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 | 802,466,587 | 21,849 | jinghancc | > @jinghancc Yes, you are right, I contributed these codes 3 years ago and I almost forgot that. Could you send a pr to improve that?
@sundy-li by the way, I have another question about `windowFunnel` which is described [here](https://github.com/ClickHouse/ClickHouse/issues/21835), do you mind taking a look when you... | 2021-03-19T02:11:30 | 2021-03-19T02:11:30 | {} | NONE |
ClickHouse | ClickHouse | 802,520,834 | 18,772 | Zhile | @azat Does this also happens in 20.12? If yes, can we also cherry-pick the fix to 20.12? | 2021-03-19T03:29:14 | 2021-03-19T03:29:14 | {} | NONE |
ClickHouse | ClickHouse | 802,538,861 | 16,726 | bohutang | I guess no plans yet. | 2021-03-19T04:12:27 | 2021-03-19T04:12:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,811,567 | 21,905 | mic4ael | The configuration which works on 20.8.x should also work on 21.2.x, right? For some reason it doesn't | 2021-03-19T12:52:05 | 2021-03-19T12:52:05 | {} | NONE |
ClickHouse | ClickHouse | 802,822,191 | 21,909 | vorillaz | Hey @den-crane, thanks for the response.
Is there any chance I could use aggregated materialized views along with collapsing rows without manually data syncing with a cron Job? | 2021-03-19T13:11:14 | 2021-03-19T13:11:14 | {} | NONE |
ClickHouse | ClickHouse | 803,230,937 | 21,815 | TCeason | Ok I will test this issue later. | 2021-03-20T02:41:57 | 2021-03-20T02:41:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,186,227 | 21,851 | djmarcin | I should have mentioned that this also fails for whitespace characters such as 0x20 and 0x0a. While the comma case is apparently fixed by this workaround, the space cases do not appear to be. --edit: It looks like 0x20 does work but 0x0a does not. 0x03 also does not appear to work.
This is successful:
```
(echo -n... | 2021-03-18T18:27:42 | 2021-03-18T18:44:07 | {} | NONE |
ClickHouse | ClickHouse | 802,495,231 | 21,894 | entropical | The drop was taking forever. I think it was attempting to catch up on the pending propogations first.
I moved the directory to a safe location, and the drop completed a few seconds later. I then recreated the distributed table, and started my writers. For the last hour it looks like the propagations are working fine... | 2021-03-19T02:38:48 | 2021-03-19T02:38:48 | {} | NONE |
ClickHouse | ClickHouse | 802,800,354 | 18,655 | alexey-milovidov | #21479 | 2021-03-19T12:30:50 | 2021-03-19T12:30:50 | {} | MEMBER |
ClickHouse | ClickHouse | 802,820,087 | 21,909 | den-crane | Materialized view is an insert trigger. It knows nothing about an engine of the base table.
https://den-crane.github.io/Everything_you_should_know_about_materialized_views_commented.pdf
https://youtu.be/ckChUkC3Pns?t=9326 | 2021-03-19T13:07:19 | 2021-03-19T13:07:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,892,158 | 21,835 | vdimir | > Do you think it's reasonable to introduce a new mode to make the timestamp `strictly increasing` to avoid counting duplicate timestamps into result?
I think adding such option may be possible, but it seems that function will do some additional work (filtering), for which it is not responsible. Why filtering rows w... | 2021-03-19T14:54:21 | 2021-03-19T14:54:21 | {} | MEMBER |
ClickHouse | ClickHouse | 803,044,480 | 21,896 | alexey-milovidov | `isInt64orUInt64FieldType` | 2021-03-19T18:56:29 | 2021-03-19T18:56:39 | {} | MEMBER |
ClickHouse | ClickHouse | 802,298,100 | 21,677 | alesapin | `queue` test is awesome. Found bug with snapshots apply, and something more interesting (currently not investigated):
```
INFO [2021-03-18 23:50:16,269] jepsen test runner - jepsen.core {:perf {:latency-graph {:valid? true},
:rate-graph {:valid? true},
:valid? true},
:workload {:ok-count 696,
... | 2021-03-18T21:04:40 | 2021-03-18T21:04:55 | {} | MEMBER |
ClickHouse | ClickHouse | 799,289,385 | 19,225 | maimaip | > > when i upgrade from 20.10.2.20 to 21.3.2 but i use comperss code is LZ4HC(6).then same error is happen。after 2 hour,my clickhouse-sever is restart every once in a while。
> > 
>
> Hi, this looks quite w... | 2021-03-15T10:03:25 | 2021-03-19T06:05:34 | {} | NONE |
ClickHouse | ClickHouse | 802,623,929 | 21,142 | Jokser | @alesapin Could you please check the change again? | 2021-03-19T07:43:50 | 2021-03-19T07:43:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,642,169 | 18,772 | azat | > I found the fix is not in 20.12 branch. Can you help to confirm?
> https://github.com/ClickHouse/ClickHouse/blob/20.12/src/Storages/MergeTree/MergeList.cpp
Yes, you are right, by some reason it has been reverted in 83ee3cbf538d372900b324a3617cc7242ef49564
Maybe @alesapin can comment on this?
> Does the new ve... | 2021-03-19T08:18:31 | 2021-03-19T08:18:31 | {} | MEMBER |
ClickHouse | ClickHouse | 802,798,651 | 21,905 | mic4ael | > It was intentional #18740 ?
>
> So you saying that even implicit incl does not work anymore?
>
> ```
> /etc/clickhouse-server/config.d/config.xml
> <remote_servers incl="clickhouse_remote_servers" optional="true" />
> ```
>
> BTW, You don't need metrica.xml or incl for remote_servers (incl useful for ... | 2021-03-19T12:27:46 | 2021-03-19T12:27:46 | {} | NONE |
ClickHouse | ClickHouse | 802,808,205 | 21,905 | alexey-milovidov | `incl` is still supported.
We only removed it from the default configuration. | 2021-03-19T12:45:28 | 2021-03-19T12:45:28 | {} | MEMBER |
ClickHouse | ClickHouse | 802,811,219 | 21,905 | mic4ael | But this still doesn't explain why my configuration was not respected. I don't think this should be closed since for some reason I am still experiencing the problem on the versions 21.x | 2021-03-19T12:51:21 | 2021-03-19T12:51:21 | {} | NONE |
ClickHouse | ClickHouse | 802,813,299 | 21,905 | den-crane | I'll check explicit incl
```
/etc/clickhouse-server/config.d/config.xml
<remote_servers incl="clickhouse_remote_servers" optional="true" />``` | 2021-03-19T12:55:12 | 2021-03-19T12:55:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 803,058,942 | 21,893 | rlazimi-dev | Thanks for the insight.
```<bind_dn>uid=bind_user,cn=users,cn=accounts,dc=test,dc=com</bind_dn>```
I deliberately hardcoded the username here because I want a specific account to perform binding (my LDAP does not allow anonymous binding). I want other users, that are not allowed to bind, to still log in however.
... | 2021-03-19T19:24:41 | 2021-03-19T19:31:23 | {} | NONE |
ClickHouse | ClickHouse | 803,262,837 | 20,343 | filimonov | Hm i guess you right. The parser may 'eat' whitespaces | 2021-03-20T06:59:46 | 2021-03-20T06:59:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,074,829 | 19,785 | filimonov | Let's make in as a FR #21885 | 2021-03-18T16:27:31 | 2021-03-18T16:27:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,084,609 | 21,875 | iassasin | > JSONEachRow это не JSON
Ок, теперь я понял свою ошибку, сначала я подумал, что вы про форматы КХ JSON и JSONEachRow, а тут речь именно про сам формат JSON. Правда из документации этот тонкий момент совсем не очевиден :(
Теперь я понял и пример с потерей точности, это, конечно, рабочий вариант, но в моем случае пр... | 2021-03-18T16:31:17 | 2021-03-18T16:31:27 | {} | NONE |
ClickHouse | ClickHouse | 802,403,224 | 21,849 | sundy-li | @jinghancc Yes, you are right, I contributed these codes 3 years ago and I almost forgot that. Could you send a pr to improve that? | 2021-03-19T00:09:35 | 2021-03-19T00:10:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,501,141 | 21,898 | den-crane | You do understand that Clickhouse does not support updates as daily-basis operations?
Alter update / delete are designed for rare admin operations like GDPR.
Try ReplacingMergeTree / CollapsingMErgeTree / external dictionaries instead of alter. | 2021-03-19T02:50:34 | 2021-03-19T02:50:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,525,369 | 21,835 | sundy-li | I think the test 2 case is reasonable. It's using stable sort, we can accept some timestamp by default (not in strict mode), so (1, 100, 123) -> (2, 100, 123) results to level 2.
I test these 3 cases in 20.3, and it works well, I did not catch the latest options like `strict` and `strict_order`. Maybe so... | 2021-03-19T03:39:40 | 2021-03-19T03:44:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,529,679 | 21,835 | jinghancc | > I think the test 2 case is reasonable. It's using stable sort, we can accept some timestamp by default (not in strict mode), so (1, 100, 123) -> (2, 100, 123) results to level 2.
>
> I test these 3 cases in 20.3, and it works well, I did not catch the latest options like `strict` and `strict_order`. Maybe some cha... | 2021-03-19T03:49:02 | 2021-03-19T03:49:02 | {} | NONE |
ClickHouse | ClickHouse | 802,553,365 | 18,772 | Zhile | Many thanks for the info. @azat | 2021-03-19T04:47:36 | 2021-03-19T04:47:36 | {} | NONE |
ClickHouse | ClickHouse | 802,629,807 | 21,361 | cyberhuman | I think #21264 may be related. We've just upgraded to a version that contains the fix and it looks good so far 🤞🏻 | 2021-03-19T07:55:38 | 2021-03-19T07:55:38 | {} | NONE |
ClickHouse | ClickHouse | 802,785,498 | 21,886 | Avogar | > Many integration tests from test_distributed_respect_user_timeouts are flaky under tsan, can this be related?
>
> https://clickhouse-test-reports.s3.yandex.net/21886/055242fe95d8f8750355fe68d6453e0abc305957/integration_tests_(thread).html#fail1
I see the same problem on master, I think it's not related to this ... | 2021-03-19T12:03:48 | 2021-03-19T12:03:48 | {} | MEMBER |
ClickHouse | ClickHouse | 802,946,364 | 21,893 | traceon | Your configuration doesn't look correct.
First, the ldap server should be configured like this:
```xml
<ldap_servers>
<ldap_test>
<host>ldap.test.co</host>
<port>636</port>
<bind_dn>uid={user_name},cn=users,cn=accounts,dc=test,dc=com</bind_dn>
<ver... | 2021-03-19T16:12:39 | 2021-03-19T16:17:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,267,123 | 21,889 | CurtizJ | 1. Bug with extra unneeded calls of `addArena`, which slightly affected performance.
2. Bug in `FinishAggregatingInOrderAlgorithm`, when some of the inputs may stuck and aggregated data from them were not merged. Unfortunately, I was unable to create a small example for tests to reproduce this bug. | 2021-03-18T20:25:52 | 2021-03-18T20:25:52 | {} | MEMBER |
ClickHouse | ClickHouse | 802,391,245 | 21,894 | den-crane | >This is safe, right, because the Distributed table doesn't actually contain data?
Yes. It is safe. It contains only not propagated inserts. That's why I suggest to stop insertion.
Check that it stops to propagate.
Check the size of these subdirectories: `du -sh /var/lib/clickhouse/data/noc/distributedtablename/... | 2021-03-18T23:44:29 | 2021-03-18T23:44:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,408,060 | 21,894 | entropical | On one server, both of these directories are very small.
On the other server, the one with the IP address in as the name is small. The one with the domain name appears to be so huge I can't even ls or du it.
> Remove all remains of Distributed folder
Hypothetically, if there are orphaned bin files here, can I re... | 2021-03-19T00:20:21 | 2021-03-19T00:20:21 | {} | NONE |
ClickHouse | ClickHouse | 802,429,484 | 21,815 | bohutang | Can your provider the MySQL cnf and your charset? | 2021-03-19T01:01:43 | 2021-03-19T01:01:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,446,036 | 21,894 | entropical | I stopped my writers, and am now dropping the replicatedtable. It has taken about 10 minutes so far... | 2021-03-19T01:33:53 | 2021-03-19T01:33:53 | {} | NONE |
ClickHouse | ClickHouse | 802,674,261 | 18,772 | Zhile | Is this safe to merge into 20.12? I see it's reverted because of backward incapability. @azat
If it's safe I think we need to merge to 20.12 too. @alesapin
Thanks. | 2021-03-19T09:11:22 | 2021-03-19T09:11:22 | {} | NONE |
ClickHouse | ClickHouse | 802,373,537 | 21,894 | entropical | > You don't have actual Replication.
> It's documented behavior for this setup.
Yeah, I did know this, but up until very recently it has been sufficient for our purpose.
> It's about /var/lib/clickhouse/data/db/distributedtablename/.....
Right! I noticed this directory when I had the failed upgrade.
I have... | 2021-03-18T23:10:29 | 2021-03-18T23:10:29 | {} | NONE |
ClickHouse | ClickHouse | 802,594,316 | 20,179 | vmihailenco | Thanks. CH is always a step ahead :)
Ideally thread_pool_size should not be larger than number of cpu (or 2 * num_cpu), but I guess I can tweak this myself.
Closing since this seems to be fixed in 21.3 with combination of:
- s3_min_upload_part_size
- thread_pool_size | 2021-03-19T06:37:32 | 2021-03-19T06:37:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,823,724 | 20,956 | SaltTan | This seems to be fixed.
I saw different funny numbers on 20.9, but not on 21.3. | 2021-03-19T13:13:51 | 2021-03-19T13:13:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,880,955 | 21,479 | alexey-milovidov | It may require various unusual coditions to reproduce, e.g. creating and destroying large number of tables concurrently. | 2021-03-19T14:38:42 | 2021-03-19T14:38:42 | {} | MEMBER |
ClickHouse | ClickHouse | 803,336,525 | 21,937 | den-crane | Mutation always mutate all parts. But executes only renaming (using hardlinks) for parts which no need to mutate.
You can check files inodes.
hashsum unchanged:
```
┌─name─────────────────────────────────────┬─active─┬─hash_of_all_files────────────────┬─hash_of_uncompressed_files───────┬─uncompressed_hash_of_comp... | 2021-03-20T13:56:08 | 2021-03-20T13:56:37 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,387,393 | 21,846 | akuzm | ubsan fuzzer -- https://github.com/ClickHouse/ClickHouse/issues/21897 | 2021-03-18T23:36:25 | 2021-03-18T23:36:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,441,967 | 21,815 | yuuch | > Can your provider the MySQL cnf and your charset?
```
[mysqld]
datadir=/data/cyw_data/mysql_package/mysql-8.0.23/build/data/
socket=/data/cyw_data/mysql_package/mysql-8.0.23/build/mysql.sock
# Disabling symbolic-links is recommended to prevent assorted security risks
symbolic-links=0
# Settings user and group ... | 2021-03-19T01:25:49 | 2021-03-19T01:25:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,558,153 | 18,772 | Zhile | Hi, @azat
I found the fix is not in 20.12 branch. Can you help to confirm?
https://github.com/ClickHouse/ClickHouse/blob/20.12/src/Storages/MergeTree/MergeList.cpp
Does the new version of 20.12 come from 20.12 branch?
BTW, do you know how to generate a new release for 20.12?
| 2021-03-19T04:59:21 | 2021-03-19T04:59:21 | {} | NONE |
ClickHouse | ClickHouse | 802,770,065 | 21,886 | akuzm | Many integration tests from test_distributed_respect_user_timeouts are flaky under tsan, can this be related?
https://clickhouse-test-reports.s3.yandex.net/21886/055242fe95d8f8750355fe68d6453e0abc305957/integration_tests_(thread).html#fail1 | 2021-03-19T11:36:10 | 2021-03-19T11:36:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,799,716 | 19,301 | alexey-milovidov | @qoega Memory Sanitizer cannot work for dynamically loaded third-party shared libraries (that were obviously built without instrumentation).
How this issue was open for two month unnoticed? | 2021-03-19T12:29:37 | 2021-03-19T12:29:37 | {} | MEMBER |
ClickHouse | ClickHouse | 802,825,110 | 21,909 | den-crane | To be honest I did not read your issue. I don't have time to read/think.
I just replied with my snippet.
I see you have some sign column, can you mutliple ` sign * value as value` in Materialized view? | 2021-03-19T13:16:07 | 2021-03-19T13:16:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,069,585 | 21,677 | alesapin | **Currently have workloads:**
- [x] `CAS register` workload -- read/write/compare-and-swap on a single node. jepsen validates that history was linearizable. The heaviest test.
- [x] `Set` workload on a single node -- store string representation of `Set` in a single node. Trying to add new elements: read node + deseri... | 2021-03-17T13:12:42 | 2021-03-19T20:27:03 | {
"rocket": 1
} | MEMBER |
ClickHouse | ClickHouse | 803,270,731 | 21,815 | TCeason | I have test it.
First run a MySQL Container:
```
sudo docker run -d -e MYSQL_ROOT_PASSWORD=123 mysql:5.7 mysqld --datadir=/var/lib/mysql --server-id=2 --log-bin=/var/lib/mysql/mysql-bin.log --gtid-mode=ON --enforce-gtid-consistency --slave-parallel-type=LOGICAL_CLOCK --slave-parallel-workers=16 --master-info-r... | 2021-03-20T08:17:59 | 2021-03-20T08:17:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 803,282,620 | 21,722 | olgarev | @tavplubix Hi, will you review this PR, please? | 2021-03-20T09:59:26 | 2021-03-20T09:59:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 803,319,726 | 21,677 | alesapin | Wow, seems like I have one more inconsistency:
```
2021-03-20 11:42:14,911{GMT} INFO [jepsen results] jepsen.store: Wrote /home/alesap/code/cpp/ClickHouse/tests/jepsen.nukeeper/store/clickhouse-keeper quorum=true total-queue drop-data-corruptor/20210320T114129.000+0300/results.edn
2021-03-20 11:42:14,921{GMT} ... | 2021-03-20T13:31:41 | 2021-03-20T13:31:41 | {} | MEMBER |
ClickHouse | ClickHouse | 802,384,657 | 21,846 | akuzm | _Fuzzer failed (1)_ -- https://github.com/ClickHouse/ClickHouse/pull/21896 | 2021-03-18T23:30:54 | 2021-03-18T23:30:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,424,363 | 21,894 | den-crane | >Hypothetically, if there are orphaned bin files here,
>can I reintroduce them after I recreate the Distributed table, and it is working?
> I'm not too concerned if not, just working out my options.
If you sure that both directories are the same destination you can move files from the one to anther.
And yes, you... | 2021-03-19T00:53:01 | 2021-03-19T00:53:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,529,470 | 18,772 | azat | >@azat Does this also happens in 20.12? If yes, can we also cherry-pick the fix to 20.12?
@Zhile yes, it has been backported to 20.12 already (since v20.12.8.5-stable):
- 083eff544c75ee686bf3d6f133ef7e86b304458f
- 8f3228dd9192a9c824e762d2aa12674082000057 | 2021-03-19T03:48:36 | 2021-03-19T03:48:36 | {} | MEMBER |
ClickHouse | ClickHouse | 802,543,315 | 21,690 | tavplubix | Functional stateless tests (release, DatabaseReplicated) — fail: 0, passed: 2781, skipped: 105
Functional stateful tests (release, DatabaseReplicated) — fail: 0, passed: 108, skipped: 3 | 2021-03-19T04:22:46 | 2021-03-19T04:22:46 | {} | MEMBER |
ClickHouse | ClickHouse | 802,650,862 | 1,576 | akonyaev90 | Hello everyone!
Any news about the new pipeline?
Is this in the new LTS 21.3? | 2021-03-19T08:33:17 | 2021-03-19T08:33:17 | {
"eyes": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 803,095,703 | 21,935 | NikitaZakharov | @kitaisreal yes, exactly
Dictionary definition in XML config:
```
<dictionary>
<name>NAME</name>
<source>
<clickhouse>
<host>host</host>
<port>port</port>
<user>user</user>
<password>password</password>
<db>database</db>
... | 2021-03-19T20:14:09 | 2021-03-19T20:14:09 | {} | NONE |
ClickHouse | ClickHouse | 802,061,468 | 21,875 | den-crane | cast(id1 as Float64) был приведен как пример потери точности в доказательство что Decimal при парсинге как раз наоборот ничего не теряет | 2021-03-18T16:18:46 | 2021-03-18T16:18:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,192,466 | 21,851 | djmarcin | Our current workaround which seems to work is to pad any initial proto that is smaller than 128 bytes to greater than 128 bytes with a dummy field so that the initial byte of the initial length delimiter is always outside the ASCII range. | 2021-03-18T18:36:52 | 2021-03-18T18:36:52 | {} | NONE |
ClickHouse | ClickHouse | 802,362,012 | 21,894 | den-crane | You don't have actual Replication.
Distributed table propagates inserts into 2 CH nodes. Sometimes inserts can fail into one replica and succeed into another.
There is no surprise that Replicas has different data. It's documented behavior for this setup.
>Poco::Exception. Code: 1000, e.code() = 0, e.displayText(... | 2021-03-18T22:47:41 | 2021-03-18T22:49:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,530,034 | 21,329 | depressed-pho | > We already have some functions that overlaps with some implemented in this PR.
> See IPAddressDictionary.cpp and FunctionsCoding.h
>
> Could you consider reusing this functions?
Yes. Will do.
> > ClickHouse doesn't support union types. We cannot use the native IPv4 and IPv6 types for this reason.
>
> Why... | 2021-03-19T03:49:54 | 2021-03-19T03:49:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,849,261 | 21,891 | nvartolomei | Please backport at least to the latest LTS if it's possible to do with automation. | 2021-03-19T13:52:56 | 2021-03-19T13:52:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 803,067,644 | 21,888 | rf | The fuzzer failures seem to just be flakiness.. if not, if someone with more familiarity with those could point at a particular spot in the logs I should be looking at, I'd appreciate it. | 2021-03-19T19:35:32 | 2021-03-19T19:35:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 803,081,715 | 21,935 | kitaisreal | @NikitaZakharov as I understand dictionary has 1 key UInt64 and 1 attribute Float64 ? | 2021-03-19T19:55:12 | 2021-03-19T19:55:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,059,611 | 21,875 | den-crane | @iassasin
еще раз, JSONEachRow это не JSON. Забудьте на секунду про существование JSON. Чем вы там генерите текстовый файл и какие там проблемы вне скоупа Кликхауз.
JSONEachRow это особый текстовый формат, который издалека напоминает JSON.
И вот у нас есть текст "id":111111.111111111111111111, тут не может бы... | 2021-03-18T16:16:19 | 2021-03-18T16:17:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,208,168 | 21,851 | djmarcin | Some more info:
If I add an additional newline character to the query in the URL, then the data starting with 0x0a (newline) starts working, but that causes the data starting with 0x20 (space) to _stop_ working, so it seems like there's still something wonky going on with the parser.
```
$ (echo -n -e "\x0a\xa2\x0... | 2021-03-18T18:57:13 | 2021-03-18T18:57:13 | {} | NONE |
ClickHouse | ClickHouse | 802,362,581 | 21,894 | den-crane | >I did some research into the first error, but I don't really understand which directory, or which string it is referring to.
It's about `/var/lib/clickhouse/data/db/distributedtablename/.....` | 2021-03-18T22:48:48 | 2021-03-18T22:48:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,698,334 | 21,677 | alesapin | Got this once with bridge partitioner:
So one of the nodes doesn't recover in one minute. It's not an error, but something strange. | 2021-03-19T09:49:07 | 2021-03-19T09:49:07 | {} | MEMBER |
ClickHouse | ClickHouse | 803,057,082 | 20,969 | alexey-milovidov | Let's generate the addresses with `parseRemoteDescription` function.
The user should specify replicas with `|` character, example: `host{1|2|3}`.
Multiple shards cannot be specified.
(actually it can be reasonable to allow shards and create StorageDistributed on top of StorageMySQL)
This also makes sense for `p... | 2021-03-19T19:21:28 | 2021-03-19T19:21:28 | {} | MEMBER |
ClickHouse | ClickHouse | 802,307,335 | 21,677 | alesapin | > `queue` test is awesome. Found bug with snapshots apply, and something more interesting (currently not investigated):
>
> ```
> INFO [2021-03-18 23:50:16,269] jepsen test runner - jepsen.core {:perf {:latency-graph {:valid? true},
> :rate-graph {:valid? true},
> :valid? true},
> :workload {:ok... | 2021-03-18T21:18:47 | 2021-03-18T21:28:52 | {} | MEMBER |
ClickHouse | ClickHouse | 802,489,142 | 21,815 | bohutang | Ping @TCeason for help | 2021-03-19T02:28:00 | 2021-03-19T02:28:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,826,142 | 21,677 | alesapin | Now I've almost finished with tests and faced with an insurmountable obstacle: how to disable logger in zookeeper client:
```
16:13:22.295 [main-SendThread(localhost:9181)] DEBUG org.apache.zookeeper.SaslServerPrincipal - Canonicalized address to localhost ... | 2021-03-19T13:17:49 | 2021-03-19T13:18:55 | {} | MEMBER |
ClickHouse | ClickHouse | 802,944,325 | 21,677 | den-crane | try to run ZK java with -Dlog4j.configuration=/etc/..../log4j.properties
| 2021-03-19T16:09:33 | 2021-03-19T16:11:57 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,216,769 | 21,690 | tavplubix | `01396_inactive_replica_cleanup_nodes_zookeeper` - need fix test
`01149_zookeeper_mutation_stuck_after_replace_partition` - something strange happened, test is probably flaky
`01079_parallel_alter_add_drop_column_zookeeper` - Timeout - DDLWorker hung due to `All connection tries failed while connecting to ZooKeeper`
... | 2021-03-18T19:11:16 | 2021-03-18T19:11:16 | {} | MEMBER |
ClickHouse | ClickHouse | 802,810,779 | 21,886 | akuzm | OK, let's merge then. | 2021-03-19T12:50:29 | 2021-03-19T12:50:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,089,569 | 21,851 | filimonov | Duplicate of https://github.com/ClickHouse/ClickHouse/issues/20343#issuecomment-778065373
Workaround - remove SETTINGS clause, pass settings via URL, i.e.
http://clickhouse:8123/?format_schema=storage:TableSchema&query=INSERT%20INTO%20t%20FORMAT%20Protobuf
| 2021-03-18T16:33:57 | 2021-03-18T16:33:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,203,224 | 20,221 | mastertheknife | @filimonov @alexey-milovidov
I have made a test. I wrote a C++ file that throws some exception, and compiled it. One time with linking to clickhouse's libunwind, and one time with the OS one, and one time without.
It seems its an issue with the ClickHouse one. Perhaps its based on an old version with bugs.
```
[... | 2021-03-18T18:49:32 | 2021-03-18T18:49:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,466,159 | 21,849 | jinghancc | > @jinghancc Yes, you are right, I contributed these codes 3 years ago and I almost forgot that. Could you send a pr to improve that?
Hi @sundy-li , for sure, I will send a PR soon, thank you for contribution again, this function is pretty useful!! | 2021-03-19T02:10:31 | 2021-03-19T02:11:48 | {} | NONE |
ClickHouse | ClickHouse | 802,483,862 | 21,835 | jinghancc | Proposal: we could add a new mode called `strictly increasing` to avoid duplicate timestamps, so that we change the matching pattern from
```
timestamp of cond1 <= timestamp of cond2 <= ... <= timestamp of condN <= timestamp of cond1 + window.
```
to
```
timestamp of cond1 < timestamp of cond2 < ... < timestamp o... | 2021-03-19T02:24:42 | 2021-03-19T02:24:42 | {} | NONE |
ClickHouse | ClickHouse | 802,527,204 | 21,853 | azat | Tested manually, everything works after installing `clickhouse-common-static` and `bash-completion` | 2021-03-19T03:43:42 | 2021-03-19T03:43:42 | {} | MEMBER |
ClickHouse | ClickHouse | 802,647,401 | 6,286 | fuwhu | @bakey
we are also working on this, let us collaborate. | 2021-03-19T08:27:32 | 2021-03-19T08:27:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,841,513 | 21,909 | vorillaz | @den-crane I appreciate your time and efforts. I don't get your point about the sign values and aggregating the count for pageviews, can you elaborate a bit more? | 2021-03-19T13:41:30 | 2021-03-19T13:41:30 | {} | NONE |
ClickHouse | ClickHouse | 802,204,302 | 21,773 | alexey-milovidov | The fix will be shipped in the next patch release tomorrow. | 2021-03-18T18:51:16 | 2021-03-18T18:51:48 | {} | MEMBER |
ClickHouse | ClickHouse | 802,685,138 | 21,904 | godliness | ```
➜ build git:(master) cmake .. -DCMAKE_C_COMPILER=`brew --prefix llvm`/bin/clang-11 -DCMAKE_CXX_COMPILER=`brew --prefix llvm`/bin/clang++ -DCMAKE_PREFIX_PATH=`brew --prefix llvm` -DCMAKE_BUILD_TYPE=Debug
-- The C compiler identification is Clang 11.0.1
-- The CXX compiler identification is Clang 11.0.1
-- Detec... | 2021-03-19T09:28:25 | 2021-03-19T09:28:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,796,732 | 21,905 | den-crane | It was intentional https://github.com/ClickHouse/ClickHouse/pull/18740 ?
So you saying that even explicit incl does not work anymore?
```
/etc/clickhouse-server/config.d/config.xml
<remote_servers incl="clickhouse_remote_servers" optional="true" />
```
BTW, You don't need metrica.xml or incl for remote_... | 2021-03-19T12:24:28 | 2021-03-19T12:54:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,878,639 | 21,479 | sundy-li | How to reproduce?
```
CREATE TABLE test
(
`key` UInt64,
`value` UInt64,
`t` Tuple(UInt64, Float64),
`c` AggregateFunction(groupBitmap, UInt64)
)
ENGINE = EmbeddedRocksDB
PRIMARY KEY key
Query id: 53b60681-0253-4303-a9c8-3852037621b3
Ok.
0 rows in set. Elapsed: 0.017 sec.
ubuntu :) ... | 2021-03-19T14:35:23 | 2021-03-19T14:35:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 803,045,052 | 21,677 | alesapin | > Now I've almost finished with tests and faced with an insurmountable obstacle: how to disable logger in zookeeper client:
>
> ```
> 16:13:22.295 [main-SendThread(localhost:9181)] DEBUG org.apache.zookeeper.SaslServerPrincipal - Canonicalized address to localhost ... | 2021-03-19T18:57:34 | 2021-03-19T19:02:50 | {} | MEMBER |
ClickHouse | ClickHouse | 803,162,675 | 21,893 | traceon | > I deliberately hardcoded the username here because I want a specific account to perform binding (my LDAP does not allow anonymous binding). I want other users, that are not allowed to bind, to still log in however.
The config I suggested doesn't rely on anonymous binding. The Bind DN will be constructed for each u... | 2021-03-19T22:09:18 | 2021-03-19T22:09:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,136,088 | 20,221 | mastertheknife | I also seeing the same issue on CentOS 8 (with clang 10) with 21.2.5.5. it works fine on x86_64, crashes on startup on aarch64 when MySQLFactory tries to initialize SSL, and that throws an exception because it can't load the private key, and it fails to catch the exception.
Exception handling is broken here when runni... | 2021-03-18T17:14:12 | 2021-03-18T17:14:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,758,436 | 15,162 | jobosk | > @jobosk Seems that it's different case.
>
> Now conditions that is not in the form `t1.col1 = t2.col2` where `t1` and `t2` different tables is not supported, except case of `INNER` join (for such cases conditions moved to `WHERE` on query processing, see #18720). Conditions that depends on one table in `ON` and `W... | 2021-03-19T11:17:15 | 2021-03-19T11:17:15 | {
"eyes": 1
} | NONE |
ClickHouse | ClickHouse | 802,800,260 | 21,479 | alexey-milovidov | @qoega You should file a bug to RocksDB. | 2021-03-19T12:30:38 | 2021-03-19T12:30:38 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 803,352,694 | 21,909 | den-crane | ```
ReplacingMergeTree(sign)
VALUES (2, 1, 200, 1, '/page2'),
VALUES (2, -1, 200, 1, '/page2'), (2, 1, 333, 1, '/page2');
```
You have confused CollapsingMT and ReplacingMergeTree? Because ReplacingMergeTree use the version policy, not the sign.
But anyway. You a inserting values into SummingMT, if you nega... | 2021-03-20T14:05:58 | 2021-03-20T14:07:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,094,568 | 21,875 | den-crane | Делать вы можете любые тикеты, только не bug, а feature request.
Подозреваю все эти проблемы из-за супербыстрого парсинга.
Да можно выводить нормальную ошибку, но тогда надо больше cpu тратить на if else if, это деньги на лишние сервера и электричество (я не шучу). | 2021-03-18T16:37:00 | 2021-03-18T16:37:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,948,982 | 21,893 | traceon | > and suppose rlazimi-dev is a user in LDAP (and does not appear anywhere in clickhouse), that is not a bind user
> and suppose anonymous binding in my LDAP server is not allowed.
The bind will be performed on behalf of `rlazimi-dev`, and group search will be done while bound as `rlazimi-dev`. "Admin" user is not ... | 2021-03-19T16:16:34 | 2021-03-19T16:16:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,001,083 | 21,549 | alesapin | `test_replace_partition/test.py::test_replace_after_replace_failover` not related to changes. | 2021-03-10T07:11:41 | 2021-03-10T07:11:41 | {} | MEMBER |
ClickHouse | ClickHouse | 795,290,874 | 20,213 | nikitamikhaylov | Upgrade. | 2021-03-10T11:37:14 | 2021-03-10T11:37:14 | {} | MEMBER |
ClickHouse | ClickHouse | 795,677,410 | 21,520 | alexey-milovidov | Ok, there is noticeable performance improvement. | 2021-03-10T16:18:02 | 2021-03-10T16:18:02 | {} | MEMBER |
ClickHouse | ClickHouse | 796,434,721 | 21,615 | Derek-BMW | how can i solve this? | 2021-03-11T04:15:27 | 2021-03-11T04:15:27 | {} | NONE |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.