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 | 780,725,429 | 20,588 | tavplubix | Stress test (thread) - looks like https://github.com/ClickHouse/ClickHouse/pull/20448#issuecomment-779795053 | 2021-02-17T17:36:43 | 2021-02-17T17:36:43 | {} | MEMBER |
ClickHouse | ClickHouse | 781,007,250 | 20,168 | alexey-milovidov | I found another case how `MergeTree` tables can work better than `Memory` tables.
It happens due do dynamic block size selection.
Example:
```
milovidov-desktop :) SELECT avg(blockSize()) FROM test.hits_memory WHERE NOT ignore(URL)
SELECT avg(blockSize())
FROM test.hits_memory
WHERE NOT ignore(URL)
Quer... | 2021-02-18T02:55:00 | 2021-02-18T02:55:00 | {} | MEMBER |
ClickHouse | ClickHouse | 781,092,202 | 20,864 | azat | Actually docker image was updated, but it seems that fast test check does not use per-PR docker images
```
$ docker run --rm -it yandex/clickhouse-fasttest:20864 which gdb
/usr/bin/gdb
``` | 2021-02-18T06:31:24 | 2021-02-18T06:31:24 | {} | MEMBER |
ClickHouse | ClickHouse | 780,339,549 | 19,362 | alexey-milovidov | https://clickhouse-test-reports.s3.yandex.net/20543/0c34d1c565b964353284425c0b54a98f837098bc/fuzzer_ubsan/server.log | 2021-02-17T06:36:31 | 2021-02-17T06:36:31 | {} | MEMBER |
ClickHouse | ClickHouse | 780,489,242 | 20,619 | kitaisreal | @qoega as current dictionaries maintainer I will took this one. | 2021-02-17T11:21:08 | 2021-02-17T11:21:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,495,601 | 20,609 | filimonov | Duplicate of https://github.com/ClickHouse/ClickHouse/issues/20608 | 2021-02-17T11:33:21 | 2021-02-17T11:33:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,570,165 | 20,548 | filimonov | OK, strange. And what is number of consumers - inside same clickhouse server / overall? | 2021-02-17T13:52:30 | 2021-02-17T13:52:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,879,969 | 14,156 | alexey-milovidov | We can add it to Docker containers.
There is no reason to add dependency on tzdata in regular ClickHouse packages.
tzdata is either already installed on server or user can install it manually if they need ORC.
| 2021-02-17T21:55:08 | 2021-02-17T21:55:08 | {} | MEMBER |
ClickHouse | ClickHouse | 780,880,171 | 14,156 | alexey-milovidov | @filimonov Could you please edit our Dockerfiles? | 2021-02-17T21:55:28 | 2021-02-17T21:55:28 | {} | MEMBER |
ClickHouse | ClickHouse | 780,899,685 | 20,543 | Enmk | > Looks like this patch does not solve the issue:
>
> `SELECT toDateTime64(CAST('10000000000.1' AS Decimal64(1)), 1)`
Addressed that with bc4d5a09b997a450a4c5e4c36bc69ecb4c332a49 | 2021-02-17T22:35:15 | 2021-02-17T22:35:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,909,197 | 20,139 | sonirico | :cookie: | 2021-02-17T22:56:32 | 2021-02-17T22:56:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 781,007,601 | 19,970 | etah000 | How about Lucene by JNI or Solr? Are there any technical problems if we integrate it into Clickhouse in the same way as greenplum DB does？ | 2021-02-18T02:56:01 | 2021-02-18T02:56:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,000,416 | 20,539 | myrrc | Looks like it's the `harmful` contrib that was causing the bugs, removed it from the coverage builds as we won't need it anyway. | 2021-02-16T17:36:09 | 2021-02-16T17:36:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,180,701 | 20,548 | chenziliang | `No offset stored` is usually not an error if client (`StorageKafka` in this case) doesn't consume any message from Kafka and didn't get the offset updated. I think this is probably a bug in cppkafka which shall treat this error code (RD_KAFKA_RESP_ERR__NO_OFFSET) as no error https://github.com/mfontanini/cppkafka/blob... | 2021-02-16T23:18:56 | 2021-02-16T23:19:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,239,246 | 20,555 | den-crane | but null is not equal null?
```
SELECT if(NULL = NULL, 1, 0)
┌─if(equals(NULL, NULL), 1, 0)─┐
│ 0 │
└──────────────────────────────┘
psql:
select case when null=null then 1 else 0 end;
case
------
0
(1 row)
``` | 2021-02-17T01:51:01 | 2021-02-17T01:51:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,357,553 | 19,685 | vdimir | `Functional stateless tests (release, wide parts enabled) — fail: 1`
`01396_inactive_replica_cleanup_nodes_zookeeper` looks unrelated and also have failed in masted in past week. | 2021-02-17T07:14:48 | 2021-02-17T07:15:10 | {} | MEMBER |
ClickHouse | ClickHouse | 780,467,454 | 20,612 | vdimir | Now `Nullabe(String)` not supported by function `splitByChar`. Workaround is to covert `Nullable(String)` into `String` e.g. with `COALESCE` and (maybe you also want to add `WHERE event is not NULL`) :
```
SELECT
JSONExtractString(event, 'event_time') AS s,
event,
splitByChar('T', JSONExtractString(COA... | 2021-02-17T10:42:09 | 2021-02-17T10:42:09 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 780,556,138 | 20,548 | filimonov | > No offset stored is usually not an error if client (StorageKafka in this case) doesn't consume any message from Kafka and didn't get the offset updated. I think this is probably a bug in cppkafka which shall treat this error code (RD_KAFKA_RESP_ERR__NO_OFFSET)
That's true, but normally we don't try to do commit u... | 2021-02-17T13:28:54 | 2021-02-17T13:28:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,596,602 | 20,608 | Exus | > Do you have any lxd level memory adjustments / limits?
Yes
```
host # lxc config get ch-gr-stat-2 limits.memory
3GB
```
> What is the size of ram available for clickhouse?
This container has no other services
```
container # free
total used free shared buff/cache ... | 2021-02-17T14:33:44 | 2021-02-17T14:33:44 | {} | NONE |
ClickHouse | ClickHouse | 780,598,539 | 20,612 | den-crane | splitByChar('T', assumeNotNull( ... ) ) | 2021-02-17T14:36:43 | 2021-02-17T14:36:43 | {
"+1": 36,
"-1": 2,
"heart": 2,
"hooray": 1,
"laugh": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,602,932 | 20,585 | alesapin | Seems like https://github.com/ClickHouse/ClickHouse/pull/20627 is the reason. | 2021-02-17T14:43:12 | 2021-02-17T14:43:12 | {} | MEMBER |
ClickHouse | ClickHouse | 780,614,035 | 18,506 | filimonov | cross-reference older discussion: https://github.com/ClickHouse/ClickHouse/issues/1531 | 2021-02-17T14:59:08 | 2021-02-17T14:59:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,677,205 | 20,585 | alesapin | Ok, seems like disks in CI slightly slower than on my machine. | 2021-02-17T16:25:46 | 2021-02-17T16:25:46 | {} | MEMBER |
ClickHouse | ClickHouse | 780,809,496 | 20,608 | azat | >Typo in file name ?oynchronous_metric_log.sql
So lots of memory corruptions.
Your setup looks like a test environment, that does not require lots of data to reproducer.
Maybe you can download clickhouse with ASAN, that will likely provide more details for such memory corruptions (if this is the problem in code)... | 2021-02-17T19:50:26 | 2021-02-17T19:50:26 | {} | MEMBER |
ClickHouse | ClickHouse | 780,208,581 | 19,294 | zhicwu | > it's users responsibility to name columns.
Right but the default behavior is sort of "unexpected"(compare to MySQL/PostgreSQL etc.), and it's tedious to create alias for each of the column just for removing the prefix...
> Please don't. You will break the current behavior. You cannot distinguish nested columns ... | 2021-02-17T00:33:14 | 2021-02-17T00:33:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,409,202 | 20,457 | filimonov | Since it affects IN (select ... ) - may be it also related to https://github.com/ClickHouse/ClickHouse/issues/20558 ? | 2021-02-17T09:02:25 | 2021-02-17T09:02:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,479,125 | 20,040 | RohanJ24 | > Dump how?
By using following command
`clickhouse-mysql --src-host={mysql_host} --src-user={mysql_user} --src-password={mysql_password} --dst-host={ch_host} --dst-user={ch_user} --dst-password={ch_password} --migrate-table --src-tables={db}.{table_name} --dst-schema {db} --dst-table {table_name}` | 2021-02-17T11:02:50 | 2021-02-17T11:02:50 | {} | NONE |
ClickHouse | ClickHouse | 780,611,493 | 20,608 | filimonov | > 3GB
Clickhouse defaults are not suitable for 3Gb (you may need to tune a lot of setting if you need to make it work in 3Gb)
See https://github.com/ClickHouse/ClickHouse/issues/18506#issuecomment-751267712
> lxc config get ch-gr-stat-2 limits.memory
> 3GB
> But I see that during startup clickhouse has some p... | 2021-02-17T14:55:35 | 2021-02-17T14:55:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,650,001 | 20,627 | alesapin | Fast test OK | 2021-02-17T15:47:05 | 2021-02-17T15:47:05 | {} | MEMBER |
ClickHouse | ClickHouse | 780,748,763 | 20,608 | Exus | I've raised container memory limit capped max_server_memory_usage.
```
root@ch-gr-stat-1:~# grep '<max_server_memory_usage>' /etc/clickhouse-server/config.xml
<max_server_memory_usage>14000000000</max_server_memory_usage>
root@ch-gr-stat-1:~# free
total used free shared buff... | 2021-02-17T18:14:03 | 2021-02-17T18:14:03 | {} | NONE |
ClickHouse | ClickHouse | 780,752,762 | 20,608 | Exus | > @Exus Check if setting `async_socket_for_remote` to zero helps. You should set it globally in `users.xml` in `default` profile.
```
<log_queries>1</log_queries>
+ <async_socket_for_remote>0</async_socket_for_remote>
</default>
```
Will check if it helps | 2021-02-17T18:20:23 | 2021-02-17T18:20:23 | {} | NONE |
ClickHouse | ClickHouse | 781,010,852 | 20,168 | alexey-milovidov | Experiment with compressing into arenas: https://github.com/ClickHouse/ClickHouse/compare/in-memory-compression-arenas?expand=1 | 2021-02-18T03:04:21 | 2021-02-18T03:04:21 | {} | MEMBER |
ClickHouse | ClickHouse | 780,141,866 | 20,498 | azat | >Marker check Pending — Starting checks (No special conditions apply)
Hanged? | 2021-02-16T21:54:16 | 2021-02-16T21:54:16 | {} | MEMBER |
ClickHouse | ClickHouse | 780,348,260 | 20,182 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/pull/20599 | 2021-02-17T06:54:54 | 2021-02-17T06:54:54 | {} | MEMBER |
ClickHouse | ClickHouse | 780,390,685 | 6,051 | ilyas-pro | For now we use that way:
```sql
create table issue6051_Table
(
dt Date,
col UInt32,
colAlias UInt32 DEFAULT col -- todo: change to alias after issue 6501 fix
)
engine = MergeTree() partition by toYYYYMM(dt) order by (dt);
```
So, values stored twice but Merge table works fine. | 2021-02-17T08:27:24 | 2021-02-17T08:27:24 | {} | NONE |
ClickHouse | ClickHouse | 780,494,424 | 20,608 | filimonov | Do you have any lxd level memory adjustments / limits?
What is the size of ram available for clickhouse?
What is the actual usage?
Do you pass some of Linux capabilities to clickhouse?
Do use use swap? | 2021-02-17T11:30:59 | 2021-02-17T11:30:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 779,637,777 | 20,462 | azat | 72c61da does not helps with avoiding using docker cache for "new" layers, maybe it should be reset somehow? (@alesapin ?) | 2021-02-16T07:13:33 | 2021-02-17T18:00:39 | {} | MEMBER |
ClickHouse | ClickHouse | 780,742,984 | 19,310 | tavplubix | The fix is incorrect | 2021-02-17T18:04:17 | 2021-02-17T18:04:17 | {} | MEMBER |
ClickHouse | ClickHouse | 780,827,759 | 20,608 | alexey-milovidov | > Typo in file name ?oynchronous_metric_log.sql
It looks like a hardware error either with RAM or SSD/HDD.
Note that it is possible even with ECC RAM if the amount of errors is very high.
I have no idea how it's related to LSD. | 2021-02-17T20:19:58 | 2021-02-17T20:20:25 | {} | MEMBER |
ClickHouse | ClickHouse | 781,005,985 | 18,762 | alexey-milovidov | Test failures are related somehow: https://clickhouse-test-reports.s3.yandex.net/18762/0d2d2dc2ed06ea70274a384846e2d421ea0f0b9e/functional_stateless_tests_flaky_check_(address).html#fail1 | 2021-02-18T02:51:02 | 2021-02-18T02:51:02 | {} | MEMBER |
ClickHouse | ClickHouse | 780,405,226 | 7,100 | alexey-milovidov | Fixed. | 2021-02-17T08:55:07 | 2021-02-17T08:55:07 | {} | MEMBER |
ClickHouse | ClickHouse | 780,445,463 | 19,965 | KochetovNicolai | Well, there is another one pr about reinterpret #20611
Probably, we should merge it first, and then merge docs changes here. | 2021-02-17T10:05:03 | 2021-02-17T10:05:03 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 780,480,986 | 20,619 | filimonov | Reproduces on 21.1, 21.2, 21.3.1.5944-testing | 2021-02-17T11:06:04 | 2021-02-17T11:06:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,483,393 | 20,531 | CurtizJ | Patch release for 21.2 will be at this week. | 2021-02-17T11:10:28 | 2021-02-17T11:10:28 | {} | MEMBER |
ClickHouse | ClickHouse | 780,691,974 | 20,576 | kssenii | @andreyshev, vhost support is added. | 2021-02-17T16:47:25 | 2021-02-17T16:47:25 | {} | MEMBER |
ClickHouse | ClickHouse | 780,040,202 | 14,448 | den-crane | @moeHaydar your issue is not related to the current issue. Create a new one. | 2021-02-16T18:41:41 | 2021-02-16T18:41:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,616,284 | 18,506 | filimonov | For lot of columns https://github.com/ClickHouse/ClickHouse/issues/6943
| 2021-02-17T15:02:26 | 2021-02-17T15:02:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,782,376 | 3,700 | kssenii | Fixed in https://github.com/ClickHouse/ClickHouse/pull/17301. Fixed since 21.1 version. | 2021-02-17T19:07:35 | 2021-02-17T19:08:57 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 780,801,696 | 20,462 | azat | >72c61da does not helps with avoiding using docker cache for "new" layers, maybe it should be reset somehow? (@alesapin ?)
Let's try install it as a separate layer then - c879719e7c1ed188cdd81a73148575cdbbacaf13 | 2021-02-17T19:38:12 | 2021-02-17T19:38:12 | {} | MEMBER |
ClickHouse | ClickHouse | 781,058,596 | 20,087 | ervikrant06 | My sincere apologies but still I don't see the impact of retentions on metrics which are stored on 16th Feb .. These metrics are still with 30s granulaity instead of 1m.
~~~
SELECT
Value,
Time,
Date,
Timestamp
FROM shard_01.graphite
WHERE Path LIKE 'gpfs.gpfs2graphite.skae.clients.testnode100.... | 2021-02-18T05:09:47 | 2021-02-18T05:09:47 | {} | NONE |
ClickHouse | ClickHouse | 780,377,623 | 20,548 | tobzsc | I just went through some further debugging. @filimonov gave me the advice to check it with a version lower than 21.2 as there was a change with librdkafka. So I checked several lower versions all with the same issue. But here are now some further information:
I enabled sending crashes to sentry in the config and the... | 2021-02-17T08:00:41 | 2021-02-17T08:00:41 | {} | NONE |
ClickHouse | ClickHouse | 780,379,209 | 20,596 | kssenii | I checked locally. It's just if adding a test, need to either add vhost to existing rabbitmq-docker-compose and adding non-default vhost to all tests or adding one more rabbitmq instance with a non-default vhost. I thought it's too much for a single test. | 2021-02-17T08:03:56 | 2021-02-17T08:03:56 | {} | MEMBER |
ClickHouse | ClickHouse | 780,463,718 | 20,619 | Enmk | Not reproducible on `20.11.6.6` Docker image:
```
# clickhouse client
ClickHouse client version 20.11.6.6 (official build).
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 20.11.6 revision 54442.
02a1d090ddc9 :) CREATE DATABASE ch_crash_test;
CREATE DATABASE ch_crash_test... | 2021-02-17T10:35:51 | 2021-02-17T10:35:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,474,234 | 20,531 | mateka | Hi! Is there an ETA for stable release of version 21.2 with this fix or stable release of version 21.3? | 2021-02-17T10:54:26 | 2021-02-17T10:54:26 | {} | NONE |
ClickHouse | ClickHouse | 780,507,936 | 20,531 | mateka | Great! Thanks for the info! :) | 2021-02-17T11:58:19 | 2021-02-17T11:58:19 | {} | NONE |
ClickHouse | ClickHouse | 780,510,075 | 20,497 | vdimir | Error reproduced with setting `enable_optimize_predicate_expression = 1` (enabled by default). We are going to fix it. | 2021-02-17T12:02:15 | 2021-02-17T12:02:15 | {} | MEMBER |
ClickHouse | ClickHouse | 780,784,525 | 18,420 | kssenii | Fixed in #17301. Fixed since 21.1 version. | 2021-02-17T19:09:49 | 2021-02-17T19:09:49 | {} | MEMBER |
ClickHouse | ClickHouse | 780,798,082 | 19,673 | azat | >Integration tests (thread) — fail: 1, passed: 1113, error: 0
Does not looks related
>Marker check Pending — Starting checks (No special conditions apply)
Hanged?
>Stress test (thread) — Killed by signal (in clickhouse-server.log)
Not related, but should be fixed, but it is not possible to know the real ... | 2021-02-17T19:31:55 | 2021-02-17T19:31:55 | {} | MEMBER |
ClickHouse | ClickHouse | 780,855,347 | 20,462 | azat | >Let's try install it as a separate layer then - c879719
Still no luck, moved into separate PR - #20864 | 2021-02-17T21:10:22 | 2021-02-17T21:16:33 | {} | MEMBER |
ClickHouse | ClickHouse | 780,131,581 | 20,217 | vzakaznikov | > LGTM, but maybe we should add output examples for queries to make it more clear?
Thanks! I will add output examples. | 2021-02-16T21:34:09 | 2021-02-16T21:34:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,348,795 | 20,597 | alexey-milovidov | @ucasFL Could you please take a look? | 2021-02-17T06:55:54 | 2021-02-17T06:55:54 | {} | MEMBER |
ClickHouse | ClickHouse | 780,424,653 | 20,552 | asafcombo | thanks @den-crane | 2021-02-17T09:29:42 | 2021-02-17T09:29:42 | {} | NONE |
ClickHouse | ClickHouse | 780,507,377 | 19,151 | filimonov | @alexeykulakov ? | 2021-02-17T11:57:18 | 2021-02-17T11:57:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,655,628 | 20,628 | abyss7 | No need for review - it's python utils | 2021-02-17T15:55:18 | 2021-02-17T15:55:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,439,248 | 20,106 | vitlibar | There was a problem unrelated to this PR:
https://clickhouse-test-reports.s3.yandex.net/20106/5f8a6ab9c109a82ab044b6ee573f86320175839a/integration_tests_(asan).html#fail1
```
Failed to start cluster:
Command '['docker-compose', '--project-name', 'roottestgrpcprotocolssl', '--file', '/compose/docker_compose_net.y... | 2021-02-17T09:54:40 | 2021-02-17T17:13:54 | {} | MEMBER |
ClickHouse | ClickHouse | 780,748,135 | 20,608 | alexey-milovidov | @filimonov
> Clickhouse defaults are not suitable for 3Gb
Low memory should not lead to crashes with signal 11. | 2021-02-17T18:13:00 | 2021-02-17T18:13:00 | {} | MEMBER |
ClickHouse | ClickHouse | 780,753,149 | 20,620 | alexey-milovidov | Changelog entry needed. | 2021-02-17T18:21:03 | 2021-02-17T18:21:03 | {} | MEMBER |
ClickHouse | ClickHouse | 780,777,475 | 20,336 | abyss7 | > Must backport not working?
Should work, I'm looking into it. | 2021-02-17T18:59:44 | 2021-02-17T18:59:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,924,725 | 20,608 | Exus | > Your setup looks like a test environment, that does not require lots of data to reproducer.
> Maybe you can download clickhouse with ASAN, that will likely provide more details for such memory corruptions (if this is the problem in code), but note that it will work slower (and likely should not be used in production... | 2021-02-17T23:35:52 | 2021-02-17T23:35:52 | {} | NONE |
ClickHouse | ClickHouse | 780,476,186 | 20,040 | RohanJ24 |
> It does not look like CH error.
Then what will be the data type of clickhouse which accepts only "time" data.
Example : 20:00:00 | 2021-02-17T10:57:58 | 2021-02-17T10:57:58 | {} | NONE |
ClickHouse | ClickHouse | 780,669,683 | 20,611 | alexey-milovidov | Ok.
BTW, why don't allow to reinterpret DateTime64 <-> Decimal64? | 2021-02-17T16:15:03 | 2021-02-17T16:15:03 | {} | MEMBER |
ClickHouse | ClickHouse | 780,747,868 | 20,608 | alexey-milovidov | @Exus Check if setting `async_socket_for_remote` to zero helps. You should set it globally in `users.xml` in `default` profile. | 2021-02-17T18:12:33 | 2021-02-17T18:12:33 | {} | MEMBER |
ClickHouse | ClickHouse | 780,921,634 | 20,866 | CurtizJ | If function is marked with `State` combinator, arena will be added [here](https://github.com/ClickHouse/ClickHouse/blob/80e6ad7aae080ad9d3ab3d49a67214f8d713c11a/src/Interpreters/Aggregator.cpp#L1209-L1225) to `ColumnAggregateFunction`, otherwise it will be destroyed after aggregating or merging blocks and `use-after-fr... | 2021-02-17T23:27:40 | 2021-02-17T23:27:40 | {} | MEMBER |
ClickHouse | ClickHouse | 780,042,753 | 20,552 | den-crane | No such feature.
You can use input function, but insert will be more complicated. https://clickhouse.tech/docs/en/sql-reference/table-functions/input/ | 2021-02-16T18:46:20 | 2021-02-16T18:46:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 779,665,732 | 19,294 | zhicwu | @alexey-milovidov, I tent to agree that this is inconvenient. I think it's not limited to JDBC driver as I got same results using clickhouse-client. If there's no plan to change the behavior from server-side, I'll consider a new setting in JDBC driver to optionally remove the prefix.
```sql
select * from (select 1 ... | 2021-02-16T08:17:20 | 2021-02-17T00:16:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,346,221 | 20,024 | alexey-milovidov | Does not make sense to be backported. | 2021-02-17T06:51:06 | 2021-02-17T06:51:06 | {} | MEMBER |
ClickHouse | ClickHouse | 780,413,272 | 18,510 | alexey-milovidov | @json-lrj Any help will be appreciated. | 2021-02-17T09:09:42 | 2021-02-17T09:09:42 | {} | MEMBER |
ClickHouse | ClickHouse | 780,530,305 | 20,621 | inakisoriamrf | To clarify, this only happens with some random rows, most of the information is properly copied
For example:
```
┌────m_averageScroll──┐
│ 17.26086956521739 │
│ 47 │
│ 0 │
│ 40.35897435897436 │
│ 42.57142857142857 │
│ 38.333333333333336 │
│ 43.46666666666667 │
│ 50.72222222222222 ... | 2021-02-17T12:42:02 | 2021-02-17T14:18:37 | {} | NONE |
ClickHouse | ClickHouse | 780,068,729 | 19,965 | olgarev | @alexey-milovidov Hi, this PR was approved but not merged into master. Can we close the ticket? | 2021-02-16T19:33:31 | 2021-02-16T19:33:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,339,664 | 20,543 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/issues/19362 | 2021-02-17T06:36:49 | 2021-02-17T06:36:49 | {} | MEMBER |
ClickHouse | ClickHouse | 780,846,973 | 14,156 | den-crane | hit the same issue with the latest CH 21.2
```
clickhouse-client -q "select number int, toString(number) str, 1/number flt, toDecimal64( 1/(number+1) , 9) dcml,
toDateTime('2020-01-01 00:00:00') + number time from numbers(100) format ORC"> test_sample.orc
CREATE TABLE test.orc_stream_compressed (int Int64, str... | 2021-02-17T20:54:59 | 2021-02-17T20:55:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,128,403 | 20,500 | filimonov | Is is only me who see the bad enum definition in one more place?
```
ATTACH MATERIALIZED VIEW akabylbekov.enum_alter
(
`date` Date,
`enum_type` Enum8('none' = 0, 'one' = 1, 'two' = 2),
`sales_id` UInt32,
`campaign_id` UInt32
)
ENGINE = SummingMergeTree()
PARTITION BY toYYYYMMDD(date)
ORDE... | 2021-02-16T21:27:53 | 2021-02-16T21:27:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,340,781 | 19,362 | alexey-milovidov | Changed settings:
max_threads = '1',
max_distributed_connections = '1',
totals_mode = 'after_having_auto',
group_by_two_level_threshold = '1000',
distributed_aggregation_memory_efficient = '1',
optimize_skip_unused_shards = '1',
log_queries = '1',
log_queries_min_type = 'QUERY_FINISH',
distributed_p... | 2021-02-17T06:39:14 | 2021-02-17T06:39:43 | {} | MEMBER |
ClickHouse | ClickHouse | 780,346,190 | 20,022 | alexey-milovidov | Does not make sense to be backported. | 2021-02-17T06:51:02 | 2021-02-17T06:51:02 | {} | MEMBER |
ClickHouse | ClickHouse | 780,549,878 | 20,462 | azat | Lack of backtrace in stress tests happens one more time - https://clickhouse-test-reports.s3.yandex.net/20477/18a147b3179e5931d89395d5e46b4d833d4bef05/stress_test_(memory).html#fail1 | 2021-02-17T13:18:13 | 2021-02-17T13:18:13 | {} | MEMBER |
ClickHouse | ClickHouse | 780,612,281 | 18,448 | filimonov | Any updates here?
See also:
#16918
#2261 + #2325
#18448
#17556
#20608 | 2021-02-17T14:56:41 | 2021-02-17T14:56:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,978,074 | 19,896 | depressed-pho | Rebased and resolved a conflict. | 2021-02-18T01:44:33 | 2021-02-18T01:44:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 781,011,146 | 20,306 | bharatnc | @alesapin does this look good to you ? | 2021-02-18T03:05:11 | 2021-02-18T03:05:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,124,597 | 20,548 | filimonov | Discussed in slack. Not clear why it have 'no offset stored'...
Will try to look deeper, if you will find some extra info - please let me know. | 2021-02-16T21:19:42 | 2021-02-16T21:19:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,238,131 | 19,294 | den-crane | >Right but the default behavior is sort of "unexpected"(compare to MySQL/PostgreSQL etc.)
Right, but this clickhouse-JDBC behavior is compatible with clickhouse-client / clickhouse-cpp / clickhouse-go / clickhouse-python / clickhouse-odbc. | 2021-02-17T01:47:39 | 2021-02-17T01:47:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,462,874 | 20,619 | qoega | @kitaisreal interested? | 2021-02-17T10:34:24 | 2021-02-17T10:34:24 | {} | MEMBER |
ClickHouse | ClickHouse | 780,565,622 | 20,548 | tobzsc | I consume a single topic. The issue was solved by deleting the complete topic which consisted of two partitions and recreating it again with six partitions. Really weird. | 2021-02-17T13:45:01 | 2021-02-17T13:45:01 | {} | NONE |
ClickHouse | ClickHouse | 780,756,535 | 20,599 | alexey-milovidov | Need help from @CurtizJ to fix bug in -Distinct aggregate function combinator. | 2021-02-17T18:26:26 | 2021-02-17T18:26:26 | {} | MEMBER |
ClickHouse | ClickHouse | 780,358,695 | 20,555 | alexey-milovidov | Works as expected, because NULL not equals to NULL. | 2021-02-17T07:17:10 | 2021-02-17T07:17:10 | {} | MEMBER |
ClickHouse | ClickHouse | 780,596,177 | 20,621 | filimonov | 20.5.2 was broken. :|
```
ClickHouse client version 20.5.2.7 (official build):
echo -ne '\x22\x59\x1e\x00\x00\x00\x00\x00\x00\xb1\x00\x00\x00\x00\x00\x00\x00\x22\x0a' | clickhouse-local --structure='scroll AggregateFunction(avg, UInt8)' --input-format=CSV -q 'select finalizeAggregation(scroll) from table'
43.8926... | 2021-02-17T14:33:08 | 2021-02-17T14:33:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 780,731,004 | 20,627 | alexey-milovidov | Assertion failed when using float in PREWHERE. Need to create a task... | 2021-02-17T17:46:08 | 2021-02-17T17:46:08 | {} | MEMBER |
ClickHouse | ClickHouse | 781,089,201 | 20,864 | azat | fasttest image is still not being updated, and this is not the first time, in #19751 it hadn't been update too (looks like some caching issue).
Can someone take a look? (maybe @alesapin ?) | 2021-02-18T06:24:28 | 2021-02-18T06:24:28 | {} | MEMBER |
ClickHouse | ClickHouse | 767,589,885 | 14,150 | Akazz | After internal discussion we decided to decline this implementation due to a number of reasons (most of them are mentioned in the discussions above). The task of parallelizing IO-writes does not seem to belong to this entity (MergeTreeDataPartWriterWide) | 2021-01-26T14:43:23 | 2021-01-26T14:43:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 767,635,737 | 19,644 | taiyang-li | > Do you know if there are some standards or implementations in another SQL databases?
@qoega Havn't refer to other SQL database implementations yet. Just because in clickhouse you can construct array value from expression like `[1, 2, 3, 4]` , I think it maybe reasonable to construct a map value from expression l... | 2021-01-26T15:50:15 | 2021-01-26T15:50:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 767,702,425 | 19,647 | alexey-milovidov | Yes, we can try to use AVX-512 there.
Details:
There are two components where SIMD is used:
- copy;
- shuffle.
For shuffle it is using specific "shuffle" intruction from SSSE3. There is no such instruction in AVX2.
As far as I remember, there is only shuffle separately inside two halves of the AVX register.... | 2021-01-26T17:30:06 | 2021-01-26T17:32:02 | {
"rocket": 1
} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.