexey-milovidov | Note that Snap, AppImage, Flatpak are always worse than our no-dependencies single binary build.
We can consider building Snap only for inclusion in the official Ubuntu repository. | 2021-02-10T13:35:51 | 2021-02-10T13:35:51 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 776,718,222 | 17,708 | alexey-milovidov | Ok. This task is easy to do. Maybe someone can help... | 2021-02-10T13:50:28 | 2021-02-10T13:50:28 | {} | MEMBER |
ClickHouse | ClickHouse | 776,701,165 | 19,580 | alesapin | > Todo: implement `sync` request:
> https://github.com/python-zk/kazoo/blob/master/kazoo/protocol/serialization.py#L263-L271
> https://github.com/apache/zookeeper/blob/master/zookeeper-jute/src/main/resources/zookeeper.jute#L179-L184
> https://zookeeper.apache.org/doc/r3.4.2/api/org/apache/zookeeper/ZooKeeper.html#s... | 2021-02-10T13:20:44 | 2021-02-10T13:20:44 | {} | MEMBER |
ClickHouse | ClickHouse | 776,791,838 | 20,231 | fuwhu | > Fuzzer error: #18793
@CurtizJ
How shall i fix this error or just leave it as it is? | 2021-02-10T15:28:37 | 2021-02-10T15:29:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,834,075 | 20,203 | filimonov | https://github.com/ClickHouse/ClickHouse/pull/15983/ ? | 2021-02-10T16:27:15 | 2021-02-10T16:27:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,999,059 | 18,369 | azat | @sliontc this is completely different issue, it is better to create new issue for this problem. | 2021-02-10T20:16:48 | 2021-02-10T20:16:48 | {} | MEMBER |
ClickHouse | ClickHouse | 777,171,774 | 20,089 | bharatnc | Simplified the logic, but it still lives inside the `Server.cpp` file (mainly to accommodate various installation scenarios as stated above), but can easily relocate it to `Install.cpp` file if you want me to. Thank you. | 2021-02-11T02:42:48 | 2021-02-11T02:42:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,928,960 | 18,282 | abyss7 | Additional consideration about access rights: in practice they can change for an user between inserts. But checking them too often will beat the whole idea. So the point is to check them not too often for each user - maybe the interval should be adjustable via setting. | 2021-02-10T18:46:21 | 2021-02-10T18:46:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,079,569 | 18,247 | SaltTan | I would check the logs on all nodes to see what DDLWorker threads were doing. | 2021-02-10T22:31:17 | 2021-02-10T22:31:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,367,012 | 20,220 | stigsb | Could you explain a bit more? What is your question? Can you post ClickHouse server logs for the same time period?
| 2021-02-11T11:03:46 | 2021-02-11T11:03:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,620,679 | 20,274 | akuzm | The Fast Test is broken (doesn't handle some GitHub behavior correctly), I'll fix it. | 2021-02-10T10:47:24 | 2021-02-10T10:47:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,689,916 | 20,253 | traceon | > We can successful bind user and login to clickhouse with i.i.ivanov or s.v.petrov.
Where is this specified? How do the records above associate to these user names? | 2021-02-10T13:00:36 | 2021-02-10T13:00:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,808,349 | 20,283 | filimonov | Confirmed. It's a bug. | 2021-02-10T15:52:07 | 2021-02-10T15:52:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,268,305 | 20,224 | alesapin | msan https://github.com/ClickHouse/ClickHouse/issues/20314 | 2021-02-11T08:14:33 | 2021-02-11T08:14:33 | {} | MEMBER |
ClickHouse | ClickHouse | 777,351,042 | 19,719 | nikitamikhaylov | > Just to be sure: Are we talking about JSON or JSONEachRow?
JSONEachRow, because JSON is not supported for input.
| 2021-02-11T10:38:00 | 2021-02-11T10:38:00 | {} | MEMBER |
ClickHouse | ClickHouse | 776,741,151 | 18,369 | sliontc | > @sliontc What error do you have?
> Can you post the error with the stacktrace included? (and also post the structure of the related tables)
I use the MaterializeMySQL Engine to replicate data from MySQL. At the beginning, I can got select query executed and data returned. But later when execute select query, the ... | 2021-02-10T14:25:06 | 2021-02-10T14:25:06 | {} | NONE |
ClickHouse | ClickHouse | 776,827,146 | 20,276 | alesapin | Done https://clickhouse-test-reports.s3.yandex.net/19912/0979308e68ac9a1ad0d49df019955dca3cd8cf50/push_to_dockerhub.html. Will check during next release. | 2021-02-10T16:17:38 | 2021-02-10T16:17:38 | {
"hooray": 2
} | MEMBER |
ClickHouse | ClickHouse | 776,835,692 | 20,203 | SaltTan | 20.11.5.18
```
SELECT
create_time,
num_tries,
last_attempt_time
FROM system.replication_queue
WHERE table = 'i20203_2'
┌─────────create_time─┬─num_tries─┬───last_attempt_time─┐
│ 2021-02-10 16:26:56 │ 36 │ 2021-02-10 16:28:36 │
└─────────────────────┴───────────┴─────────────────────┘
``... | 2021-02-10T16:29:29 | 2021-02-10T16:29:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,917,754 | 20,175 | amosbird | Test failure looks unrelated. | 2021-02-10T18:28:19 | 2021-02-10T18:28:19 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 776,993,459 | 20,215 | azat | >Do I understand correctly, that we can safely execute StorageDistributedDirectoryMonitor::shutdownAndDropAllData() and StorageDistributedDirectoryMonitor::flushAllData() concurrently?
@tavplubix nice catch!
We can, but this will allow removing files that are sending right now, so not a good idea, fixed in 809fa7e4... | 2021-02-10T20:12:27 | 2021-02-10T20:12:27 | {} | MEMBER |
ClickHouse | ClickHouse | 777,128,754 | 20,300 | alexey-milovidov | What about Sign column? | 2021-02-11T00:19:28 | 2021-02-11T00:19:28 | {} | MEMBER |
ClickHouse | ClickHouse | 777,176,056 | 19,356 | bharatnc | Duplicate https://github.com/ClickHouse/ClickHouse/issues/19950.
Try this fix:
> run client in multiline/multiquery more clickhouse-client -mn
(https://github.com/ClickHouse/ClickHouse/issues/19950#issuecomment-772770123)
| 2021-02-11T03:01:33 | 2021-02-11T03:01:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,214,589 | 17,623 | bharatnc | > > Pluggable auth with tokens
>
> Anyone working on that? I have some draft ideas on how this should look like but no time to work on it.
>
> I was thinking about SASL for native protocol (interested in OAUTHBEARER mechanism) and JWT support for HTTP protocol (pretty much same same code as for the OAUTHBEARER). ... | 2021-02-11T05:39:58 | 2021-02-11T05:39:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,242,646 | 16,991 | vdimir | Issue from initial comment with StorageJoin that works only once fixed with PR #20079 | 2021-02-11T07:05:55 | 2021-02-11T07:05:55 | {} | MEMBER |
ClickHouse | ClickHouse | 777,270,081 | 20,224 | alesapin | abort: https://github.com/ClickHouse/ClickHouse/issues/20315 | 2021-02-11T08:17:55 | 2021-02-11T08:17:55 | {} | MEMBER |
ClickHouse | ClickHouse | 777,313,683 | 20,139 | sonirico | @nvartolomei Hi! Any progress on this? Looks like some automated checks have failed, but they don't seem to be related to this PR. | 2021-02-11T09:36:46 | 2021-02-11T11:30:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,425,359 | 17,623 | traceon | When integrating SASL in native TCP protocol, bear in mind that Kerberos support may eventually be added there too, so the integration better be reasonably generic. | 2021-02-11T12:39:08 | 2021-02-11T12:39:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,708,508 | 20,278 | alexey-milovidov | Sanitizers cannot instrument pre-built shared libraries, including ODBC drivers.
Your report shows the context from inside the ODBC driver.
Our test suite has the tests with ODBC in skip list for sanitizers. | 2021-02-10T13:33:36 | 2021-02-10T13:33:36 | {} | MEMBER |
ClickHouse | ClickHouse | 776,718,927 | 20,034 | alexey-milovidov | Ok, if there is no such issue in 20.10, it means that versions 21.1 and 21.2 are definitely alright. | 2021-02-10T13:51:39 | 2021-02-10T13:51:39 | {} | MEMBER |
ClickHouse | ClickHouse | 776,771,649 | 20,285 | filimonov | EOS consumer (isolation.level=read_committed) is enabled by default since librdkafka 1.2.0, so for clickhouse - since 20.2
See
* https://github.com/edenhill/librdkafka/commit/6b2a1552ac2a4ea09d915015183f268dd2df96e6
* https://github.com/ClickHouse/ClickHouse/pull/9000/commits/9de5dffb5c97eb93545ae25eaf87ec195a590... | 2021-02-10T15:07:40 | 2021-02-10T15:49:44 | {
"+1": 1,
"eyes": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 776,809,953 | 20,283 | filimonov | 20.12 and before are not affected. Thanks for the report! | 2021-02-10T15:54:18 | 2021-02-10T15:54:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 777,235,841 | 16,991 | vdimir | > Add new ANY JOIN logic for StorageJoin consistent with JOIN operation. To upgrade without changes in behaviour you need add SETTINGS any_join_distinct_right_table_keys = 1 to Engine Join tables metadata or recreate these tables after upgrade. #8400
@den-crane
Cannot reproduce it with provided queries. Is there o... | 2021-02-11T06:46:55 | 2021-02-11T06:47:22 | {} | MEMBER |
ClickHouse | ClickHouse | 777,281,709 | 20,301 | azat | >Functional stateless tests (ANTLR debug) — fail: 1, passed: 84, skipped: 3
This is the problem of ANTLR parser (after #20175) | 2021-02-11T08:42:23 | 2021-02-11T08:42:23 | {} | MEMBER |
ClickHouse | ClickHouse | 771,694,048 | 19,299 | akuzm | Fuzzer https://github.com/ClickHouse/ClickHouse/issues/19579 | 2021-02-02T14:58:53 | 2021-02-02T14:58:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 771,818,509 | 19,897 | funemy | And regarding `ccache`,
It can potentially cause problems. Say, if some part of the program is cached and thus not re-compiled, then that part will be missing from the bc file, thus Coderrect cannot analyze that part. | 2021-02-02T17:29:14 | 2021-02-02T17:29:14 | {} | NONE |
ClickHouse | ClickHouse | 771,855,792 | 19,897 | alexey-milovidov | > I've compiled clickhouse, and below is the bc file I got. clickhouse.bc only has 276K, so I assume it's just an entry program that dispatches workloads to other programs?
Yes, it is the entry point, but it calls other tools directly - all of them are statically linked inside it. | 2021-02-02T18:11:34 | 2021-02-02T18:11:34 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.