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 | 668,458,365 | 13,290 | KochetovNicolai | 01413_rows_events is broken on master | 2020-08-04T08:26:04 | 2020-08-04T08:26:04 | {} | MEMBER |
ClickHouse | ClickHouse | 668,460,242 | 13,079 | alesapin | Not important | 2020-08-04T08:29:47 | 2020-08-04T08:29:47 | {} | MEMBER |
ClickHouse | ClickHouse | 668,535,015 | 12,771 | acmnu | @traceon Could you please provide a document with exact list of supported OS and Distros?
| 2020-08-04T11:14:40 | 2020-08-04T11:14:40 | {} | NONE |
ClickHouse | ClickHouse | 668,578,452 | 12,544 | ritaank | @UnamedRus Since the user actions are not valid here (invalid syntax), it seems the behavior is not well defined.
```
clicktest :) CREATE ROW POLICY c_user ON default.test.2
Syntax error: failed at position 41:
create row policy c_user on default.test.2
Expected one of: WITH CHECK, AS, USING, Comma, ON, FO... | 2020-08-04T12:53:51 | 2020-08-04T12:53:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,629,023 | 13,321 | alesapin | Ok, now we have clear error. | 2020-08-04T14:27:00 | 2020-08-04T14:27:00 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 668,656,399 | 13,331 | SaltTan | Also the issue does not reproduce when I group by a single column, only when I group by two columns. And they don't have to be String, I found this problem when grouping by a Date and an Int32. | 2020-08-04T15:13:54 | 2020-08-04T15:13:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,733,550 | 13,191 | akuzm | SQL-fuzzy failure: https://github.com/ClickHouse/ClickHouse/issues/13335 | 2020-08-04T17:42:25 | 2020-08-04T17:42:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,809,857 | 13,143 | alexey-milovidov | > What if we add a special prefix to all the settings leaf_
Let's made it suffix: max_rows_to_read_leaf :)
> which will be renamed by the root executor before sending them to leafs and it will ignore itself
Let's don't rename anything. Just send settings as usual, apply along with `max_rows_to_read` but don't ... | 2020-08-04T20:30:31 | 2020-08-04T20:30:31 | {} | MEMBER |
ClickHouse | ClickHouse | 668,813,592 | 7,563 | alexey-milovidov | @564064202 ClickHouse does not have masters and slaves, all replicas are equal.
To create a new replica of existing table on a server, run CREATE TABLE command with the same zookeeper path and different replica path in ReplicatedMergeTree parameters. | 2020-08-04T20:38:56 | 2020-08-04T20:38:56 | {} | MEMBER |
ClickHouse | ClickHouse | 668,838,289 | 13,234 | azat | >Performance — 3 errors, 1 too long, 3 faster, 1 slower, 241 unstable
3 errors, but I see only 2:
decimal_aggregates | decimal_aggregates-err.log:clickhouse_driver.errors.ServerException: Code: 407.
-- | --
decimal_aggregates | decimal_aggregates-err.log:DB::Exception: Decimal math overflow: While executing Con... | 2020-08-04T21:38:39 | 2020-08-04T21:38:39 | {} | MEMBER |
ClickHouse | ClickHouse | 668,087,777 | 12,622 | akuzm | Fuzz test: https://github.com/ClickHouse/ClickHouse/issues/13281 | 2020-08-03T15:30:29 | 2020-08-03T15:30:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,090,315 | 12,771 | traceon | And, in case of aarch64 Linux, they are built and run in a relatively recent Ubuntu. | 2020-08-03T15:35:10 | 2020-08-03T15:35:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,091,386 | 12,771 | ilejn | > Ubuntu 8.04
> CentOS 5
> FreeBSD 11
>
> The binary of ClickHouse is generally expected to work in at least these versions. No need to build ClickHouse there, but the hardcoded config headers of third-party libs should enable/disable features in a way that will produce a binary that is runnable on at least these ... | 2020-08-03T15:37:16 | 2020-08-03T15:37:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,107,121 | 13,091 | alexey-milovidov | > sql_fuzzy
https://github.com/ClickHouse/ClickHouse/issues/12508 | 2020-08-03T16:07:29 | 2020-08-03T16:07:29 | {} | MEMBER |
ClickHouse | ClickHouse | 668,137,014 | 13,256 | den-crane | Check that you have configured access_control_path in config.xml
```
<!-- Path to folder where users and roles created by SQL commands are stored. -->
<access_control_path>/var/lib/clickhouse/access/</access_control_path>
``` | 2020-08-03T17:09:31 | 2020-08-03T17:09:31 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,137,848 | 13,256 | handgunman | Спасибо! Его не было в старом конфиге, который я использовал, потопу проглядел. | 2020-08-03T17:11:17 | 2020-08-03T17:11:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,481,491 | 12,741 | filimonov | it seems optimize_move_to_prewhere is still disabled: https://github.com/ClickHouse/ClickHouse/blob/cd372de4175ff38dbb52df54b363041aa3001275/src/Interpreters/InterpreterSelectQuery.cpp#L331
Test
```sql
create table test engine=MergeTree ORDER BY tuple() AS select intDiv(number,1000) = 0 as a, number from numb... | 2020-08-04T09:11:21 | 2020-08-04T09:23:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,628,345 | 12,771 | ilejn | @traceon I agree that we need an integration test that covers new functionality (and it will be done).
I agree that added functionality should be available on all platforms and I suggest doing this outside of this PR.
I don't quite follow you regarding other points. Could you list platforms/OSes/whatever [not covered... | 2020-08-04T14:25:47 | 2020-08-04T14:25:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,631,809 | 13,331 | alexey-milovidov | From changelog:
> Fix incompatibility of two-level aggregation between versions 20.1 and earlier. This incompatibility happens when different versions of ClickHouse are used on initiator node and remote nodes and the size of GROUP BY result is large and aggregation is performed by a single String field. It leads to ... | 2020-08-04T14:31:44 | 2020-08-04T14:31:44 | {} | MEMBER |
ClickHouse | ClickHouse | 668,732,160 | 12,508 | azat | test_version_update_after_mutation - https://clickhouse-test-reports.s3.yandex.net/13234/d28c4c4a6c82f7812562ba9ba7698e657670ee45/integration_tests_(asan).html#fail1 | 2020-08-04T17:39:29 | 2020-08-04T17:39:29 | {} | MEMBER |
ClickHouse | ClickHouse | 668,805,292 | 13,143 | nvartolomei | What if we add a special prefix to all the settings `leaf_` which will be renamed by the root executor before sending them to leafs and it will ignore itself? With this approach we will be able to send limits for both roots and leafs. | 2020-08-04T20:20:31 | 2020-08-04T20:20:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 667,450,436 | 13,182 | den-crane | @redrampage thank you.
I slightly minimized the example.
```
$ cat badFixedStringSort.sql
drop table if exists badFixedStringSort;
CREATE TABLE IF NOT EXISTS badFixedStringSort (uuid5_old FixedString(16), subitem String)
engine=MergeTree order by tuple();
system stop merges badFixedStringSort;
INSER... | 2020-08-01T01:54:53 | 2020-08-03T17:26:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,195,845 | 11,249 | nikitamikhaylov | > @nikitamikhaylov It's not a bugfix and no bug was present. Not necessary to backport (only to fulfill tests).
But it fixes the data race...
I am kinda perfectionist, want the tests and builds in LTS branch to be green except some | 2020-08-03T19:16:02 | 2020-08-03T19:16:02 | {} | MEMBER |
ClickHouse | ClickHouse | 668,426,027 | 13,203 | azat | And FWIW librdkafka has been upgraded to 1.5 (in #12991), worth rebasing against upstream/master or merging it. | 2020-08-04T07:18:28 | 2020-08-04T07:18:28 | {} | MEMBER |
ClickHouse | ClickHouse | 668,595,397 | 12,544 | vitlibar | > @UnamedRus Since the user actions are not valid here (invalid syntax), it seems the behavior is not well defined.
>
> ```
> clicktest :) CREATE ROW POLICY c_user ON default.test.2
>
> Syntax error: failed at position 41:
>
> create row policy c_user on default.test.2
>
> Expected one of: WITH CHECK, AS, ... | 2020-08-04T13:25:31 | 2020-08-04T13:25:31 | {} | MEMBER |
ClickHouse | ClickHouse | 668,608,115 | 13,311 | 4ertus2 | It's hard to backport | 2020-08-04T13:50:26 | 2020-08-04T13:50:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,757,471 | 13,203 | azat | >@azat tombstone == empty message so that is covered.
>There is no test for broken message though
Egh, sorry mix up those two, how about test for broken message then? (does not looks complex at first glance) | 2020-08-04T18:33:24 | 2020-08-04T18:33:24 | {} | MEMBER |
ClickHouse | ClickHouse | 668,815,491 | 6,063 | alexey-milovidov | I have to mention that silly automatic comments from "robot-clickhouse" and from "blinkov" is just a personal idea of one of our friends and we strongly discourage this idea. | 2020-08-04T20:43:31 | 2020-08-04T20:43:31 | {} | MEMBER |
ClickHouse | ClickHouse | 668,073,969 | 13,254 | traceon | These Testflows failures are known, and fixed in separate PR: #12736 | 2020-08-03T15:05:49 | 2020-08-03T15:05:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,543,046 | 13,321 | alesapin | Waiting for stress tests only. | 2020-08-04T11:33:53 | 2020-08-04T11:33:53 | {} | MEMBER |
ClickHouse | ClickHouse | 668,614,731 | 12,771 | traceon | @ilejn I am not saying there is anything wrong with tests or the code. Just what is the state of the codebase should be in before and after this PR is merged.
Also, the rule of thumb for any new functionality - it should be covered with at least some tests, that show that the changes actually have some positive effe... | 2020-08-04T14:01:50 | 2020-08-04T14:01:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,126,246 | 12,991 | filimonov | https://github.com/ClickHouse/ClickHouse/pull/13285 | 2020-08-03T16:47:01 | 2020-08-03T16:47:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,157,715 | 13,045 | vzakaznikov | @vitlibar, this PR is ready for your review.
| 2020-08-03T17:51:40 | 2020-08-03T17:51:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,189,762 | 11,249 | alexey-milovidov | @nikitamikhaylov It's not a bugfix and no bug was present. Not necessary to backport (only to fulfill tests). | 2020-08-03T19:02:04 | 2020-08-03T19:02:04 | {} | MEMBER |
ClickHouse | ClickHouse | 668,410,364 | 13,156 | azat | @vitlibar can you take a look please? | 2020-08-04T06:40:23 | 2020-08-04T06:40:23 | {} | MEMBER |
ClickHouse | ClickHouse | 668,644,792 | 12,771 | traceon | > ClickHouse must be runnable on:
>
> - ...
> - Ubuntu 8.04 x86_64 and above (when built on **any** x86_64 Linux and linked statically with third-parties and runtime)
> - CentOS 5 x86_64 and above (when built on **any** x86_64 Linux and linked statically with third-parties and runtime)
> - FreeBSD 11 ... | 2020-08-04T14:53:45 | 2020-08-04T14:53:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,856,692 | 13,234 | akuzm | > > Performance — 3 errors, 1 too long, 3 faster, 1 slower, 241 unstable
>
> 3 errors, but I see only 2:
>
> decimal_aggregates decimal_aggregates-err.log:clickhouse_driver.errors.ServerException: Code: 407.
> decimal_aggregates decimal_aggregates-err.log:DB::Exception: Decimal math overflow: While executing Con... | 2020-08-04T22:30:43 | 2020-08-04T22:30:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,088,874 | 12,771 | ilejn | > Have you checked that, in the generated header files (at least for Linux), the set of enabled flags is compatible with those that would be enabled if you generated/configured the libs in a typical Ubuntu and CentOS/RedHat, and the oldest Ubuntu being 14? Will the libs be enabled in macOS and FreeBSD builds? Will the ... | 2020-08-03T15:32:26 | 2020-08-03T15:32:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,219,625 | 12,003 | filimonov | It also has plenty of options / switches | 2020-08-03T20:13:51 | 2020-08-03T20:14:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,496,828 | 13,325 | KochetovNicolai | Fixed code does not present in this branch. | 2020-08-04T09:43:45 | 2020-08-04T09:43:45 | {} | MEMBER |
ClickHouse | ClickHouse | 668,582,439 | 12,550 | Akazz | After some digging in a DRAMATIC slowdown in `array_index_low_cardinality.xml` performance test has been discovered. Run time increase is so substantial, that it prevents the whole "Performance Test" from completing in due time causing TIMEOUT for its CI task. | 2020-08-04T13:02:01 | 2020-08-04T13:02:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,607,889 | 13,313 | 4ertus2 | It's hard to backport | 2020-08-04T13:50:01 | 2020-08-04T13:50:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,792,263 | 13,321 | alesapin | ` AST-based query fuzzer — Received signal 11 ` Reported and fixed. | 2020-08-04T19:50:29 | 2020-08-04T19:50:29 | {} | MEMBER |
ClickHouse | ClickHouse | 668,839,686 | 13,234 | alexey-milovidov | Two integration test failures are unrelated but has to be investigated later:
test_adaptive_granularity/test.py::test_version_update_two_nodes | FAIL
test_version_update_after_mutation/test.py::test_mutate_and_upgrade | FAIL
| 2020-08-04T21:42:41 | 2020-08-04T21:42:41 | {} | MEMBER |
ClickHouse | ClickHouse | 668,106,840 | 12,508 | alexey-milovidov | Use after free in sql_fuzzy test:
https://clickhouse-test-reports.s3.yandex.net/13091/006f9a831f481cc3f642758a6fceb2e06334b6a5/functional_stateless_tests_(thread).html#fail1 | 2020-08-03T16:06:58 | 2020-08-03T16:06:58 | {} | MEMBER |
ClickHouse | ClickHouse | 668,112,797 | 12,999 | akuzm | Superseded by #13192 | 2020-08-03T16:19:09 | 2020-08-03T16:19:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,418,863 | 13,302 | blinkov | @adevyatova there're merge conflicts | 2020-08-04T07:01:52 | 2020-08-04T07:01:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,496,965 | 13,323 | KochetovNicolai | Fixed code does not present in this branch. | 2020-08-04T09:44:02 | 2020-08-04T09:44:02 | {} | MEMBER |
ClickHouse | ClickHouse | 668,525,667 | 13,061 | vitlibar | Test failures:
integration `test_storage_rabbitmq/test.py::test_rabbitmq_many_materialized_views`- reported in https://github.com/ClickHouse/ClickHouse/issues/12508
stateless `00834_kill_mutation` - fixed in https://github.com/ClickHouse/ClickHouse/pull/13167
| 2020-08-04T10:51:13 | 2020-08-04T10:51:24 | {} | MEMBER |
ClickHouse | ClickHouse | 668,652,764 | 12,771 | traceon | Sounds good to me. | 2020-08-04T15:07:25 | 2020-08-04T15:07:25 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,653,308 | 13,331 | SaltTan | I use 20.3.13.127, I think it's recent enough.
#10952 was backported to 20.3.10
Also I don't think 20.5 was ever affected by #10952 | 2020-08-04T15:08:26 | 2020-08-04T15:22:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,689,029 | 13,302 | adevyatova | @blinkov I know, we fix it | 2020-08-04T16:13:17 | 2020-08-04T16:13:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,150,887 | 13,271 | den-crane | It's not a bug.
https://clickhouse.tech/docs/en/sql-reference/functions/date-time-functions/#timeslotsstarttime-duration-size
For a time interval starting at ‘StartTime’ and continuing for ‘Duration’ seconds, it returns an array of moments in time, consisting of points from this **interval rounded down** to the ‘Si... | 2020-08-03T17:38:43 | 2020-08-03T17:38:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,443,334 | 13,203 | oandrew | @azat tombstone == empty message so that is covered.
There is no test for broken message though | 2020-08-04T07:55:31 | 2020-08-04T07:55:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,483,113 | 12,550 | Akazz | It turned out that Performance test `TIMEOUT` was due to incorrect syntax in the newly added test queries!
As for `01413_rows_events` test failures after commit 059a3dc, i'm at my wit's end here. @myrrc Need to investigate. | 2020-08-04T09:14:46 | 2020-08-04T09:14:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,754,230 | 12,741 | azat | >It, works, but a little bit dumb:
Not a bug, feature:
https://github.com/ClickHouse/ClickHouse/blob/18ef0c0c964530866f85fec3adb8ee8ac23b0b10/src/Interpreters/InterpreterSelectQuery.cpp#L426
P.S. backported to 20.3 also works | 2020-08-04T18:26:16 | 2020-08-04T18:26:16 | {} | MEMBER |
ClickHouse | ClickHouse | 668,091,194 | 12,771 | traceon | In case of macOS, no lower limit that I know of, but a working build with those libs enabled would be nice to have too. | 2020-08-03T15:36:51 | 2020-08-03T15:36:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,187,536 | 13,234 | alexey-milovidov | e823dd9 was wrong. | 2020-08-03T18:57:03 | 2020-08-03T18:57:03 | {} | MEMBER |
ClickHouse | ClickHouse | 668,526,737 | 13,045 | vitlibar | There are conflicts, could you resolve them please? | 2020-08-04T10:54:04 | 2020-08-04T10:54:04 | {} | MEMBER |
ClickHouse | ClickHouse | 668,581,953 | 13,318 | boy-xiaozhang | > The log you've pasted is incomplete. Need full log from startup to shutdown. Also pay attention to the log messages with severity Warning or Error.
Hello, the scenario is as follows. In the case of large amounts of data, the Replicated meter will exit unexpectedly, and the restart log will only print the informati... | 2020-08-04T13:01:00 | 2020-08-04T13:01:00 | {} | NONE |
ClickHouse | ClickHouse | 668,733,253 | 13,304 | azat | Now the test can be enabled in fasttest run:
https://github.com/ClickHouse/ClickHouse/blob/5c51162b29aa4ef81d89727a489bab6c24f4be50/docker/test/fasttest/run.sh#L158-L159 | 2020-08-04T17:41:48 | 2020-08-04T17:41:48 | {} | MEMBER |
ClickHouse | ClickHouse | 668,831,773 | 13,096 | alexey-milovidov | No visible benefits. | 2020-08-04T21:21:51 | 2020-08-04T21:21:51 | {} | MEMBER |
ClickHouse | ClickHouse | 668,075,112 | 13,101 | 4ertus2 | We recomend to try the latest stable version, reproduce an issue on it and write if it does or not here.
Anycase we neew `CREATE` statements for tables and views to fix the issue. | 2020-08-03T15:08:00 | 2020-08-03T15:09:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,149,202 | 12,622 | alexey-milovidov | Stress test:
MergeTreeMutationStatus.cpp:19
\- fixed in #13169 | 2020-08-03T17:35:06 | 2020-08-03T17:35:19 | {} | MEMBER |
ClickHouse | ClickHouse | 668,355,459 | 13,320 | templarzq | storage join Table used setting：SETTINGS join_any_take_last_row = 1 | 2020-08-04T03:13:41 | 2020-08-04T03:13:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,549,484 | 12,771 | traceon | @acmnu From what I know:
ClickHouse must be buildable on:
- all oses/distros/arches/configs from CI (can be verified by looking at CI checks on PR)
ClickHouse must be runnable on:
- all os/distros/arches from CI (can be verified by looking at CI checks on PR)
- Ubuntu 8.04 x86_64 and above (when built on **any... | 2020-08-04T11:50:41 | 2020-08-04T11:58:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,719,491 | 11,755 | xmariachi | Hi @4ertus2 , was my previous comment answering your question well enough?
I'd love to collaborate if anything else can be done from our end. | 2020-08-04T17:13:33 | 2020-08-04T17:13:48 | {} | NONE |
ClickHouse | ClickHouse | 668,739,910 | 12,741 | azat | >it seems optimize_move_to_prewhere is still disabled:
No, the `row_policy_filter` is cleaned earlier.
>Test
It, works, but a little bit dumb:
```sql
/* "a = 10000" is row-level filter */
ANALYZE
SELECT *
FROM test
WHERE a = 1
┌─explain───────────────────────┐
│ SELECT │
│ ... | 2020-08-04T17:55:45 | 2020-08-04T18:03:05 | {} | MEMBER |
ClickHouse | ClickHouse | 668,864,041 | 13,343 | akuzm | @nikitamikhaylov FYI
You'll probably have to change all this when adding parallel formatting, so it's good that we have more test cases now. | 2020-08-04T22:52:03 | 2020-08-04T22:52:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,089,304 | 12,771 | traceon | Ubuntu 8.04
CentOS 5
FreeBSD 11
The binary of ClickHouse is generally expected to work in at least these versions. No need to build ClickHouse there, but the hardcoded config headers of third-party libs should enable/disable features in a way that will produce a binary that is runnable on at least these versions o... | 2020-08-03T15:33:13 | 2020-08-03T15:33:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,156,683 | 12,003 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/pull/12003#issuecomment-651203701
This result looks very promising, we can incorporate this tool with a simple exception list. | 2020-08-03T17:49:13 | 2020-08-03T17:49:13 | {} | MEMBER |
ClickHouse | ClickHouse | 668,188,382 | 13,253 | 4ertus2 | Cannot reproduce on current master + #13291
```
DROP TABLE IF EXISTS t0;
DROP TABLE IF EXISTS t1;
CREATE TABLE t0 (x UInt32, y UInt32) engine = Memory;
CREATE TABLE t1 (x UInt32, y UInt32) engine = Memory;
SELECT * FROM t0 ASOF LEFT JOIN t1
ON (t1.x = t0.x) AND greater(('\0llall\0ll\0ll\0llall\0llall\0lla... | 2020-08-03T18:59:00 | 2020-08-03T19:00:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,573,940 | 13,319 | alexey-milovidov | No, it's not implemented. TTL is applicable only for tables of MergeTree family.
But it's a good feature to implement. | 2020-08-04T12:44:04 | 2020-08-04T12:44:04 | {} | MEMBER |
ClickHouse | ClickHouse | 668,587,200 | 11,249 | nikitamikhaylov | Backported manually in 06784ed | 2020-08-04T13:11:20 | 2020-08-04T13:11:20 | {} | MEMBER |
ClickHouse | ClickHouse | 668,631,307 | 13,331 | alexey-milovidov | It can happen if you don't use the latest patch release for 20.3, 20.4 or 20.5. | 2020-08-04T14:30:54 | 2020-08-04T14:30:54 | {} | MEMBER |
ClickHouse | ClickHouse | 668,658,377 | 13,333 | alesapin | Actually, we get correct key condition `Key condition: (column 0 in [4294967280, +inf)), (column 0 in (-inf, 0]), and` And it's always false for all ranges, but for some reason `mayBeTrueAfter` https://github.com/clickhouse/ClickHouse/blob/master/src/Storages/MergeTree/KeyCondition.cpp#L1372-L1378 return true. | 2020-08-04T15:17:12 | 2020-08-04T15:17:12 | {} | MEMBER |
ClickHouse | ClickHouse | 668,737,016 | 13,326 | filimonov | Please provide a reproducible sample. | 2020-08-04T17:49:32 | 2020-08-04T17:49:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,739,285 | 13,304 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/commit/18ef0c0c964530866f85fec3adb8ee8ac23b0b10 | 2020-08-04T17:54:25 | 2020-08-04T17:54:25 | {} | MEMBER |
ClickHouse | ClickHouse | 668,839,368 | 13,234 | alexey-milovidov | The only slower test is just an outlier. | 2020-08-04T21:41:50 | 2020-08-04T21:41:50 | {} | MEMBER |
ClickHouse | ClickHouse | 668,086,280 | 12,771 | traceon | Have you checked that, in the generated header files (at least for Linux), the set of enabled flags is compatible with those that would be enabled if you generated/configured the libs in a typical Ubuntu and CentOS/RedHat? Will the libs be enabled in macOS and FreeBSD builds? Will the generated headers work for such bu... | 2020-08-03T15:27:47 | 2020-08-03T15:31:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,141,627 | 12,622 | alexey-milovidov | Regression in performance test is Ok because the logic of the function was changed. | 2020-08-03T17:19:38 | 2020-08-03T17:19:38 | {} | MEMBER |
ClickHouse | ClickHouse | 667,932,658 | 10,851 | bohutang | We have done some performance tests for this MaterializeMySQL engine.
## Test hardware
* ClickHouse, 192.168.0.4, 8vCPU*16G (cloud vm instance)
* MySQL 5.7, 192.168.0.3, 8vCPU*16G (cloud vm instance)
* benchyou, 192.168.0.2, 8vCPU*16G (cloud vm instance)
## Test method
We use [benchyou](https://github.co... | 2020-08-03T10:00:44 | 2020-08-04T09:09:25 | {
"+1": 9
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,831,494 | 13,096 | alexey-milovidov | Query Fuzzer: #13342 | 2020-08-04T21:21:09 | 2020-08-04T21:21:09 | {} | MEMBER |
ClickHouse | ClickHouse | 668,082,739 | 12,771 | filimonov | Any progress with the integration test? | 2020-08-03T15:21:42 | 2020-08-03T15:21:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,192,716 | 11,781 | alexey-milovidov | We usually have snapshot size in a range of 10..30 GiB, the worst case network between our servers is 1Gbit, if we consider it overloaded, so download rate will be 50 MB/sec, the download will take 600 seconds.
I've cut two zeros here: https://github.com/ClickHouse/ClickHouse/commit/5cc886b5ba97494a98ce5fa8d00668ce7... | 2020-08-03T19:08:33 | 2020-08-03T19:08:33 | {} | MEMBER |
ClickHouse | ClickHouse | 668,460,407 | 13,078 | alesapin | Not important | 2020-08-04T08:30:07 | 2020-08-04T08:30:07 | {} | MEMBER |
ClickHouse | ClickHouse | 668,503,646 | 13,327 | templarzq | db.T2 column id type: int32 | 2020-08-04T09:58:35 | 2020-08-04T09:58:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,572,383 | 13,318 | alexey-milovidov | The log you've pasted is incomplete. Need full log from startup to shutdown. Also pay attention to the log messages with severity Warning or Error. | 2020-08-04T12:40:47 | 2020-08-04T12:40:47 | {} | MEMBER |
ClickHouse | ClickHouse | 668,609,624 | 12,771 | ilejn | @traceon we willingly check the thing against any platform/os, while I am not sure that I understand what is wrong with CI tests that are already passed.
In https://clickhouse-test-reports.s3.yandex.net/12771/ff92812ac57d59e9c43cc90b93c0048630304ebe/compatibility_check.html one can see that centos:5 and ubuntu:12.04 m... | 2020-08-04T13:53:04 | 2020-08-04T13:53:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,416,416 | 12,195 | azat | >BTW, when the insertion destination is a materialized view,should we set max_block_size to min_insert_block_size_rows_for_materialized_views?
Yes | 2020-08-04T06:56:02 | 2020-08-04T06:56:02 | {} | MEMBER |
ClickHouse | ClickHouse | 668,596,241 | 12,550 | myrrc | Looks like the current implementation with a non-constant right argument produces a 1k-10k times slower result (with speed about 10kb/sec), so one should either fallback to the default implementation if the right argument is non-constant (thus providing a speedup with const right-hand arguments) or mutate the column an... | 2020-08-04T13:27:02 | 2020-08-04T13:27:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,651,535 | 12,771 | ilejn | @traceon , Ok, we can ignore https://clickhouse-test-reports.s3.yandex.net/12771/ff92812ac57d59e9c43cc90b93c0048630304ebe/compatibility_check.html (e.g. because we do not know what "OK" means) and ensure that the branch is
* runnable on Ubuntu 8.04 x86_64 when built on Ubuntu 20.04
* runnable on CentOS 5 x86_64 whe... | 2020-08-04T15:05:12 | 2020-08-04T15:05:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,087,907 | 12,771 | akonyaev90 | I'm not sure if we can check the assembly on all the above systems at all. | 2020-08-03T15:30:42 | 2020-08-03T15:30:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,608,008 | 13,312 | 4ertus2 | It's hard to backport | 2020-08-04T13:50:14 | 2020-08-04T13:50:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,272,568 | 13,277 | alexey-milovidov | #13303 | 2020-08-03T22:29:30 | 2020-08-03T22:29:30 | {} | MEMBER |
ClickHouse | ClickHouse | 668,459,959 | 13,071 | alesapin | Not important | 2020-08-04T08:29:13 | 2020-08-04T08:29:13 | {} | MEMBER |
ClickHouse | ClickHouse | 668,775,687 | 13,327 | den-crane | @templarzq Do you have reproducible example?
Does it work before 20.6?
Do you expect that joinGet have to return something instead of 0 in case of `left` ? | 2020-08-04T19:14:11 | 2020-08-04T19:14:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 668,878,372 | 13,203 | oandrew | Actually it's more complex than I thought.
I tried registering a schema with bad name (for #12661 ) but none of the tools allow you to do that (they all have validations).
So we would need to mock the schema registry inside the test.
I did verify it locally though.
On top of that I noticed that IRowInputForma... | 2020-08-04T23:29:34 | 2020-08-04T23:32:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,477,748 | 14,399 | den-crane | > The memory of the server is 8G
Try to add 16 GB swap space. | 2020-09-03T13:08:29 | 2020-09-03T13:08:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 686,515,070 | 14,445 | den-crane | Check how many rows in these parts. Currently TTL leaves parts empty (with 0 rows), because CH parts design (work in progress to address this).
select rows from system.parts where ... | 2020-09-03T14:04:26 | 2020-09-03T14:11:13 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.