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 | 725,248,238 | 16,395 | macEar | Timeout options helped, thanks!
The only question left is aboiut proxy (above). | 2020-11-11T07:07:36 | 2020-11-11T07:07:36 | {} | NONE |
ClickHouse | ClickHouse | 725,253,706 | 16,849 | alesapin | You are using an outdated version. I recommend updating to the latest stable or LTS version. Alter was significantly improved since 19.11. | 2020-11-11T07:21:10 | 2020-11-11T10:50:47 | {} | MEMBER |
ClickHouse | ClickHouse | 725,370,622 | 16,853 | azat | >I would say something like this
Doesn't this the same? (But a little bit longer though)
```
CREATE MATERIALIZED VIEW data2_mv
ENGINE = AggregatingMergeTree()
ORDER BY key AS
SELECT
key,
CAST(value, 'SimpleAggregateFunction(max, Int)') AS value
FROM data1
```
As for performance:
```sql
INSE... | 2020-11-11T11:29:27 | 2020-11-11T11:29:27 | {} | MEMBER |
ClickHouse | ClickHouse | 725,409,144 | 16,868 | akuzm | Do you have some simplified query + data that allows you to consistently reproduce the problem? It would be very helpful if we could reproduce it as well.
Also please show the query that fails, and attach full text of exception -- the message itself + all subsequent stack trace lines that start with 1., 2., 3., etc. | 2020-11-11T12:58:23 | 2020-11-11T12:58:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,411,524 | 15,966 | sovietotaku | >3 zookeeper nodes is right
Yep, it is. And number of zk nodes should be: n+1, where "1" will be "eyewitness". Or just odd cont with full working nodes.
I solved all of my issues and failover works now just I need.
Last problem was just missconfig (forgot to add third server in <zookeeper> section of Click's confi... | 2020-11-11T13:03:37 | 2020-11-11T13:04:01 | {} | NONE |
ClickHouse | ClickHouse | 725,415,100 | 16,850 | akuzm | You have to also configure a Storage Policy that determines how the data is spread across the disks.
Please refer to this guide: https://altinity.com/blog/2019/11/29/amplifying-clickhouse-capacity-with-multi-volume-storage-part-2 | 2020-11-11T13:12:04 | 2020-11-11T13:12:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,573,391 | 16,869 | filimonov | After discussion in a telegram, we decided to remove docs right now, and the function itself will be removed / rewriten / replaced in future releases.
https://github.com/ClickHouse/ClickHouse/pull/16901 | 2020-11-11T18:00:51 | 2020-11-11T18:00:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,732,026 | 16,578 | fibersel | > Maybe you will be also interested in the internals of `pixz` tool.
I will keep it in mind, but now I am going to deal with main task. | 2020-11-12T00:02:13 | 2020-11-12T00:02:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,808,095 | 16,827 | akuzm | The error message is bad, we should improve it. | 2020-11-10T16:18:38 | 2020-11-10T16:18:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,849,130 | 16,825 | danlark1 | @alexey-milovidov you can merge, the tests from 435f410 are good and after that I fixed the performance test which I checked locally, it should be good | 2020-11-10T17:25:22 | 2020-11-10T17:25:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,875,805 | 15,146 | tavplubix | Fixed in #15767 | 2020-11-10T18:12:45 | 2020-11-10T18:12:45 | {} | MEMBER |
ClickHouse | ClickHouse | 725,383,601 | 16,840 | Akazz | Current [perftest failure](https://clickhouse-test-reports.s3.yandex.net/16840/8353cf78b6a8ed5ae3568abee81f7e483e143d13/performance_comparison/report.html#fail1) seems irrelevant.
Sync checks are known to fail currently. | 2020-11-11T12:00:10 | 2020-11-11T12:00:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,426,440 | 16,747 | akuzm | This is already supported:
https://clickhouse.tech/docs/en/operations/server-configuration-parameters/settings/#max_server_memory_usage_to_ram_ratio | 2020-11-11T13:35:28 | 2020-11-11T13:35:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,455,130 | 16,876 | alexey-milovidov | Tests:
`SELECT toTypeName(-toUInt64(1))`
It should be the same as:
`SELECT toTypeName(0 - toUInt64(1))`
I think that returning `Int64` (with overflow) is correct. | 2020-11-11T14:30:40 | 2020-11-11T14:30:40 | {} | MEMBER |
ClickHouse | ClickHouse | 725,510,149 | 16,853 | amosbird | test error looks weird
```
2020-11-11 13:29:07 01247_optimize_distributed_group_by_sharding_key: [ FAIL ] 2.43 sec. - return code 101
2020-11-11 13:29:07 Received exception from server (version 20.12.1):
2020-11-11 13:29:07 Code: 101. DB::Exception: Received from localhost:9000. DB::Exceptio... | 2020-11-11T16:07:02 | 2020-11-11T16:07:02 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 725,601,687 | 16,853 | azat | >That's the motivation (and also refer to the example of column transformers).
Got it, thanks!
>It's a good warning but not a persuasive example.
Yeah, this is synthetic and I posted them just to add at least some details | 2020-11-11T18:59:09 | 2020-11-11T18:59:09 | {} | MEMBER |
ClickHouse | ClickHouse | 725,694,775 | 16,578 | fibersel | > LGTM.
>
> There are some leftover comments that are related to another xz implementation... we can try it in the next PR.
I had sent mail to contributor of fast-lzma2 maintainer, but he has not responded yet.I am not sure that fast-lzma2 is compatible with classic *.xz format and it's repository does not contai... | 2020-11-11T22:23:48 | 2020-11-11T22:23:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,849,988 | 16,898 | vladimir-miller | Sorry. Github create 2 Issues - original issue is here https://github.com/ClickHouse/ClickHouse/issues/16897 | 2020-11-12T05:37:19 | 2020-11-12T05:37:19 | {} | NONE |
ClickHouse | ClickHouse | 724,599,039 | 15,713 | ghost | @amosbird
Hello, do you have a plan to improve it?
Nowadays I am looking at some variations of TopN.
ClickHouse's initiator node collects all intermediate states from other nodes and merges them. If a query limits the results by `limit 10` all result-sets except top 10 will throw away after finished merging.
... | 2020-11-10T10:03:21 | 2020-11-11T08:44:42 | {} | NONE |
ClickHouse | ClickHouse | 725,407,468 | 16,433 | nvartolomei | @akuzm I think they should be a separate concept. When I think roles I think authorization concerns (authorization to read/modify data) and not resource allocation. I think keeping them separate will make both documenting and implementing easier.
Here is a hypothetical example I have in my mind which could possible ... | 2020-11-11T12:54:39 | 2020-11-11T12:54:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,410,565 | 16,857 | akuzm | If you set `max_bytes_before_external_group_by` lower, say, 50 GB -- does it still fail? 10 GB?
My guess is that this setting is based on the size of input rows and not the amount of required memory, which might turn out to be larger. | 2020-11-11T13:01:34 | 2020-11-11T13:01:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,413,981 | 15,713 | amosbird | > I wonder how ClickHouse merges high-cardinality intermediate states. If high-cardinality intermediate states makes merging heavy, is it worse to reduce high-cardinality intermediate states in remote nodes? But reducing intermediate states in remote nodes will get loss of accurate result. But I think the trade off bet... | 2020-11-11T13:09:22 | 2020-11-11T13:09:22 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 725,472,245 | 16,854 | dependabot-preview[bot] | OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting `@dependabot ignore this major version` or `@dependabot ignore this minor version`.
If you change your mind, just re-... | 2020-11-11T15:00:30 | 2020-11-11T15:00:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,593,718 | 16,621 | filimonov | BTW - i think that libhdfs can use both gsasl and cyrus sasl, may be we should switch to cyrus everywhere?
( we use cyrus for Kafka and for openldap and there is no choice there) | 2020-11-11T18:42:22 | 2020-11-11T18:43:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,686,248 | 10,767 | jakesylvestre | In case it's helpful to anyone coming across this issue, in our case we were inserting dates instead of unix timestamps | 2020-11-11T22:03:18 | 2020-11-11T22:03:18 | {} | NONE |
ClickHouse | ClickHouse | 725,712,066 | 16,578 | alexey-milovidov | You can remove commented code before we'll merge. | 2020-11-11T23:09:37 | 2020-11-11T23:09:37 | {} | MEMBER |
ClickHouse | ClickHouse | 725,869,453 | 16,907 | hexiaoting | which version do you use?
It runs ok on 20.12.1.1 | 2020-11-12T06:31:00 | 2020-11-12T06:31:00 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,411,955 | 16,851 | akuzm | What version of ClickHouse do you use? Do you have some example queries that can reproduce the problem? | 2020-11-11T13:04:32 | 2020-11-11T13:04:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,639,102 | 16,585 | rogeriochaves | it worked perfectly thank you very much! BTW this gist also helped me seeing the tables sizes: https://gist.github.com/sanchezzzhak/511fd140e8809857f8f1d84ddb937015 | 2020-11-11T20:16:35 | 2020-11-11T20:16:35 | {} | NONE |
ClickHouse | ClickHouse | 725,730,658 | 16,578 | alexey-milovidov | I've tested decompression performance, it's the same as for `xz` tool. | 2020-11-11T23:57:59 | 2020-11-11T23:57:59 | {} | MEMBER |
ClickHouse | ClickHouse | 725,229,410 | 10,223 | l1t1 | the pdf at 2020-11-11 is still has the problem, page 23 shows
| 2020-11-11T06:28:40 | 2020-11-11T06:28:40 | {} | NONE |
ClickHouse | ClickHouse | 725,388,082 | 16,853 | amosbird | > Doesn't this the same? (But a little bit longer though)
That's the motivation (and also refer to the example of column transformers).
> As for performance:
It's a good warning but not a persuasive example. `-SimpleState` is mainly for defining tables and materialized views (and projections in the future). It... | 2020-11-11T12:10:06 | 2020-11-11T12:10:23 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 725,457,328 | 16,876 | alexey-milovidov | I expect that promotion to Float64 was a bug in some old version but now the behaviour is correct.
The principle:
- arithmetic operations on integers promotes to the larger type, up to 64bit types, but not larger;
- promotion to floating point types never performed if arguments are integers;
- result of subtracti... | 2020-11-11T14:34:44 | 2020-11-11T14:36:19 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 725,615,506 | 16,902 | den-crane | Could not reproduce.
And such Dictionary is impossible in CH 20.11
```sql
create table src ( id Nullable(UInt16), date Date, Base Decimal64(9) )
Engine=MergeTree order by tuple();
insert into src select number, today(), 1.0 from numbers(1000);
insert into src select null, today(), 1.1 from numbers(1);
... | 2020-11-11T19:27:14 | 2020-11-11T19:27:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,711,907 | 16,578 | alexey-milovidov | Yes, we can merge before trying fast-lzma2. | 2020-11-11T23:09:12 | 2020-11-11T23:09:12 | {} | MEMBER |
ClickHouse | ClickHouse | 725,348,044 | 16,853 | amosbird | > Something like this only?
I would say something like this
```
create materialized view data_mv engine=AggregatingMergeTree() order by key as select key, maxSimpleState(value) value from data group by key;
``` | 2020-11-11T10:41:39 | 2020-11-11T10:41:39 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 725,429,948 | 16,857 | filimonov | See https://github.com/ClickHouse/ClickHouse/issues/10461#issuecomment-623006030 | 2020-11-11T13:42:35 | 2020-11-11T13:42:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,450,152 | 16,674 | alexey-milovidov | @akuzm It should be similar to the issue with TSVWithNames, CSVWithNames.
Note that Parquet I/O formats work fine, the only issue is with File engine. | 2020-11-11T14:21:21 | 2020-11-11T14:21:21 | {} | MEMBER |
ClickHouse | ClickHouse | 725,531,193 | 13,252 | CurtizJ | Correct version #16895. | 2020-11-11T16:44:14 | 2020-11-11T16:44:14 | {} | MEMBER |
ClickHouse | ClickHouse | 725,615,592 | 16,853 | azat | >test error looks weird
Indeed:
```
2020.11.11 13:29:06.062314 [ 5605 ] {} <Debug> TCPHandler: Processed in 0.013796241 sec.
...
2020.11.11 13:29:06.465597 [ 5605 ] {} <Error> TCPHandler: Code: 101, e.displayText() = DB::Exception: Unexpected packet Data received from client, Stack trace:
```
Looks like pr... | 2020-11-11T19:27:23 | 2020-11-11T19:27:23 | {} | MEMBER |
ClickHouse | ClickHouse | 725,881,649 | 16,578 | akuzm | This PR broke the Fast Test.
https://github.com/ClickHouse/ClickHouse/pull/16908
I'll merge the fix for now, so that we have at least some tests, but it's not entirely correct. CMake should be able to finish when the `xz` submodule is not cloned -- it's not an essential one. | 2020-11-12T06:59:39 | 2020-11-12T06:59:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,879,338 | 15,182 | tavplubix | 16 is for `MYSQL_TYPE_BIT`, currently it's not supported in ClickHouse | 2020-11-10T18:19:34 | 2020-11-10T18:19:34 | {} | MEMBER |
ClickHouse | ClickHouse | 724,952,459 | 16,844 | abyss7 | Maybe it will be easier to just declare functionals for nth element, partial sort, etc. under defines and use them everywhere? | 2020-11-10T20:37:05 | 2020-11-10T20:37:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,327,270 | 16,853 | amosbird | @filimonov You can achieve that via column transformers
```
CREATE MATERIALIZED VIEW xxx_mv TO xxx AS SELECT x,y, COLUMNS(<your favorite regexp>) APPLY (xxxState) FROM yyy;
```
It's in question whether I should extend it to `ORDER BY` for storage definition. | 2020-11-11T10:00:00 | 2020-11-11T10:01:18 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 725,412,434 | 16,848 | akuzm | Do you have some sample data + queries that we can use to reproduce the problem? | 2020-11-11T13:05:39 | 2020-11-11T13:05:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,528,710 | 16,895 | CurtizJ | This PR contains also #16875 and should be merged after it. | 2020-11-11T16:39:49 | 2020-11-11T16:39:49 | {} | MEMBER |
ClickHouse | ClickHouse | 725,524,821 | 16,867 | nvartolomei | Regarding second problem you need to remove `database_pull` from copier config so it defaults to `default_database` as specified in remote_servers config.
Later edit: Or set it to empty: `<database_pull></database_pull>`, don't recall exactly. | 2020-11-11T16:32:56 | 2020-11-11T17:25:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,827,809 | 3,950 | filimonov | Implemented in #15601, available since 20.10.
Can be enabled with `insert_quorum_parallel=1`.
The feature in its current form has a drawback - it can't work together with the sequential consistency model.
That also means that replicas can see the parts even if a quorum were not satisfied for them (insert is... | 2020-11-10T16:50:22 | 2020-11-10T16:50:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,841,194 | 16,839 | den-crane | All in this issue are UB.
And AggregatingMergeTree is not expected to work with such design.
And it will never be fixed.
1.
"index" SimpleAggregateFunction(any, Int32), "date" SimpleAggregateFunction(max, Int32)
PARTITION BY toYYYYMM(toDateTime("date")) ORDER BY ("index")
2. where -> prewhere works befor... | 2020-11-10T17:12:13 | 2020-11-10T19:20:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,987,863 | 15,718 | azat | >Looks like not a bug but a consequence of using too many separate mmaps in debug build.
@alexey-milovidov Indeed, this may be related.
>It's also indicates the deficiency in jemalloc: it does not correctly handle mmap errors.
However, at first glance return value of munmap/mmap are checked everywhere.
Althou... | 2020-11-10T21:50:20 | 2020-11-10T21:50:20 | {} | MEMBER |
ClickHouse | ClickHouse | 725,227,785 | 16,827 | l1t1 | select count(*)worked at version 20.6, but 20.10 does not work
https://github.com/ClickHouse/ClickHouse/issues/12482 | 2020-11-11T06:24:15 | 2020-11-11T06:24:15 | {} | NONE |
ClickHouse | ClickHouse | 725,457,719 | 16,836 | abyss7 | All build problems seem unrelated | 2020-11-11T14:35:33 | 2020-11-11T14:35:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,472,218 | 16,855 | dependabot-preview[bot] | OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting `@dependabot ignore this major version` or `@dependabot ignore this minor version`.
If you change your mind, just re-... | 2020-11-11T15:00:28 | 2020-11-11T15:00:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,731,203 | 16,578 | alexey-milovidov | Maybe you will be also interested in the internals of `pixz` tool. | 2020-11-11T23:59:53 | 2020-11-11T23:59:53 | {} | MEMBER |
ClickHouse | ClickHouse | 725,179,485 | 5,485 | kiwimg | when the window function will be supported in Clickhouse?
I hope to see the function early! | 2020-11-11T04:11:38 | 2020-11-11T04:11:38 | {} | NONE |
ClickHouse | ClickHouse | 725,308,056 | 16,395 | Jokser | @macEar
Glad to hear it :)
About tunnel mode you can read here https://en.wikipedia.org/wiki/HTTP_tunnel
This is how a client usually works with a secured endpoint through HTTP.
I can make a simple change that turns on tunnel mode if proxy is HTTP while endpoint is HTTPS, but this change is a bit hard to test bec... | 2020-11-11T09:22:13 | 2020-11-11T09:22:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,434,938 | 16,807 | filimonov | > so in short it means if im stoping cilckhouse-server before rsync /data /metadata i should be fine later just restore this folders with correct rights, correct?
If you don't use zookeeper / replicated tables - yes.
> how i can find is there any replicated tables? this is legacy db not created be me and its firs... | 2020-11-11T13:52:14 | 2020-11-11T13:52:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,451,298 | 16,747 | alexey-milovidov | @akuzm We have per query `max_memory_usage` that is 10 GB by default regardless to the server memory amount. | 2020-11-11T14:23:31 | 2020-11-11T14:23:31 | {} | MEMBER |
ClickHouse | ClickHouse | 725,004,135 | 15,419 | myrrc | 01459_manual_write_to_replicas not related to the PR. | 2020-11-10T22:24:38 | 2020-11-10T22:24:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,089,728 | 15,806 | hexiaoting | @CurtizJ ready for review. | 2020-11-11T02:33:10 | 2020-11-11T02:33:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,324,882 | 16,853 | filimonov | I also prefer to do
```
CREATE TABLE xxx Engine=AggregatingMergeTree() ORDER BY ... AS AS SELECT x,y, xxxState(...) FROM yyy WHERE 0=1;
CREATE MATERIALIZED VIEW xxx_mv TO xxx AS SELECT x,y, xxxState(...) FROM yyy;
```
to avoid typing definitions for every columns. And i was missing that functionality. | 2020-11-11T09:55:08 | 2020-11-11T09:55:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,447,398 | 16,674 | akuzm | Looks like it is completely broken...
```
create table p(a int) engine File(Parquet);
insert into p select * from numbers(10);
insert into p select * from numbers(11);
select * from p;
Received exception from server (version 20.12.1):
Code: 33. DB::Exception: Received from localhost:9000. DB::Exception: E... | 2020-11-11T14:16:01 | 2020-11-11T14:16:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,472,871 | 16,846 | alexey-milovidov | > Performance — 3 errors, 1 too long, 2 faster, 85 unstable
Same as in master. | 2020-11-11T15:01:40 | 2020-11-11T15:01:40 | {} | MEMBER |
ClickHouse | ClickHouse | 725,528,966 | 10,471 | dzeban | Still reproduced on the latest stable version 20.10.3.30, so fix in #15538 doesn't help here. | 2020-11-11T16:40:14 | 2020-11-11T16:40:14 | {} | NONE |
ClickHouse | ClickHouse | 725,415,944 | 16,867 | nikitamikhaylov | Will try to reproduce this behaviour in our integration tests. | 2020-11-11T13:13:39 | 2020-11-11T13:13:39 | {} | MEMBER |
ClickHouse | ClickHouse | 725,483,446 | 16,869 | filimonov | Since it works on single block - maybe we should hide it for now using the same setting as proposed here: https://github.com/ClickHouse/ClickHouse/issues/16484 ? | 2020-11-11T15:21:15 | 2020-11-11T15:26:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,798,060 | 16,338 | ucasfl | cmake failed in submodule. | 2020-11-12T03:01:44 | 2020-11-12T03:01:44 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 724,789,797 | 16,832 | akuzm | Machine translation:
Hello there:
I have a batch of offline data servers, each with 24 disks. I want to use them to build a clickhouse cluster. I am considering a storage strategy. There are several questions to ask you:
1. Is it possible to avoid raid0-50, and plan the number of shards and replicas in the click... | 2020-11-10T15:50:12 | 2020-11-10T15:50:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,926,151 | 15,856 | VitalyVPinchuk | Hello @vladimir-golovchenko
Could you send me a message via email, please?
| 2020-11-10T19:45:50 | 2020-11-10T19:45:50 | {} | NONE |
ClickHouse | ClickHouse | 725,205,942 | 16,827 | l1t1 | some links about this question
https://github.com/ClickHouse/ClickHouse/issues/873
>Enabled in master with some restrictions:
you need to list all selected columns. Asterisks (* / t.*) do not work
prewhere does not work
complex aliases in JOIN ON section do not work
so i add column name in count, it works
```
... | 2020-11-11T05:22:10 | 2020-11-11T05:31:08 | {} | NONE |
ClickHouse | ClickHouse | 725,339,750 | 16,853 | azat | > It's useful for defining MaterializedView of AggregatingMergeTree engine, and will benefit projections too.
@amosbird what is not possible to do right now that will be allowed after this patch? Right now the type for the materialized view will use original type:
```sql
create table data1 (key Int, value Simple... | 2020-11-11T10:25:07 | 2020-11-11T10:25:07 | {} | MEMBER |
ClickHouse | ClickHouse | 725,427,357 | 16,870 | filimonov | check also the list of [prio-major](https://github.com/ClickHouse/ClickHouse/issues?q=is%3Aopen+is%3Aissue+label%3Aprio-major) issues. | 2020-11-11T13:37:26 | 2020-11-11T13:37:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,447,289 | 16,837 | alesapin | Hmm, what's wrong with `test_distributed_ddl` ? | 2020-11-11T14:15:48 | 2020-11-11T14:15:48 | {} | MEMBER |
ClickHouse | ClickHouse | 724,798,051 | 16,825 | danlark1 | One query became 5-10% slower, 15-20 became significantly faster, up to 20x
```
SELECT MobilePhoneModel, MobilePhoneModel FROM hits_100m_single ORDER BY MobilePhoneModel, MobilePhoneModel LIMIT 10 FORMAT Null
```
It is expected as Floyd-Rivest is not performing way faster when there are many equal elements in t... | 2020-11-10T16:02:59 | 2020-11-10T16:30:53 | {
"rocket": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,947,233 | 16,627 | alesapin | Test failures are not related to changes. Need to fix integration tests... | 2020-11-10T20:26:41 | 2020-11-10T20:26:41 | {} | MEMBER |
ClickHouse | ClickHouse | 725,238,281 | 15,621 | avzhurav | Hi, @alexey-milovidov provided some information about an internal connection usage (with own access rights) between distributed and replicated tables in the last webinar. This feature is not documented yet. Does it possible to get more details how this internal connection correlates with the current active user, pleas... | 2020-11-11T06:50:33 | 2020-11-11T06:50:33 | {} | NONE |
ClickHouse | ClickHouse | 725,437,511 | 16,856 | excitoon | In `AssumeRole` mode S3 client tries to create HTTP client by his own means which leads to crash because in some code I expect HTTP client to be an instance of specific subclass:
```
2020.11.11 15:13:40.626835 [ 49 ] {} <Fatal> BaseDaemon: 6. /build/obj-x86_64-linux-gnu/../src/IO/S3/PocoHTTPClient.cpp:83: DB::S3::P... | 2020-11-11T13:57:15 | 2020-11-11T13:57:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,680,709 | 16,787 | kitaisreal | @alexey-milovidov should we make this default replacement for our precise formatting ? Without introducing conditional cmake variable ```USE_FAST_FLOAT``` ? | 2020-11-11T21:50:44 | 2020-11-11T21:50:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,908,432 | 15,718 | alexey-milovidov | It looks similar with the assertion found here: #16836
`<jemalloc>: ../contrib/jemalloc/include/jemalloc/internal/arena_inlines_a.h:45: Failed assertion: "newarena != NULL"`
https://clickhouse-test-reports.s3.yandex.net/16836/6e8945d78c665e8b0155fcb8d8367320f538ebaa/fuzzer/server.log
We can see the possible reason... | 2020-11-10T19:12:33 | 2020-11-10T19:12:33 | {} | MEMBER |
ClickHouse | ClickHouse | 725,189,595 | 16,843 | amosbird | Heh, it's somehow similar to https://github.com/ClickHouse/ClickHouse/pull/7549 | 2020-11-11T04:27:22 | 2020-11-11T04:27:22 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 725,254,478 | 16,851 | vegastar002 | but the locked table still cann't use and delete . How to delete it ? | 2020-11-11T07:23:05 | 2020-11-11T07:23:05 | {} | NONE |
ClickHouse | ClickHouse | 725,413,994 | 16,847 | akuzm | Please refer to these comments explaining the possible reasons for this problem:
https://github.com/ClickHouse/ClickHouse/issues/3174#issuecomment-423435071
https://github.com/ClickHouse/ClickHouse/issues/6720#issuecomment-526045768
| 2020-11-11T13:09:24 | 2020-11-11T13:09:24 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,731,748 | 8,828 | alexey-milovidov | Wow, it works!
```
SELECT count()
FROM url('https://clickhouse-datasets.s3.yandex.net/hits/tsv/hits_v1.tsv.xz', 'TSV', '`WatchID` UInt64, `JavaEnable` UInt8, `Title` String, `GoodEvent` Int16, `EventTime` DateTime, `EventDate` Date, `CounterID` UInt32, `ClientIP` UInt32, `ClientIP6` FixedString(16), `RegionID` UIn... | 2020-11-12T00:01:25 | 2020-11-12T00:01:25 | {
"hooray": 2
} | MEMBER |
ClickHouse | ClickHouse | 725,424,455 | 15,073 | sundy-li | > alexey-milovidov: I wonder why don't allow multi-column primary key?
eg: primary key (a, b, c)
- I think it's useless, we can define multi-column as a single Tuple column instead.
- It's hard to apply index for multi-column primary key, such as `where a = 1 and b in (2, 3) and c in (4, 5)`, this may rewrite to `... | 2020-11-11T13:31:31 | 2020-11-12T05:51:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,939,376 | 9,117 | flipper86 | > > I tried to achieve the same, couldn't change the system.zookeeper table. I'm currently stuck because I'm trying to get create a replicated table, but I get an error saying the zookeeper path already exists, yet I can't do any other operation because there is no table (no data nor metadata in clickhouse). So the zoo... | 2020-11-10T20:11:02 | 2020-11-10T20:11:02 | {
"eyes": 2,
"rocket": 1
} | NONE |
ClickHouse | ClickHouse | 725,400,739 | 16,610 | zhang2014 | > Performance — 2 errors, 1 too long, 8 faster, 91 unstable
Some queries have unexpected duration - unrelated
The test 'read_in_order_many_parts' is too slow to run as a whole. Investigate whether the create and fill queries can be sped up - unrelated
| 2020-11-11T12:39:37 | 2020-11-11T12:39:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,419,934 | 16,814 | akuzm | Test failures are known and not related. | 2020-11-11T13:22:13 | 2020-11-11T13:22:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,077,190 | 16,578 | fibersel | `HTTPHandlers.cpp` has too many changes because I did clang-format -i src/Server/HTTPHandler.cpp before commiting. | 2020-11-11T01:53:45 | 2020-11-11T14:03:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,469,803 | 16,869 | alexey-milovidov | @filimonov If it is totally wrong, we can simply delete this function, and only then commit the replacement. | 2020-11-11T14:56:14 | 2020-11-11T14:56:14 | {} | MEMBER |
ClickHouse | ClickHouse | 725,474,674 | 16,876 | yohannj | Hello Alexei,
Thank you, the reasons behind ClickHouse behavior is very clear.
Regarding avoiding Int / UInt -> Float. This behavior is still there on the latest LTS version (20.3.21.2).
As for promoting up to 64bit, since we got Int128 and Int256 in 20.10, is there any purpose in not using them here?
For the l... | 2020-11-11T15:04:58 | 2020-11-11T15:04:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,874,585 | 16,830 | hexiaoting | @anzi1983 check your configurate file: config1.xml
Chinese: 检查一下你的配置文件config1.xml, 里面的<compression incl="clickhouse_compression">和<remote_servers incl="clickhouse_remote_servers">是否配置正确。 | 2020-11-12T06:43:09 | 2020-11-12T06:43:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 724,959,866 | 16,789 | alexey-milovidov | We don't use rebase for another reasons:
- it makes difficult to find commits from PR in master and find PRs for commit in master;
- it makes difficult to check if PR was merged;
- if multiple PRs are based on the same set of commits and will be rebased to master, commits will duplicate;
- GitHub cannot automatical... | 2020-11-10T20:53:01 | 2020-11-10T20:56:14 | {} | MEMBER |
ClickHouse | ClickHouse | 725,238,872 | 16,851 | vegastar002 | Ooooo, beacause I use tinyLog, why dataX cann't write tinyLog table?
I change to mergeTree it's ok. | 2020-11-11T06:51:57 | 2020-11-11T06:51:57 | {} | NONE |
ClickHouse | ClickHouse | 725,482,436 | 16,869 | filimonov | > @filimonov If it is totally wrong, we can simply delete this function, and only then commit the replacement
From one side can't imagine how that can be used on the bigger datasets in current implementation (and wondering why nobody complain on that), but for single block it should work ok.
But removing that may... | 2020-11-11T15:19:25 | 2020-11-11T15:19:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 725,510,475 | 16,869 | alexey-milovidov | It's too obscure, let's remove it. | 2020-11-11T16:07:34 | 2020-11-11T16:07:34 | {} | MEMBER |
ClickHouse | ClickHouse | 725,557,732 | 15,073 | alexey-milovidov | @sundy-li
> It's hard to apply index for multi-column primary key
> If it's really needed in the future for some reasons, we can add it later.
Ok.
> It's strange that this PR view in github involves lots of changes when I pull rebase from origin/master.
I don't know the reason but don't worry... | 2020-11-11T17:32:24 | 2020-11-11T17:32:24 | {} | MEMBER |
ClickHouse | ClickHouse | 725,770,547 | 16,851 | vegastar002 | > What version of ClickHouse do you use? Do you have some example queries that can reproduce the problem?
my version is clickhouse 20.8.
the sql of tb1 :
create table tb1
(
)
engine = TinyLog
then I use dataX start import data, it will report error
[https://note.youdao.com/ynoteshare1/index.html?id=d28b4... | 2020-11-12T01:34:12 | 2020-11-12T01:34:12 | {} | NONE |
ClickHouse | ClickHouse | 729,170,480 | 16,532 | den-crane | @atercattus I think you should share such part with @alesapin | 2020-11-17T20:04:34 | 2020-11-17T20:04:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 729,242,070 | 17,054 | l1t1 | got it. thanks | 2020-11-17T22:05:18 | 2020-11-17T22:05:18 | {} | NONE |
ClickHouse | ClickHouse | 729,316,595 | 17,161 | vzakaznikov | @alexey-milovidov, Hi Alexey, this PR should fix unstable test in the LDAP test suite. | 2020-11-18T01:27:34 | 2020-11-18T04:15:54 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.