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 | 580,424,231 | 7,773 | vzakaznikov | This issue should be closed as it is fixed in the latest release.
| 2020-01-30T19:49:52 | 2020-01-30T19:49:52 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,629,345 | 8,699 | orloffv | I can write a test if needed | 2020-01-31T08:07:48 | 2020-01-31T08:07:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,717,154 | 8,699 | alexey-milovidov | I'm not able to finish this task. Assigned to @4ertus2 | 2020-01-31T12:37:49 | 2020-01-31T12:37:49 | {} | MEMBER |
ClickHouse | ClickHouse | 579,631,944 | 8,553 | YiuRULE | With trying to check the commit of the version `19.17`, from the `19.17` branch, it seem it's this PR who is responsible of this bug: ClickHouse/ClickHouse#8051 (or commit 107633ff1ae50bb445f2237a7078a19f5afaf504 on the branch `19.17`)
| 2020-01-29T07:34:44 | 2020-01-29T07:34:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,806,016 | 8,889 | akuzm | It would be very helpful to reproduce it on our side. Do you have some sample data for which this reproduces consistently? You can share it with us privately through clickhouse-feedback@yandex-team.com. | 2020-01-29T15:23:48 | 2020-01-29T15:23:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,714,156 | 1,254 | alexey-milovidov | No. | 2020-01-31T12:27:40 | 2020-01-31T12:27:40 | {} | MEMBER |
ClickHouse | ClickHouse | 579,162,851 | 6,358 | alexey-milovidov | It's not too common to add shards in the middle of the list.
And if it happened, the behaviour is easy understandable and explainable. | 2020-01-28T09:47:30 | 2020-01-28T09:47:30 | {} | MEMBER |
ClickHouse | ClickHouse | 579,467,758 | 8,886 | filimonov | May relate #8573 | 2020-01-28T21:36:09 | 2020-01-29T08:51:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,657,528 | 8,573 | filimonov | Also - one of customizations seems related: https://github.com/ClickHouse-Extras/openssl/commit/a38119230ea8491c28dd80963a8dfc3a4fffe213 | 2020-01-29T08:56:36 | 2020-01-29T08:56:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,371,001 | 1,576 | TimonKK | After upgrade to 20.1.2.4 "Immortal" queries appeared. Only server restart helps | 2020-01-30T17:42:32 | 2020-01-30T17:42:32 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 580,531,826 | 8,903 | den-crane | It's not
```
create table X (A Int64, `AMap.key` Array(String), `AMap.val` Array(Int64))
Engine=SummingMergeTree order by A;
insert into X select 1 A, [toString(number%3)] keyMap, [number] valMap from numbers(10);
SELECT
A,
sumMap(AMap.key, AMap.val)
FROM X
GROUP BY A
┌─A─┬─sumMap(AMap.key, AMa... | 2020-01-31T00:54:54 | 2020-01-31T00:55:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,646,819 | 8,888 | alesapin | Maybe you can provide a table definition of `t` and sample data? | 2020-01-29T08:25:15 | 2020-01-29T08:25:15 | {} | MEMBER |
ClickHouse | ClickHouse | 579,848,197 | 8,765 | alexey-milovidov | BTW, non-trivial static function variables contain spinlock and mutex - cannot be used in signal handlers. | 2020-01-29T16:42:49 | 2020-01-29T16:42:49 | {} | MEMBER |
ClickHouse | ClickHouse | 579,854,025 | 8,857 | alexey-milovidov | Now it works like a charm. Even with data in garbage format `json.gz` it is processed with more than 500 MB/sec on old E5-2650v2 and RAID-5 of 8 HDDs. | 2020-01-29T16:55:05 | 2020-01-29T16:55:05 | {
"rocket": 1
} | MEMBER |
ClickHouse | ClickHouse | 580,506,648 | 8,903 | den-crane | It would be nice to have the same in SummingMergeTree (Nested Structures)
https://clickhouse.tech/docs/en/operations/table_engines/summingmergetree/#nested-structures | 2020-01-30T23:15:41 | 2020-01-30T23:15:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,128,036 | 8,870 | muligolan | opened by mistake | 2020-01-28T08:09:42 | 2020-01-28T08:09:42 | {} | NONE |
ClickHouse | ClickHouse | 579,137,378 | 8,573 | dstruck | > check `cat /proc/sys/kernel/random/entropy_avail `
>
> Try rng-tools or haveged at test cluster https://www.cyberciti.biz/open-source/debian-ubuntu-centos-linux-setup-additional-entropy-for-server-using-aveged-rng-tools-utils/ to check that an entropy starvation is the reason.
We have already **haveged** runnin... | 2020-01-28T08:38:06 | 2020-01-28T08:38:06 | {} | NONE |
ClickHouse | ClickHouse | 579,159,910 | 8,736 | alexey-milovidov | Yes, a comment in code will be very helpful. | 2020-01-28T09:40:03 | 2020-01-28T09:40:03 | {} | MEMBER |
ClickHouse | ClickHouse | 579,186,321 | 8,823 | KochetovNicolai | It's because I have forgotten to update progress in some cases, so the number of read rows was 0. And for 0 rows we don't print some a message in log (that was checked in test). | 2020-01-28T10:46:18 | 2020-01-28T10:46:18 | {} | MEMBER |
ClickHouse | ClickHouse | 579,224,783 | 8,573 | filimonov | Do you use any non-standard ssl-related settings?
If you have adjusted the list of ciphers (or other ssl-related settings) in `<server>` section [here](https://github.com/ClickHouse/ClickHouse/blob/e27a3170bc61981f824fd02776d6be4942e7ecce/dbms/programs/server/config.xml#L45-L57), then you should make the same adjustm... | 2020-01-28T12:35:14 | 2020-01-28T12:43:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,116,086 | 8,897 | filimonov | https://github.com/aws/aws-sdk-cpp/issues/931#issuecomment-412679977 | 2020-01-30T07:16:42 | 2020-01-30T07:17:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,521,092 | 8,903 | alexey-milovidov | It is possible that it is supported automatically. Could you please test (and add functional test if it's ok)? | 2020-01-31T00:10:03 | 2020-01-31T00:10:03 | {} | MEMBER |
ClickHouse | ClickHouse | 579,158,273 | 7,572 | alexey-milovidov | It was suspended for unknown reason, sorry. I will continue. | 2020-01-28T09:35:44 | 2020-01-28T09:35:44 | {} | MEMBER |
ClickHouse | ClickHouse | 579,163,533 | 8,864 | alexey-milovidov | They will continue to have bad exception message (because there was no create table statement). | 2020-01-28T09:49:16 | 2020-01-28T09:49:16 | {} | MEMBER |
ClickHouse | ClickHouse | 579,163,777 | 8,864 | alexey-milovidov | We can add some explanation in error message in this case too. | 2020-01-28T09:49:55 | 2020-01-28T09:50:07 | {} | MEMBER |
ClickHouse | ClickHouse | 579,647,171 | 8,884 | alesapin | Strange, why we cannot fetch merge commit? I think it's a flap. | 2020-01-29T08:26:14 | 2020-01-29T08:26:14 | {} | MEMBER |
ClickHouse | ClickHouse | 579,664,169 | 8,888 | weery28 | I create same table as select from old and requests from new works fine. | 2020-01-29T09:14:57 | 2020-01-29T09:14:57 | {} | NONE |
ClickHouse | ClickHouse | 580,190,888 | 8,889 | zombiemonkey | Will try and get a small obfuscated sample. Because of the nature of the query/environment and that it's on a 300 billion row table it might be a bit tricky to reproduce/work out. Give me a few days. | 2020-01-30T10:34:55 | 2020-01-30T10:34:55 | {} | NONE |
ClickHouse | ClickHouse | 580,301,427 | 8,903 | bgiard | Are the release integration tests hanging or crashed? I see the ASAN flavor passed a while ago. | 2020-01-30T15:16:13 | 2020-01-30T15:16:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,376,422 | 6,462 | champtar | Hello,
Upgraded a test cluster from 19.7.3.9 (with some data) to 19.14.10.16-lts, then added some more data in existing partitions.
Here is what I tried to use the TTL feature:
- `ALTER TABLE analytics.vme1_shard ON CLUSTER analytics_cluster MODIFY TTL date + INTERVAL 1 DAY;`
- set merge_with_ttl_timeout to 60 in... | 2020-01-30T17:55:40 | 2020-01-30T17:55:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,167,774 | 8,573 | dstruck | >1. do you face that issue only for Distributed queries?
Locally the query runs fine.
>2. clickhouse-client connected from the same node to secure port works properly?
When we run `clickhouse-client -s` we get the following error message
```
ClickHouse client version 20.1.2.4 (official build).
Connecting ... | 2020-01-28T10:00:18 | 2020-01-28T12:27:35 | {} | NONE |
ClickHouse | ClickHouse | 579,516,738 | 8,765 | alexey-milovidov | It is not possible to initialize singleton inside signal handler:
https://clickhouse-test-reports.s3.yandex.net/8765/df0a3ce221c794160d668a0a3a4b1e8e5c8b8d09/functional_stateless_tests_(debug)/clickhouse-server.log | 2020-01-28T23:11:57 | 2020-01-28T23:11:57 | {} | MEMBER |
ClickHouse | ClickHouse | 579,665,254 | 8,888 | jason-godden | I encountered the same problem and had to set optimize_move_to_prewhere = 0 for established queries to work. | 2020-01-29T09:17:39 | 2020-01-29T09:17:39 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 579,908,018 | 5,301 | filimonov | > ClickHouse build is pretty slow, on 40 core machine I'm able to build it in ~7m but on a local VM, which usually is more handy, it takes ~2h.
>
> Someone suggested an out of the box solution https://github.com/sakra/cotire
>
> Curious if you have experience with it or might want to look into it.
#8894 | 2020-01-29T19:00:27 | 2020-01-29T19:00:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,936,368 | 8,894 | alexey-milovidov | Build is not even started, something was stuck in CI. | 2020-01-29T20:07:19 | 2020-01-29T20:07:19 | {} | MEMBER |
ClickHouse | ClickHouse | 580,541,648 | 8,699 | alexey-milovidov | It's strange that it's passed. I remember that I faced some issues. Let's merge with master and look again. | 2020-01-31T01:33:50 | 2020-01-31T01:33:50 | {} | MEMBER |
ClickHouse | ClickHouse | 579,229,795 | 6,833 | azat | Hm, am [I missing something](https://github.com/ClickHouse/ClickHouse/blob/e27a3170bc61981f824fd02776d6be4942e7ecce/dbms/src/Common/ThreadPool.cpp#L229) or any unhandled exception from job that is scheduled in thread pool will shutdown it? (And also the exception will not be logged) | 2020-01-28T12:49:02 | 2020-01-28T12:49:33 | {} | MEMBER |
ClickHouse | ClickHouse | 579,494,769 | 8,573 | den-crane | could be related to https://github.com/ClickHouse/ClickHouse/issues/4470#issuecomment-576466122
@dstruck what Linux distro do you use?
And version of openssl ?
```
dpkg -l 'openssl'
+++-==============-================-============-=
ii openssl 1.1.1d-0+deb10u2 amd64
``` | 2020-01-28T22:31:42 | 2020-01-28T22:31:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,909,978 | 8,894 | filimonov | Also cotire readme says that
> The functionality provided by cotire has been superseded by features added to CMake 3.16. Support for pre-compiling and unity builds is now built into CMake. Thus, there will not be any further updates or support for this project.
Would be interesting to compare with option "use lates... | 2020-01-29T19:05:14 | 2020-01-29T19:05:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,153,755 | 5,301 | azat | >Yes, ~ 20% difference makes sense.
It will be interesting to see how the precompiled headers will help with modifications to headers that included in lot's of places (i.e. `Context.h`), but I guess it could not be more then 20% | 2020-01-30T09:07:14 | 2020-01-30T09:07:14 | {} | MEMBER |
ClickHouse | ClickHouse | 580,534,571 | 8,860 | alexey-milovidov | Difference in perf test: unrelated. | 2020-01-31T01:06:21 | 2020-01-31T01:06:21 | {} | MEMBER |
ClickHouse | ClickHouse | 579,732,839 | 5,301 | alexey-milovidov | Yes, ~ 20% difference makes sense.
If it's not so hard to add it, let's do. | 2020-01-29T12:25:46 | 2020-01-29T12:25:46 | {} | MEMBER |
ClickHouse | ClickHouse | 580,095,090 | 8,846 | azat | Or maybe this should be replaced with const expr evaluation for the `WITH` statement? | 2020-01-30T05:53:44 | 2020-01-30T05:53:44 | {} | MEMBER |
ClickHouse | ClickHouse | 580,542,896 | 8,835 | alexey-milovidov | It didn't work as expected, we can safely throw exception in that strange cases. | 2020-01-31T01:38:22 | 2020-01-31T01:38:22 | {} | MEMBER |
ClickHouse | ClickHouse | 580,592,086 | 8,695 | filimonov | > Such kind of requests does not need caching, cause it is 100% cache miss.
That's the idea of 'direct' layout, when every value get requested from external source w/o any caching. For now you can try to mimic that by setting cache size to some tiny value (will not work efficiently, but should work). | 2020-01-31T05:48:00 | 2020-01-31T05:48:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,159,318 | 8,812 | alexey-milovidov | @qoega Usually we don't insist on splitting. | 2020-01-28T09:38:29 | 2020-01-28T09:38:29 | {} | MEMBER |
ClickHouse | ClickHouse | 579,213,253 | 8,865 | achulkov2 | Fixed, see https://github.com/ClickHouse/ClickHouse/pull/8866. | 2020-01-28T12:02:28 | 2020-01-28T12:02:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,261,092 | 8,573 | dstruck | We use the standard settings on the server and the client:
```
<disableProtocols>sslv2,sslv3</disableProtocols>
<preferServerCiphers>true</preferServerCiphers>
```
With the docker-compose (3 zookeeper and 3 clickhouse server, debian:buster) test setup and a minimal table, we are able to replicate the issue.
... | 2020-01-28T14:05:35 | 2020-01-28T14:42:47 | {} | NONE |
ClickHouse | ClickHouse | 579,477,357 | 8,573 | filimonov | Maybe https://github.com/openssl/openssl/issues/9545 ? | 2020-01-28T21:58:42 | 2020-01-28T21:58:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,720,473 | 8,871 | muligolan | it happened few days ago and the log of the zookeeper rolled over this part.
the zookeeper servers are separated from clickhouse servers | 2020-01-29T11:49:53 | 2020-01-29T11:49:53 | {} | NONE |
ClickHouse | ClickHouse | 579,843,726 | 5,329 | filimonov | Also - when table is ordered by `(k1,k2,k3,...)` group by `k1` can use something like read_in_order and immediately send the group to the client after k1 value changed. (much more memory efficient, faster results). | 2020-01-29T16:33:30 | 2020-01-29T16:33:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,934,137 | 8,894 | kreuzerkrieg | > Switch to the latest CMake is possible (and appreciated) if we will add as simple as possible way to install it to the build instruction.
Later on I will test it with new cmake
Meanwhile, what happens to the build? looks like it is stuck | 2020-01-29T20:01:41 | 2020-01-29T20:01:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,390,365 | 6,462 | champtar | ```
# cat ./20200126_5946_5946_0/ttl.txt
ttl format version: 1
{"table":{"min":1580083200,"max":1580083200}}
```
So what we would need is a query to update the TTL of existing parts/partitions | 2020-01-30T18:30:03 | 2020-01-30T18:30:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,419,360 | 8,876 | filimonov | Superseded by #8917. | 2020-01-30T19:37:53 | 2020-01-30T19:37:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,519,194 | 8,879 | alexey-milovidov | The `quantileTiming` function is expected to take the number of milliseconds without the decimal point.
Example: 1234 - 1 second 234 milliseconds. | 2020-01-28T23:16:55 | 2020-01-28T23:16:55 | {} | MEMBER |
ClickHouse | ClickHouse | 579,665,580 | 8,573 | kung-foo | yeah, that doesn't exactly match the upstream commit https://github.com/openssl/openssl/commit/9484b67dfb0fc69326b4d94c2040751b205baa24#diff-764f070a63820217e9abd41ff93edc3bL104-R108 | 2020-01-29T09:18:26 | 2020-01-29T09:18:26 | {} | NONE |
ClickHouse | ClickHouse | 580,217,461 | 8,890 | josepowera | We got similar situation after upgrade of CH from official docker image: yandex/clickhouse-server:19.17.6.36 to image: yandex/clickhouse-server:20.1.2.4
#Connections to all replicas failed: dspconfig@192.168.1.60:3306 as user dspconfig (version 20.1.2.4 (official build)
```
2020.01.30 11:33:12.735385 [ 31 ] ... | 2020-01-30T11:51:21 | 2020-01-30T11:51:21 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 580,495,705 | 8,903 | bgiard | Thank you! That was fast! | 2020-01-30T22:39:26 | 2020-01-30T22:39:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,655,593 | 8,892 | YiuRULE | Also this issue is a bit critical as it also depends about how much data we have on a clickhouse set.
for example, this command `SELECT NULL as v1 FROM (SELECT * FROM system.numbers LIMIT 6) ORDER BY v1 DESC LIMIT 10` fail as well. | 2020-01-31T09:27:23 | 2020-01-31T09:27:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,704,809 | 8,925 | filimonov | Can you check / share the query which caused that? Most probably you will be able to find it grepping logs for `3b7f94c7-7aae-4fe6-863e-5248cfe97cc3` (query_id from stacktrace). | 2020-01-31T11:55:55 | 2020-01-31T11:55:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,705,602 | 8,924 | filimonov | I think you can find the answer just by executing both queries, don't you? :) | 2020-01-31T11:58:38 | 2020-01-31T11:58:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,157,927 | 8,842 | alexey-milovidov | @zhang2014 Thank you! Your PR has suspended for unknown reason, I will take it. | 2020-01-28T09:34:36 | 2020-01-28T09:34:36 | {} | MEMBER |
ClickHouse | ClickHouse | 579,222,833 | 8,834 | zlobober | One more attempt. | 2020-01-28T12:29:52 | 2020-01-28T12:29:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,906,075 | 7,495 | morozovsk | Заработало начиная с версии 20.1.2.4. Большое спасибо!
Пример запроса:
```sql
SELECT
toStartOfInterval(t, INTERVAL $__interval_ms/1000 second) AS time
count() as value,
level as metric
FROM error_log
WHERE t BETWEEN toDateTime($__unixEpochFrom()) and toDateTime($__unixEpochTo())
GROUP BY time, metr... | 2020-01-29T18:55:45 | 2020-01-29T18:55:45 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,318,410 | 8,824 | CurtizJ | Backport conflict in 20.1 | 2020-01-30T15:52:59 | 2020-01-30T15:52:59 | {} | MEMBER |
ClickHouse | ClickHouse | 580,394,132 | 5,486 | champtar | It would also be nice to have a way to update the current TTL of parts, for example TTL was set to 6 month, then ALTERed to 1 year, if I understand correctly right now data created before the ALTER will be deleted after 6 months, which is not what is expected IMO | 2020-01-30T18:39:19 | 2020-01-30T18:39:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,480,699 | 8,903 | alexey-milovidov | > Are the release integration tests hanging or crashed? I see the ASAN flavor passed a while ago.
Right now all tests have passed and this PR is green! | 2020-01-30T21:58:47 | 2020-01-30T21:58:47 | {} | MEMBER |
ClickHouse | ClickHouse | 580,706,776 | 1,254 | millb | As I understand this task has already been solved. | 2020-01-31T12:02:35 | 2020-01-31T12:02:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,334,677 | 8,812 | BayoNet | > Diff is not only about query profiler. Can you split PR please?
There are just markup edits and some minor updates I've made by comments to other pulls. Usually, I try not to do so.
Splitting this pull can take longer than it's reasonable to spend on it. | 2020-01-28T16:26:15 | 2020-01-28T16:26:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,656,347 | 8,573 | kung-foo | Looking at just the TLS issue in https://github.com/ClickHouse/ClickHouse/issues/8886, I can see a failed call during the handshake.
stracing the ch http handler I see:
```
09:04:36 shmget(0x72, 1, 000) = -1 ENOENT (No such file or directory)
09:04:36 uname({sysname="Linux", nodename="clickhouse-demo2",... | 2020-01-29T08:53:09 | 2020-01-29T09:06:09 | {} | NONE |
ClickHouse | ClickHouse | 579,673,359 | 8,889 | zombiemonkey | #4494 - possibly related? | 2020-01-29T09:39:33 | 2020-01-29T09:39:46 | {} | NONE |
ClickHouse | ClickHouse | 579,802,895 | 5,301 | nvartolomei | From https://github.com/sakra/cotire
> The functionality provided by cotire has been superseded by features added to CMake 3.16. Support for pre-compiling and unity builds is now built into CMake. Thus, there will not be any further updates or support for this project. | 2020-01-29T15:17:29 | 2020-01-29T15:17:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,486,865 | 8,755 | alexey-milovidov | Build was Ok at cd0e921 but now it is broken again. | 2020-01-30T22:14:25 | 2020-01-30T22:14:25 | {} | MEMBER |
ClickHouse | ClickHouse | 580,546,180 | 8,894 | alexey-milovidov | There is some incompatibility, please look at the build log: https://clickhouse-builds.s3.yandex.net/8894/f34b341427daa2fdeffabe2839a7f61cbb8394c9/build_log_599049204_1580382774.txt | 2020-01-31T01:52:03 | 2020-01-31T01:52:03 | {} | MEMBER |
ClickHouse | ClickHouse | 580,658,377 | 8,692 | LittleMaverick | @mrxotey,
There are only 2 entries in the table. But after using array join we see 4 entries in totals section.
It how array join works, yes. But i would like to see 2 entries in totals section. | 2020-01-31T09:35:29 | 2020-01-31T09:35:29 | {} | NONE |
ClickHouse | ClickHouse | 579,221,732 | 6,833 | azat | It also happened to me, and AFAIR the `GlobalThreadPool::instance().shutdown` was set to `true` by some reason, and there was nothing in logs that can lead to this (no errors before `Cannot schedule a task` and any subsequent query fails with this error after).
At that time I cannot find the culprit, once this will ... | 2020-01-28T12:26:55 | 2020-01-28T12:26:55 | {} | MEMBER |
ClickHouse | ClickHouse | 579,292,352 | 8,871 | den-crane | Do you use dedicated servers for Zookeeper nodes?
Check Zookeeper's logs in the same time period. | 2020-01-28T15:12:25 | 2020-01-28T15:12:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,648,886 | 8,888 | weery28 | > Maybe you can provide a table definition of `t` and sample data?
-- auto-generated definition
create table messages1
(
id UUID default generateUUIDv4(),
time DateTime,
ident String,
longitude Nullable(Floa... | 2020-01-29T08:31:20 | 2020-01-29T08:45:56 | {} | NONE |
ClickHouse | ClickHouse | 579,730,559 | 5,301 | kreuzerkrieg | Somehow I got back to `cotire` and on my laptop:
Clean build, clean ccache
1 h 16 m
Clean build, clean ccache, cotire
57 m 11 s
is it worth considering creating PR for it? | 2020-01-29T12:18:58 | 2020-01-29T12:18:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,790,829 | 8,897 | alexey-milovidov | @filimonov It could be related to the OpenSSL issue. | 2020-01-29T14:51:55 | 2020-01-29T14:51:55 | {} | MEMBER |
ClickHouse | ClickHouse | 579,889,174 | 7,798 | SaltTan | The same test using DDL:
```
drop dictionary if exists system.dict1;
CREATE DICTIONARY if not exists system.dict1
(
bytes_allocated UInt64,
element_count Int32,
loading_start_time DateTime
)
PRIMARY KEY bytes_allocated
SOURCE(CLICKHOUSE(HOST 'localhost' PORT 9000 USER 'default' PASSWORD '' TABLE 'dicti... | 2020-01-29T18:14:28 | 2020-01-29T18:14:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,716,963 | 8,699 | alexey-milovidov | @orloffv You have already written the test and it doesn't work for this PR. | 2020-01-31T12:37:09 | 2020-01-31T12:37:09 | {} | MEMBER |
ClickHouse | ClickHouse | 579,162,281 | 6,358 | alexey-milovidov | @filimonov Yes, I'm aware of this issue and decided that it's Ok.
If shards were added in the middle, stale data will be sent to another shards. And this is perfectly fine. | 2020-01-28T09:45:57 | 2020-01-28T09:45:57 | {} | MEMBER |
ClickHouse | ClickHouse | 579,632,811 | 8,887 | alex-zaitsev | @sibilkov , documentation is for master. TTL for storage are only available in 20.1 and later. | 2020-01-29T07:38:00 | 2020-01-29T07:38:00 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,845,777 | 8,765 | abyss7 | > It is not possible to initialize singleton inside signal handler:
>
> https://clickhouse-test-reports.s3.yandex.net/8765/df0a3ce221c794160d668a0a3a4b1e8e5c8b8d09/functional_stateless_tests_(debug)/clickhouse-server.log
It's because the `writeTraceInfo()` contents were outside of `TraceCollector` - will try to s... | 2020-01-29T16:37:58 | 2020-01-29T16:37:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,933,426 | 8,894 | alexey-milovidov | Switch to the latest CMake is possible (and appreciated) if we will add as simple as possible way to install it to the build instruction. | 2020-01-29T19:59:55 | 2020-01-29T19:59:55 | {} | MEMBER |
ClickHouse | ClickHouse | 580,110,919 | 8,159 | weiqxu | @nickevin Here is the draft
The path part of `URT` may cantain HDFS nameservice instead of host/ip and port. In this case it will read HDFS configuration values that are configured with hdfs-client.xml. By default the location of hdfs-client.xml is in working directory. Or special environment variable LIBHDFS3_CONF ... | 2020-01-30T06:59:32 | 2020-01-30T06:59:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 578,741,311 | 8,841 | filimonov | > It only works for data of single tables, not metadata.
adding metadata to a shadow dir can be also a nice idea. Can be a part of #8661
> > Why tar?
>
> Allow to manipulate backup as a single unit.
> Otherwise we will end up with huge number of files.
I'm not sure about Yandex s3 implementation, but aws s... | 2020-01-27T13:16:45 | 2020-01-30T08:21:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,204,175 | 8,859 | filimonov | Fixed in #8868 | 2020-01-30T11:10:07 | 2020-01-30T11:10:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,323,472 | 8,797 | CurtizJ | Backport conflict in 20.1 | 2020-01-30T16:03:25 | 2020-01-30T16:03:25 | {} | MEMBER |
ClickHouse | ClickHouse | 580,531,373 | 8,915 | zhang2014 | This is as expected. predicate pushdown is lazy. we always push the predicate down to the '`subquery having`' and move it to '`subquery where`' when the subquery actually execute.
this is because for the following query we cannot push down predicates to '`subquery where`':
``` sql
SELECT ... FROM (SELECT ... FRO... | 2020-01-31T00:52:50 | 2020-01-31T00:52:50 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,540,355 | 8,911 | alexey-milovidov | Let's get rid of host, port, default_database completely. | 2020-01-31T01:29:08 | 2020-01-31T01:29:08 | {} | MEMBER |
ClickHouse | ClickHouse | 580,665,748 | 2,914 | filimonov | > when table is ordered by (k1,k2,k3,...) group by k1 can use something like read_in_order and immediately send the group to the client after k1 value changed. (much more memory efficient, faster results).
Also there is one more optimization mentioned here. NOT reading primary key column file ranges if value can be ... | 2020-01-31T09:56:02 | 2020-01-31T10:06:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 580,714,388 | 1,254 | alexey-milovidov | I should have an option to format data like that:
1\thello\r\n
2\tworld\r\n
instead of
1\thello\n
2\tworld\n
| 2020-01-31T12:28:27 | 2020-01-31T12:28:27 | {} | MEMBER |
ClickHouse | ClickHouse | 579,651,790 | 8,218 | filimonov | **TLSv1.3**
* OpenSSL 1.1.1 release includes support for TLSv1.3. [[source]](https://wiki.openssl.org/index.php/TLS1.3)
(and it's a surprise for me that libressl still don't have that: https://github.com/libressl-portable/portable/issues/228 )
**SNI**
* opennssl 0.9.8f (released 11 Oct 2007) – not compiled in ... | 2020-01-29T08:39:47 | 2020-01-29T08:40:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,849,168 | 5,329 | alexey-milovidov | @filimonov This is completely unrelated task https://github.com/ClickHouse/ClickHouse/issues/2914 | 2020-01-29T16:45:04 | 2020-01-29T16:45:04 | {} | MEMBER |
ClickHouse | ClickHouse | 580,362,067 | 8,861 | CurtizJ | Backport conflict in 20.1 | 2020-01-30T17:21:52 | 2020-01-30T17:21:52 | {} | MEMBER |
ClickHouse | ClickHouse | 580,667,835 | 8,894 | kreuzerkrieg | > There is some incompatibility, please look at the build log: https://clickhouse-builds.s3.yandex.net/8894/f34b341427daa2fdeffabe2839a7f61cbb8394c9/build_log_599049204_1580382774.txt
yes, I know, yesterday when it finaly started to build I saw this problem
still checking. | 2020-01-31T10:01:49 | 2020-01-31T10:01:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 579,915,281 | 5,329 | filimonov | > @filimonov This is completely unrelated task #2914
OK. Sorry for mess. | 2020-01-29T19:17:38 | 2020-01-31T10:04:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 607,775,832 | 2,666 | chenxing-xc | Thanks Ivan, no more question from my side. @qietingfengling is not from my
team. maybe he or she still need your help.
Ivan Blinkov <notifications@github.com> 于2020年4月2日周四 上午12:46写道：
> @chenxing-xc <https://github.com/chenxing-xc>, do you have any further
> questions?
>
> —
> You are receiving this because you were ... | 2020-04-02T11:00:23 | 2020-04-02T11:00:23 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.