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 | 801,454,569 | 18,978 | tavplubix | > The part with corrupted data may produce the same checksum as the part with the untouched data
> So, the part with the corrupted data.bin file and the part with the untouched data produce the same checksums,
which obviously can't be true.
> I tried explicitly deleting the .bin file on replica2, no luck.
Just ta... | 2021-03-17T21:27:34 | 2021-03-17T21:27:34 | {} | MEMBER |
ClickHouse | ClickHouse | 801,503,702 | 21,847 | genzgd | I'll try to create a reproducible test case tomorrow, but I'm putting this out here in case someone knows the answer immediately. I did not see anything in open or closed issues that seemed related. We're in the process of upgrading our clusters to the 21.3 release, so if that fixes this it would be good to know. | 2021-03-17T23:18:57 | 2021-03-17T23:18:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,564,934 | 21,821 | spongedu | @filimonov Do we have plan to make a change here ? I think it's a little confusing using `replication_alter_partitions_sync` to control the synchronicity here. since `MODIFY_COLUMN`, `DROP_COLUMN` etc are recognized as `mutations`. as:
https://github.com/ClickHouse/ClickHouse/blob/b8a437c35adc40cf91dc6961d76c655790... | 2021-03-18T02:17:23 | 2021-03-18T03:10:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,797,962 | 21,773 | filimonov | Please provide the whole stacktrace of the exception (should be in logs).
Side note: `CODEC(ZSTD(22))` is very slow, and can significantly impact insert / merge speed (while of course it should not lead to that issue) | 2021-03-17T05:08:06 | 2021-03-17T05:12:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,887,432 | 21,058 | feidaodalian | @windylcx hi,I have also this question #21867 ，Excuse me, did you solve the problem | 2021-03-18T12:30:49 | 2021-03-18T12:30:49 | {} | NONE |
ClickHouse | ClickHouse | 800,828,484 | 21,569 | lehasm | We can change it to^
_A query may be processed faster if it is executed on several servers in parallel. But there are *a lot of cases* when query performance may degrade:_
@den-crane, do you agree? | 2021-03-17T06:21:24 | 2021-03-17T06:21:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,852,977 | 21,759 | TCeason | Because of this `https://github.com/ClickHouse/ClickHouse/pull/20961`
https://github.com/ClickHouse/ClickHouse/commit/641b1b249e6eaead8eb8468acefad1d26418b17c#diff-2265e250dfa7b4c91ff77977f8cab650df63b05f4021fdd180d06e34dcf4939bR657
`Integration tests (asan)` has been fatal.
I will look at it and try to fix i... | 2021-03-17T07:12:16 | 2021-03-17T07:12:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,900,428 | 21,817 | filimonov | > You mean that when allocating 4MB data, it reaches the limit?
Yes.
It was "the straw that broke the camel’s back" :)
So if successfully allocated 50Gb before for this (and other) tasks/queries, and when it requested 4Mb more it reached the limit.
| 2021-03-17T08:36:42 | 2021-03-17T08:37:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,905,517 | 21,817 | filimonov | ```Memory limit (total) exceeded: would use 56.02 GiB```
It does not correspond to 'free 46G'.
You need to check the memory usage and check query log when issue was happening.
> No, there are no concurrent queries, but concurrent insert.
From what i see your single insert requires about 3Gb of RAM. So if... | 2021-03-17T08:44:52 | 2021-03-17T08:44:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,911,121 | 21,579 | KochetovNicolai | Build is failed because of ci issue. | 2021-03-17T08:54:04 | 2021-03-17T08:54:04 | {} | MEMBER |
ClickHouse | ClickHouse | 800,915,359 | 21,817 | filimonov | > and what do you think about it could run normally before
Have no idea.
> will the docker processes affact it?
If you run 3 docker containers - obviously they will use more RAM altogether, comparing to a single instance.
| 2021-03-17T09:01:04 | 2021-03-17T09:01:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,035,625 | 21,831 | alexey-milovidov | What version of ClickHouse do you use? | 2021-03-17T12:16:55 | 2021-03-17T12:16:55 | {} | MEMBER |
ClickHouse | ClickHouse | 801,037,350 | 21,831 | jorisgio | 21.2.5.5
Actually it might be an instance of #19684 i will try to update | 2021-03-17T12:19:47 | 2021-03-17T12:22:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,039,769 | 6,286 | alexey-milovidov | Yes, it is unassigned.
> I can pick it up if none
It will be wonderful! | 2021-03-17T12:23:55 | 2021-03-17T12:23:55 | {} | MEMBER |
ClickHouse | ClickHouse | 801,655,433 | 21,833 | filimonov | > > Did you check windowFunnel function?
>
> @filimonov thanks for the suggestion! I just double checked that function, and that is the one I need! However, the official documentation about `window` parameter is kind of misleading, so I propose a [fix](https://github.com/ClickHouse/ClickHouse/issues/21849) for that,... | 2021-03-18T06:04:14 | 2021-03-18T06:04:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,004,113 | 21,875 | den-crane | JSONEachRow is a text format.
0.00 - string representation, it's parsed as Decimal. | 2021-03-18T15:07:10 | 2021-03-18T15:07:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,125,408 | 21,813 | akuzm | A pull request with just the `zOrder` function will be appreciated -- it is relatively easy to do, and I think we already have an implementation somewhere in the old pull requests... We will merge it soon and at least have some progress, which is good because this task is stuck for too long already.
The query optimize... | 2021-03-17T14:26:17 | 2021-03-17T14:26:17 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,099,675 | 21,806 | ravil-iskinin | I'm trying to build СH tag v21.2.6.1-stable.
Yes, I've updated submodules.
Ок, I will compile and install v21.4.1.6251-testing, also I would like to get a stable version with this fix. :))
| 2021-03-17T13:54:28 | 2021-03-17T16:14:28 | {} | NONE |
ClickHouse | ClickHouse | 801,325,786 | 21,701 | vialorn | @kssenii , thanks a lot! | 2021-03-17T18:47:08 | 2021-03-17T18:47:08 | {} | NONE |
ClickHouse | ClickHouse | 801,344,329 | 21,806 | alexey-milovidov | You don't have to build all targets (utils), just type
```
ninja clickhouse
```
Then if you want to install ClickHouse, run
```
sudo build/programs/clickhouse install
``` | 2021-03-17T19:16:32 | 2021-03-17T19:16:32 | {} | MEMBER |
ClickHouse | ClickHouse | 801,787,841 | 21,082 | BoloniniD | I considered to reopen this PR as a new one soon. I think it will be better to implement yaml methods not as a new class but as methods of old ConfigProcessor, so the current code needs to be reworked. | 2021-03-18T09:58:32 | 2021-03-18T09:58:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,890,296 | 20,179 | Jokser | @alexey-milovidov
Since 21.1 (after this fix https://github.com/ClickHouse/ClickHouse/pull/17934) memory consumption for S3 is O(1) and it's equal to max (s3_min_upload_part_size, s3_max_single_part_upload_size) for each part column file. | 2021-03-18T12:36:00 | 2021-03-18T12:36:00 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,042,024 | 21,781 | bun4uk | > @bun4uk you can try to use https://github.com/lmangani/clickhouse-metrics-prometheus to allow export custom SQL queries as Prometheus metrics
I'd like to not use it until it not stable
<img width="657" alt="CleanShot 2021-03-18 at 17 54 02@2x" src="https://user-images.githubusercontent.com/14631690/111656173-f371... | 2021-03-18T15:55:13 | 2021-03-18T15:55:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,794,652 | 21,816 | filimonov | As exception says you are trying to compare `AggregateFunction(sum, UInt8)` with `UInt8` somewhere. It looks like you forget to use -Merge combinator somewhere for you AggregatingMergeTree. | 2021-03-17T04:58:53 | 2021-03-17T04:58:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,906,316 | 21,817 | SE2AI | yes that make sense.
but I tracked the memory usage, and found it only cost a part of memory, after setting max_memory_usage to 10G，I got the same exception message, even the same number **`56.02 GiB (attempt to allocate chunk of 4219316 bytes)`**
I will test again by reducing the concurrent number. and what do y... | 2021-03-17T08:46:12 | 2021-03-17T08:54:04 | {} | NONE |
ClickHouse | ClickHouse | 800,828,135 | 21,773 | Teeeto | This run memory requirement changed but still huge. If I use string filters on different columns at the same place error still comes up with different memory requirements (was testing data corruption hypothesis)
```
2021.03.17 10:15:04.376890 [ 202508 ] {72d64fae-d654-4cc2-aaca-b1e5cdcdd99a} <Error> executeQuery: Cod... | 2021-03-17T06:20:30 | 2021-03-17T09:09:54 | {} | NONE |
ClickHouse | ClickHouse | 801,001,502 | 21,739 | Enmk | > I think it's not happy about
>
> ```
> WARNING:mkdocs.structure.pages:Documentation file 'sql-reference/statements/optimize.md' contains a link to '../sql-reference/statements/select/index.md' which is not found in the documentation files.
> WARNING:mkdocs.structure.pages:Documentation file 'sql-reference/statem... | 2021-03-17T11:17:41 | 2021-03-17T11:17:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,083,501 | 20,244 | Felixoid | As I've mentioned in the dev-chat, I've found the request that fails even on localhost.
- I run `docker run --rm --net=host --name=clickhouse yandex/clickhouse-server:21.3` in one terminal
- Then in another I run `cat POST.txt | nc localhost 8123 | tail -c 153 | gzip -d`, file [POST.txt](https://github.com/ClickHou... | 2021-03-17T13:32:59 | 2021-03-17T13:32:59 | {} | MEMBER |
ClickHouse | ClickHouse | 801,212,004 | 21,816 | amitsharma10 | @filimonov It has been failing intermittently so there is no scope for me to minimize the query to reproduce the issue. I am not able to reproduce the issue even with this query. This query runs every five minutes and fails 3-4 times a day.
| 2021-03-17T16:10:09 | 2021-03-17T16:10:09 | {} | NONE |
ClickHouse | ClickHouse | 801,565,224 | 21,833 | jinghancc | > Did you check windowFunnel function?
@filimonov thanks for the suggestion! I just double checked that function, and that is the one I need! However, the official documentation about `window` parameter is kind of misleading, so I propose a [fix](https://github.com/ClickHouse/ClickHouse/issues/21849) for that, do yo... | 2021-03-18T02:18:08 | 2021-03-18T02:18:08 | {} | NONE |
ClickHouse | ClickHouse | 802,050,660 | 21,875 | iassasin | May be I something misunderstand, but this cast `cast(id1 as Float64)` is precision lossy in general case. But now I can't provide an example of value wich show this loss. Thanks for answers. | 2021-03-18T16:05:24 | 2021-03-18T16:05:24 | {} | NONE |
ClickHouse | ClickHouse | 800,838,774 | 21,817 | SE2AI | Another small question is when I run `clickhouse stop`, the server will start automatically. It seems there is a daemon process. So how to stop the server completely?
btw, this problem appears to CH instance, but it's ok for CH cluster :) | 2021-03-17T06:42:16 | 2021-03-17T06:42:16 | {} | NONE |
ClickHouse | ClickHouse | 800,852,479 | 21,701 | kssenii | @vialorn, yes, it won't, it will be in 21.4. | 2021-03-17T07:11:09 | 2021-03-17T07:11:09 | {} | MEMBER |
ClickHouse | ClickHouse | 800,945,590 | 21,739 | filimonov | I think it's not happy about
```
WARNING:mkdocs.structure.pages:Documentation file 'sql-reference/statements/optimize.md' contains a link to '../sql-reference/statements/select/index.md' which is not found in the documentation files.
WARNING:mkdocs.structure.pages:Documentation file 'sql-reference/statements/optim... | 2021-03-17T09:48:01 | 2021-03-17T09:48:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,037,260 | 21,829 | alexey-milovidov | @azat I thought about this issue when we've added suggestions for system.errors but it does not look important. | 2021-03-17T12:19:39 | 2021-03-17T12:19:39 | {} | MEMBER |
ClickHouse | ClickHouse | 801,122,222 | 21,813 | akuzm | It probably doesn't require non-standard grammar. The first step is to create a function `zOrder(lat, lon)`, that takes several numeric values and returns the Z-order value for them. We can index on this function, e.g. `ORDER BY zOrder(lat, lon)`. Then we will teach the query optimizer to deduce restrictions on Z-order... | 2021-03-17T14:23:14 | 2021-03-17T14:23:14 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,214,682 | 21,816 | den-crane | Now it looks like your shards / replicas has different tables behind distributed table lmd.otd_sdr_deviation_new
Check that all tables are the same
```
List of inconsistent columns
with (
select groupArray(FQDN()) from clusterAllReplicas({cluster},system,one)
) as hosts
select database,
... | 2021-03-17T16:13:32 | 2021-03-17T16:13:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,324,797 | 18,978 | myrrc | Ok, as for now it seems the test started working (not failing due to the timeout), but here we have two issues:
1. The part with corrupted data may produce the same checksum as the part with the untouched data:
While running test we have the following logs for `replica1`:
```
> File data.bin, file size 67, file... | 2021-03-17T18:45:23 | 2021-03-17T18:50:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,981,965 | 21,875 | den-crane | JSONEachRow is not JSON.
JSONEachRow expects `{"id":0.00}` | 2021-03-18T14:38:10 | 2021-03-18T14:38:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,018,070 | 21,875 | iassasin | For example in NodeJS to achieve this string representation of decimal in json-string I'll need to use custom json stringification function instead of built-in `JSON.stringify`. Also in JSON specification it is not actually string, it is number (see section "Numbers" [here](https://www.ecma-international.org/wp-content... | 2021-03-18T15:24:44 | 2021-03-18T15:24:44 | {} | NONE |
ClickHouse | ClickHouse | 800,792,622 | 21,813 | filimonov | Please see https://github.com/ClickHouse/ClickHouse/pull/6286
I guess the work is paused there again, right @alexey-milovidov ?
You can definitely collaborate. | 2021-03-17T04:52:50 | 2021-03-17T04:52:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,207,961 | 21,835 | jinghancc | > Maybe introduced in #18884.
Hi @ucasFL , thanks for finding the related PR! But I guess this issue existed before this PR because duplicate events cannot be avoided by changing sorting method.
I wonder if we should change logic [here](https://github.com/ClickHouse/ClickHouse/blob/master/src/AggregateFunctions/... | 2021-03-17T16:04:53 | 2021-03-17T16:04:53 | {} | NONE |
ClickHouse | ClickHouse | 801,214,788 | 21,813 | bakey | > It probably doesn't require non-standard grammar. The first step is to create a function `zOrder(lat, lon)`, that takes several numeric values and returns the Z-order value for them. We can index on this function, e.g. `ORDER BY zOrder(lat, lon)`. Then we will teach the query optimizer to deduce restrictions on Z-ord... | 2021-03-17T16:13:40 | 2021-03-17T16:13:40 | {} | NONE |
ClickHouse | ClickHouse | 801,227,668 | 21,813 | alexey-milovidov | You can extract the existing code from #6286 into a new PR. | 2021-03-17T16:30:27 | 2021-03-17T16:30:27 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 801,350,185 | 21,770 | filimonov | Please check / fix build failures.
```
2021-03-17 09:50:53 /build/src/Processors/Formats/Impl/ArrowColumnToCHColumn.cpp:650:33: error: variable 'list_nested_type' may be uninitialized when used here [-Werror,-Wconditional-uninitialized]
2021-03-17 09:50:53 switch (list_nested_type)
2021-0... | 2021-03-17T19:26:44 | 2021-03-17T19:26:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,443,383 | 21,569 | den-crane |
ok | 2021-03-17T21:07:06 | 2021-03-17T21:07:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,649,234 | 20,249 | fastio | I create the new [pr](https://github.com/ClickHouse/ClickHouse/pull/21850). And close this one. | 2021-03-18T05:51:40 | 2021-03-18T05:51:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,697,623 | 21,850 | fastio | > @fastio looks useful, can you add a test (`tests/integration/test_sotrage_kafka`)?
Sure. | 2021-03-18T07:29:28 | 2021-03-18T07:29:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,700,179 | 21,849 | sundy-li | Not yet, the doc is right. The window is between sequence events. It's not between the first condition and last condition. | 2021-03-18T07:33:33 | 2021-03-18T07:33:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,722,694 | 20,500 | adiletkabylbekov | reproduced on version 20.8.14.4 | 2021-03-18T08:14:09 | 2021-03-18T08:14:09 | {} | NONE |
ClickHouse | ClickHouse | 801,774,657 | 21,690 | alesapin | Looks like with `force_fsync` NuKeeper doesn't work properly on sandbox HDDs. | 2021-03-18T09:37:29 | 2021-03-18T09:37:29 | {} | MEMBER |
ClickHouse | ClickHouse | 801,982,422 | 21,851 | vdimir | Could you provide table schema and sample data? | 2021-03-18T14:38:45 | 2021-03-18T14:38:45 | {} | MEMBER |
ClickHouse | ClickHouse | 800,828,641 | 21,817 | SE2AI | thanks @filimonov
The target table has 680 columns, it's a kind of wide-sparse table, and the settings of user.xml and config.xml are same as default. I just curious about that the same test DML could be performed normally before, the test code are the same. Maybe there are some differences of the OS, I run three do... | 2021-03-17T06:21:46 | 2021-03-17T08:30:43 | {} | NONE |
ClickHouse | ClickHouse | 800,795,929 | 21,816 | amitsharma10 | @filimonov
There is no AggregatingMergeTree being used here.
Here is the table definition -
```
CREATE TABLE lmd.local_otd_sdr_deviation_new
(
`eventName` String,
`customerStartTime` UInt64,
`customerEndTime` UInt64,
`actualDeliveryTime` UInt64,
`serviceProvider` String,
`carrier... | 2021-03-17T05:02:31 | 2021-03-17T09:04:16 | {} | NONE |
ClickHouse | ClickHouse | 801,565,393 | 21,814 | lxswbr | The execution is indeed completed. How do I use a longer timeout period, set it during the Python connection or configure it in the clickhouse configuration file? What is the name of the parameter? I set some parameters before and it didn't work. | 2021-03-18T02:18:38 | 2021-03-18T02:18:38 | {} | NONE |
ClickHouse | ClickHouse | 801,997,011 | 21,875 | iassasin | @den-crane but why? At least error message seems incorrect in this case. It says that can't parse json and expecting comma, but json is valid. Also in documentation nothing about this.
Last and important point - if it will stringified or parsed from float value, for decimal it can be precision-loss, so transfer it as ... | 2021-03-18T14:57:55 | 2021-03-18T14:57:55 | {} | NONE |
ClickHouse | ClickHouse | 800,803,213 | 21,817 | filimonov | How many columns do the table have? Can you share the schema? Do you have any non-standard settings?
Do you have some concurrent queries running? Can you check queries memory usage in system.query_log?
> What does the chunk of 4219316 bytes mean?
It literały means that during the insert execution clickhouse... | 2021-03-17T05:22:24 | 2021-03-17T05:22:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,155,477 | 21,831 | jorisgio | i just realized that GlobalThreadsActive >>> LocalThreads is not necessarily a problem. The comment in system.metrics is incorrect with processors pipeline, as it seems to allocate threads directly from GlobalPool without passing through a LocalThreadPool. | 2021-03-17T15:00:56 | 2021-03-17T15:00:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,142,808 | 21,839 | kssenii | Also `clickhouse-benchmark --query "select * from postgres_table" -c 20` works fine. | 2021-03-17T14:45:07 | 2021-03-17T15:54:27 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 801,695,153 | 21,833 | jinghancc | > > > Did you check windowFunnel function?
> >
> >
> > @filimonov thanks for the suggestion! I just double checked that function, and that is the one I need! However, the official documentation about `window` parameter is kind of misleading, so I propose a [fix](https://github.com/ClickHouse/ClickHouse/issues/2184... | 2021-03-18T07:24:27 | 2021-03-18T07:24:27 | {} | NONE |
ClickHouse | ClickHouse | 801,852,381 | 21,599 | garutilorenzo | Hi @filimonov
i've made few other changes:
* Dockerfile Use sticky bit on clickhouse directories, change owner to clickhouse:clickouse
* in entrypoint.sh: simplified use of gosu, reintroduced DO_CHOWN variable to to keep variables names logical/consistent, simplified logic for chown directories: setup_dirs now ... | 2021-03-18T11:42:17 | 2021-03-18T11:42:17 | {} | NONE |
ClickHouse | ClickHouse | 802,000,279 | 21,754 | akuzm | The fuzzer failure does not reproduce, others are not related. | 2021-03-18T15:02:06 | 2021-03-18T15:02:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,791,313 | 21,763 | WesleyBatista | > _From RBAC you usually expect dererminism w/o too many 'automatic' changes_
Thanks for bringing this up. I didn't have this perspective.
Unfortunately, I have no strong background on security to contribute more on the topic.
> _Can you explain why exactly is it needed? Playground currently uses single user._... | 2021-03-17T04:48:54 | 2021-03-17T04:48:54 | {} | NONE |
ClickHouse | ClickHouse | 801,237,139 | 21,811 | alexey-milovidov | > test_materialize_mysql_database
- known bad test. | 2021-03-17T16:42:52 | 2021-03-17T16:42:52 | {} | MEMBER |
ClickHouse | ClickHouse | 801,012,548 | 21,820 | filimonov | See https://github.com/ClickHouse/ClickHouse/pull/11512#discussion_r450226856
And setting `mysql_datatypes_support_level`
| 2021-03-17T11:36:36 | 2021-03-17T21:02:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,736,079 | 21,621 | winter7 | @alesapin Is there a released version for this fix please？ | 2021-03-18T08:37:05 | 2021-03-18T08:37:05 | {} | NONE |
ClickHouse | ClickHouse | 802,043,544 | 21,875 | den-crane | >You really say that precision loss for Decimal is expected behavior while importing/exporting data in clickhouse?
the opposite.
```sql
CREATE TABLE test2(id Decimal(38, 18), id1 Float64) ENGINE = MergeTree() ORDER BY id ;
INSERT INTO test2 FORMAT JSONEachRow {"id":111111.111111111111111111, "id1":111111.111111... | 2021-03-18T15:57:15 | 2021-03-18T15:57:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,036,770 | 21,829 | alexey-milovidov | Yes, it will happen if you connect with newer client to older server.
Workarounds:
- install new server version;
- ignore this message (clickhouse-client will work nevertheless);
- run client without suggestions (`-A` command line argument). | 2021-03-17T12:18:48 | 2021-03-17T12:20:16 | {} | MEMBER |
ClickHouse | ClickHouse | 801,117,521 | 21,828 | akuzm | Looks like the optimization for reading in primary key order is applied incorrectly when there is also sorting for window functions:
```
:) select date, event, count(*) over (partition by event) count from events order by date limit 10 settings optimize_read_in_order = 0;
┌────────────────date─┬─event───────────... | 2021-03-17T14:16:52 | 2021-03-17T14:17:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,286,226 | 21,843 | den-crane | duplicate https://github.com/ClickHouse/ClickHouse/issues/20865 | 2021-03-17T17:49:38 | 2021-03-17T17:49:38 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,376,736 | 21,569 | lehasm | Changed _some cases_ to _following cases_ (neutral variant in my opinion). | 2021-03-17T19:50:42 | 2021-03-17T19:50:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,404,614 | 21,806 | ravil-iskinin | @alexey-milovidov, Thank you | 2021-03-17T20:10:01 | 2021-03-17T20:10:01 | {} | NONE |
ClickHouse | ClickHouse | 801,846,869 | 16,726 | bruno-lp | Hi
Any update or planned date on this ?
I hesitate moving from mariadb to mysql to use the materializeMySQL feature | 2021-03-18T11:31:49 | 2021-03-18T11:31:49 | {} | NONE |
ClickHouse | ClickHouse | 801,859,740 | 21,781 | bun4uk | > Did you consider extracting data from `system.tables`?
Yes. I can have a total count of rows for each table from `system.tables`. But to have such dashboard, that I attached, I have to do periodical snapshots of system.tables table and count delta between current and previous table state. It requires creating som... | 2021-03-18T11:49:59 | 2021-03-18T11:49:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,784,491 | 21,772 | filimonov | Please ensure it's really hang (other queries are not working, while clickhouse-server process is active) ensure your VM have enough ram, ensure that the process was not killed by oomkiller (see `dmesg`).
We need fragment of log just before the issue:
```
sudo tail -n 1000 /var/log/clickhouse-server/clickhouse-se... | 2021-03-17T04:28:08 | 2021-03-17T04:30:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,806,987 | 21,816 | filimonov | Please try to minimize your query, removing pieces of it while the issue reproduces.
https://stackoverflow.com/help/minimal-reproducible-example
Check if it reproduces it on supported clickhouse versions. | 2021-03-17T05:31:12 | 2021-03-17T05:38:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,340,250 | 21,724 | alexey-milovidov | > Hanged? (Actually the are not related anyway)
They are completed :) | 2021-03-17T19:10:35 | 2021-03-17T19:10:35 | {} | MEMBER |
ClickHouse | ClickHouse | 801,434,498 | 21,833 | filimonov | Did you check windowFunnel function? | 2021-03-17T20:51:19 | 2021-03-17T20:51:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,839,540 | 21,082 | BoloniniD | Opened a new PR: https://github.com/ClickHouse/ClickHouse/pull/21858 | 2021-03-18T11:19:29 | 2021-03-18T11:19:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,996,844 | 21,761 | akuzm | The tests are broken in master. | 2021-03-18T14:57:43 | 2021-03-18T14:57:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,940,778 | 21,773 | filimonov | Confirmed:
```
DROP TABLE IF EXISTS Test;
CREATE TABLE Test
ENGINE = MergeTree()
PRIMARY KEY (String1,String2)
ORDER BY (String1,String2)
AS
SELECT
'String1_' || toString(number) as String1,
'String2_' || toString(number) as String2,
'String3_' || toString(number) as String3,
'String4_' || to... | 2021-03-17T09:40:17 | 2021-03-17T09:50:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,304,026 | 21,724 | azat | >Integration tests (thread) Pending — Started
Hanged? (Actually the are not related anyway) | 2021-03-17T18:15:52 | 2021-03-17T18:15:52 | {} | MEMBER |
ClickHouse | ClickHouse | 801,442,985 | 21,821 | filimonov | Historical reasons:
replication_alter_partitions_sync
was exists long before mutations were added.
https://github.com/ClickHouse/ClickHouse/commit/d3664eac890691f06068459d57895550f70aaefe | 2021-03-17T21:06:23 | 2021-03-17T21:06:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,884,283 | 21,773 | Teeeto | I was trying to build a test dataset to reproduce the issue and encountered a strange behavior which may or may not be related to the issue.
Generate test data:
```
CREATE TABLE Test
(
String1 String codec(LZ4HC(0)),
String2 String codec(LZ4HC(0)),
String3 String codec(LZ4HC(0)),
String4 String codec (LZ4HC(0... | 2021-03-17T08:08:31 | 2021-03-17T09:32:41 | {
"+1": 1,
"confused": 1
} | NONE |
ClickHouse | ClickHouse | 801,905,166 | 20,179 | vmihailenco | I can confirm that it is working in 21.2 even with default settings.
>for each part column file.
@Jokser Do you mean that each column in a part is uploaded in parallel? That could explain why memory usage is still larger than default 512mb (s3_min_upload_part_size=536870912). But it also means that memory usage v... | 2021-03-18T12:56:35 | 2021-03-18T12:56:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,036,509 | 21,781 | Slach | @bun4uk you can try to use https://github.com/lmangani/clickhouse-metrics-prometheus to allow export custom SQL queries as Prometheus metrics | 2021-03-18T15:48:24 | 2021-03-18T15:48:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 802,032,687 | 21,849 | jinghancc | > Not yet, the doc is right. The window is between sequence events. It's not between the first condition and last condition.
Hi @sundy-li , thanks for your response, but the window is between the first condition and last condition based on test query.
First example: user 123 completed event 1, event 2, event 3 wi... | 2021-03-18T15:43:35 | 2021-03-18T15:49:14 | {} | NONE |
ClickHouse | ClickHouse | 800,787,083 | 21,814 | filimonov | Optimize will be finished anyway (you can see it running in `show processlist`).
Just reconnect or use bigger timeouts. | 2021-03-17T04:35:52 | 2021-03-17T04:35:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,962,625 | 8,687 | Zhile | hi @alexey-milovidov, we also need `ARRAY JOIN x ARRAY JOIN y` feature. The workaround may be too tricky for us.
Hope we can have it in ClickHouse soon.
| 2021-03-17T10:14:36 | 2021-03-17T10:14:36 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 802,029,799 | 21,876 | baibaichen | just disable building memcpy-bench on macos? | 2021-03-18T15:39:47 | 2021-03-18T15:39:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,789,584 | 21,806 | filimonov | What version are you trying to build? Do you have submodules updated?
It looks like some release branches (21.1) are currently missing the commit https://github.com/ClickHouse/ClickHouse/commit/ab55556c5e3600528ba8f5e3d54638990b5b3a5b | 2021-03-17T04:43:44 | 2021-03-17T04:45:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,824,223 | 6,286 | bakey | @alexey-milovidov Has anyone dedicated in this feature now? I can pick it up if none | 2021-03-17T06:11:19 | 2021-03-17T06:11:19 | {} | NONE |
ClickHouse | ClickHouse | 801,038,411 | 21,815 | alexey-milovidov | CC @zhang2014 @BohuTANG | 2021-03-17T12:21:33 | 2021-03-17T12:21:33 | {} | MEMBER |
ClickHouse | ClickHouse | 801,100,464 | 21,835 | ucasfl | Maybe introduced in #18884. | 2021-03-17T13:55:32 | 2021-03-17T13:55:32 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 801,421,419 | 14,345 | den-crane | aaahahahah,
>SET bytes = max(bytes), timestamp = toStartOfHour(timestamp)
SET bytes = max(bytes), timestamp = min(toStartOfHour(timestamp)) | 2021-03-17T20:28:38 | 2021-03-17T20:28:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,892,080 | 15,162 | vdimir | @jobosk Seems that it's different case.
Now conditions that is not in the form `t1.col1 = t2.col2` where `t1` and `t2` different tables is not supported, except case of `INNER` join (for such cases conditions moved to `WHERE` on query processing, see https://github.com/ClickHouse/ClickHouse/issues/18720). Condition... | 2021-03-18T12:39:16 | 2021-03-18T12:49:46 | {} | MEMBER |
ClickHouse | ClickHouse | 783,961,591 | 20,219 | yiguolei | > @yiguolei
>
> > But it could not deal with ReplacingMergeTree Engine.
>
> Could you explain this reason more detailed? 3Q.
In ReplacingMergeTree table engine, data is distributed among all nodes according to the sharding key, the sharding key is part of "**primary key**". And Merging process will merge the r... | 2021-02-23T07:07:27 | 2021-02-23T07:07:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 784,045,451 | 12,454 | asafcombo | new issue at :
[21101](https://github.com/ClickHouse/ClickHouse/issues/21101) | 2021-02-23T09:50:34 | 2021-02-23T09:50:34 | {} | NONE |
ClickHouse | ClickHouse | 784,619,144 | 13,574 | xjewer | >Theoretically, all parts of a partition will be merged into one, and maybe it will be very large. Then "MOVE PART" is not appropriate for that large part, since we might expect some of the data (not all data) in the partition to be moved to the new shard.
what is the question?
there is a setting that can configu... | 2021-02-24T00:02:16 | 2021-02-24T00:02:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 784,712,419 | 21,115 | den-crane | CH does not merge parts by time.
Merge scheduler selects parts by own algorithm based on the current node workload / number of parts / size of parts.
https://github.com/ClickHouse/ClickHouse/issues/8322#issuecomment-567990939
>If the partitions are not merged, will the performance of clickhouse be effected?
No... | 2021-02-24T02:57:35 | 2021-02-24T02:58:20 | {
"+1": 2
} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.