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 | 794,045,132 | 21,560 | amosbird | cc @kitaisreal | 2021-03-09T15:26:54 | 2021-03-09T15:26:54 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 792,529,014 | 5,726 | blind-oracle | I guess we can close this one as the original ticket #5582 has more data. | 2021-03-08T07:13:40 | 2021-03-08T07:13:40 | {} | NONE |
ClickHouse | ClickHouse | 792,622,205 | 21,112 | kssenii | Though I think there are a few small changes needed to make MV `into` rabbitmq table work. For now it does not work, because connection is supposed to be closed once current outputStream is finished. And there are errors like yours, because no measures are taken to sustain this connection object while it becomes idle (... | 2021-03-08T09:36:46 | 2021-03-08T09:41:24 | {} | MEMBER |
ClickHouse | ClickHouse | 792,678,170 | 21,464 | lcwxz1989 | I have some confused about insert performance:
There is some suggestion In the [clickhouse doc](https://clickhouse.tech/docs/en/introduction/performance/#performance-when-inserting-data)
A: `To improve performance, you can make multiple INSERT queries in parallel, which scales linearly`. I suppose if I have scale the... | 2021-03-08T11:05:07 | 2021-03-08T11:05:07 | {} | NONE |
ClickHouse | ClickHouse | 792,845,008 | 21,524 | den-crane | Clickhouse merge scheduler does not have the goal to merge everything.
Merge scheduler balance number of parts in a partition with an amount of resources needed to merge them and plans merges if there are free resources available.
Merges are eventual and may never happen. It depends on the number of inserts tha... | 2021-03-08T15:48:06 | 2021-03-08T15:48:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,343,579 | 4,456 | MrSsunlight | > `clickhouse-odbc-bridge` always dies with server. Seems like server get abortion signal in `ExternalLoader` (class which load dictionaries including `odbc`) while loading dictionaries, after that it restarts and again loading dictionaries and get abortion. Can you collect some statistics about uptime of troublesome s... | 2021-03-09T03:53:24 | 2021-03-09T03:53:24 | {} | NONE |
ClickHouse | ClickHouse | 793,559,609 | 15,090 | shiyi23 | @filimonov @alesapin hello, is this pull request still a work in progress?
| 2021-03-09T08:43:24 | 2021-03-09T08:43:24 | {} | NONE |
ClickHouse | ClickHouse | 793,838,001 | 10,519 | pitwegner | I'm on ClickHouse server version `20.8.13.15` and get the same error (`Code: 73. DB::Exception: Unknown format Parquet: data for INSERT was parsed from stdin`) when trying to insert data using `cat file.parquet | clickhouse client -h host -u *** --password *** --query "INSERT INTO db.table FORMAT Parquet"` (Mac ClickHo... | 2021-03-09T12:44:20 | 2021-03-09T12:46:33 | {} | NONE |
ClickHouse | ClickHouse | 793,898,034 | 21,498 | CurtizJ | Also, maybe it will be useful to have virtual column `_replica_num`. | 2021-03-09T13:28:03 | 2021-03-09T13:28:03 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 793,984,712 | 21,557 | vdimir | Seems that in build from master (90d019a2812c62cd2f8d3df74808c224c414ff97) reproduced too | 2021-03-09T14:40:34 | 2021-03-09T14:40:34 | {} | MEMBER |
ClickHouse | ClickHouse | 794,004,714 | 19,575 | vitlibar | @dankondr Are you going to continue working on this PR? Do you need some help maybe? | 2021-03-09T14:55:12 | 2021-03-09T15:02:52 | {} | MEMBER |
ClickHouse | ClickHouse | 792,846,556 | 21,518 | den-crane | Please provide
```
select distinct toYYYYMM(eventDate) from op_customer_action_month_summary where eventDate>today()-30
``` | 2021-03-08T15:50:34 | 2021-03-08T15:50:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,346,722 | 21,524 | den-crane | Clickhouse merge scheduler does not have the goal to merge everything. | 2021-03-09T03:57:19 | 2021-03-09T03:57:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,349,624 | 21,502 | den-crane | >but reduce cardinality (of index column) to one day.
In this case Clickhouse will be able to use PKindex for queries without `time = ` predicate. | 2021-03-09T04:00:48 | 2021-03-09T04:00:48 | {
"heart": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,877,068 | 4,959 | thinhnp-it | Any solution for this, guys ? I have same problem in latest release clickhouse-server 21.2.5.5 | 2021-03-09T13:09:31 | 2021-03-09T13:09:49 | {} | NONE |
ClickHouse | ClickHouse | 792,580,334 | 21,112 | khlystov | > BTW, I just noticed, you are using the same table for mv and inserts?
>
> > CREATE MATERIALIZED VIEW clickhouse.clickhouse_mv TO clickhouse.test AS
> > SELECT * FROM clickhouse.input;
> > INSERT INTO clickhouse.input (key, value) VALUES(52, 5221);
>
> It is written in doc (probably not easy to notice though) ... | 2021-03-08T08:38:49 | 2021-03-08T08:38:49 | {} | NONE |
ClickHouse | ClickHouse | 792,811,874 | 19,644 | taiyang-li | > Yes, there is conflict with syntax of parametrized queries.
> Example: `$CLICKHOUSE_CLIENT --param_test 'Hello, world' --query 'SELECT {test:String}'`
That is embarrassing, I didn't expect that... | 2021-03-08T14:58:52 | 2021-03-08T14:58:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 792,887,116 | 21,515 | bakey | I try your case on version 21.3,but it didn't replay。May be just upgrade the clickhosue-server? | 2021-03-08T16:40:39 | 2021-03-08T16:40:39 | {} | NONE |
ClickHouse | ClickHouse | 793,013,524 | 21,425 | alesapin | Ok, will make one more CI run just to be sure. | 2021-03-08T19:29:01 | 2021-03-08T19:29:01 | {} | MEMBER |
ClickHouse | ClickHouse | 793,475,138 | 21,541 | Sallery-X | I think the merge operation will be excute when i insert、alter or optimize, Besides will the merging be executed regularly when the storage start up？And How often？
| 2021-03-09T07:15:12 | 2021-03-09T07:15:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,556,811 | 21,409 | alesapin | All failures already broken in branch. | 2021-03-09T08:41:51 | 2021-03-09T08:41:51 | {} | MEMBER |
ClickHouse | ClickHouse | 793,583,738 | 21,489 | alesapin | 20.11 is out of support. | 2021-03-09T09:03:31 | 2021-03-09T09:03:31 | {} | MEMBER |
ClickHouse | ClickHouse | 793,910,011 | 21,136 | traceon | @alexey-milovidov By the way, if performance of the parsing is important inside a browser (and if relevant for the target use cases at all), another option to consider could be using native binary interface and adapting `clickhouse-cpp` to be compilable into WebAssembly for running inside a browser. | 2021-03-09T13:38:19 | 2021-03-09T13:38:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,925,690 | 4,959 | thinhnp-it | > Try change server, update clickhouse dont help with this issue.
Which settings need to be changed? Or you meant reinstall CH cluster ? Thanks. | 2021-03-09T13:51:47 | 2021-03-09T13:51:47 | {} | NONE |
ClickHouse | ClickHouse | 794,173,707 | 21,562 | CurtizJ | This PR has a huge diff, but mostly it's just moving serializations to separate files. However, git history for most serializations (largest part of code of data types) is kept, because there was a move beetween files in 1e61f64. But github is unable to show it properly in the interface.
Example of blame:
 desc t1;
┌─name─┬─type──┬─default_type─┬─default_expression─┬─comment─┬─codec_expression─┬─ttl_expression─┐
│ x │ Int32 │ │ │ │ │ ... | 2021-03-08T07:57:39 | 2021-03-08T07:57:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 792,634,142 | 20,172 | UprightNeether | I meet the same problem. FixedString need a positive length,since there is no length<0 in mysql,find the column which length is 0 | 2021-03-08T09:55:54 | 2021-03-08T09:55:54 | {} | NONE |
ClickHouse | ClickHouse | 794,007,477 | 21,556 | UnamedRus | > By manually providing the PREWHERE clause, the query performance become better. But I am not sure if this is the only problem in the newest version.
Can you show results of running query on new version? (GB scanned) | 2021-03-09T14:57:22 | 2021-03-09T14:57:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,168,957 | 21,562 | CurtizJ | However, now only default serialization is available (besides serialization of subcolumns) and not full interface of getting serializations is implemented in this PR, since it's worthless now. But it will look like this:
```c++
SerializationPtr IDataType::getSerialization(const String & column_name, const Serializa... | 2021-03-09T17:03:59 | 2021-03-09T17:10:30 | {} | MEMBER |
ClickHouse | ClickHouse | 792,557,128 | 21,112 | khlystov | @kssenii still the same
2021.03.08 08:57:35.928839 [ 3215 ] {} <Error> StorageRabbitMQ (sbgi_input1): Producer's channel 1_0 error: connection lost
2021.03.08 08:57:35.928964 [ 3215 ] {} <Error> StorageRabbitMQ (sbgi_input1): Library error report: connection lost
2021.03.08 08:58:20.541654 [ 3223 ] {} <Error> Sto... | 2021-03-08T08:02:37 | 2021-03-08T08:02:37 | {} | NONE |
ClickHouse | ClickHouse | 793,132,387 | 21,516 | alexey-milovidov | Maybe `materialize` function started to be slower. | 2021-03-08T22:32:38 | 2021-03-08T22:32:38 | {} | MEMBER |
ClickHouse | ClickHouse | 793,261,989 | 21,524 | Sallery-X | > Clickhouse merge scheduler does not have the goal to merge everything.
> Merge scheduler balance number of parts in a partition with an amount of resources needed to merge them and plans merges if there are free resources available.
>
> Merges are eventual and may never happen. It depends on the number of inserts... | 2021-03-09T01:58:03 | 2021-03-09T01:58:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,330,414 | 21,505 | anmolarora | @Avogar The chain `insert_integer` -> `insertValue` might make the value loose sign.
`value` (which is passed to `insert_integer`) is not being typecasted when calling `insertValue` on the appropriate column.
https://github.com/ClickHouse/ClickHouse/blob/4125a78be697abbc54638319b6e62459e7a022d2/src/Columns/ColumnVect... | 2021-03-09T03:36:13 | 2021-03-09T03:36:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,487,026 | 5,292 | thinhnp-it | @igor-sh8 How could you fix it? I got the same error on version 20.3.8.53 (official build) | 2021-03-09T07:37:00 | 2021-03-09T07:37:00 | {} | NONE |
ClickHouse | ClickHouse | 793,772,846 | 20,726 | alesapin | Complex conflicts | 2021-03-09T11:49:51 | 2021-03-09T11:49:51 | {} | MEMBER |
ClickHouse | ClickHouse | 793,896,746 | 21,505 | Avogar | @anmolarora Template typename T in `insertValue(const T value) ` is template of the whole class ColumnVector, not the specific method:
https://github.com/ClickHouse/ClickHouse/blob/4125a78be697abbc54638319b6e62459e7a022d2/src/Columns/ColumnVector.h#L95-L96
In `insert_integer` we firstly cast column to the needed an... | 2021-03-09T13:26:55 | 2021-03-09T13:26:55 | {} | MEMBER |
ClickHouse | ClickHouse | 792,556,769 | 21,406 | hexiaoting | > toNullable('1')
@vzakaznikov Map type do not support nullable type as key. | 2021-03-08T08:02:02 | 2021-03-08T08:02:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 792,584,885 | 21,112 | kssenii | But MV in regard to RabbitMQ table are used not in this way:
> CREATE MATERIALIZED VIEW rabbitmq.sbgi_input_mv TO **rabbitmq.sbgi_input** AS
select
version as objectId,
'employee' as objectType,
location as locationStringRaw
from **lkdn_profiles.employees**;
But it should be:
> CREATE MATERIALIZED VIEW ra... | 2021-03-08T08:45:45 | 2021-03-08T08:45:45 | {} | MEMBER |
ClickHouse | ClickHouse | 792,876,321 | 21,515 | den-crane | And you have different tables CLSFD_PI_TRAFFIC VS PI_TRAFFIC_SRC
```
ENGINE = Distributed('real_3shards_1replicas', 'tables', 'CLSFD_PI_TRAFFIC', rand())
;
CREATE TABLE tables.PI_TRAFFIC_SRC
``` | 2021-03-08T16:26:19 | 2021-03-08T16:26:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,474,185 | 21,425 | alesapin | `test_testkeeper_restore_from_snapshot/test.py::test_recover_from_snapshot` flaky -- with tsan server start 7 seconds but we wait for 5 only. | 2021-03-09T07:13:17 | 2021-03-09T07:13:17 | {} | MEMBER |
ClickHouse | ClickHouse | 793,716,864 | 21,550 | amosbird | cc @azat | 2021-03-09T10:59:29 | 2021-03-09T10:59:29 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 793,902,557 | 17,762 | stufently | > > before version 20.10.2.20
> > Yes we use mutations for clean old data, but data in part fresh, will delete after 3 month.
> > Ask about scheme , maybe nda
>
> Which version did you upgrade to fix it ?
Seems problem related with physicals server , after change server for this replica solve problem.
Update ... | 2021-03-09T13:31:59 | 2021-03-09T13:31:59 | {} | NONE |
ClickHouse | ClickHouse | 792,578,077 | 21,112 | kssenii | BTW, I just noticed, you are using the same table for mv and inserts?
> CREATE MATERIALIZED VIEW clickhouse.clickhouse_mv TO clickhouse.test AS
SELECT * FROM clickhouse.input;
INSERT INTO clickhouse.input (key, value) VALUES(52, 5221);
It is written in doc (probably not easy to notice though) that it is not allow... | 2021-03-08T08:35:32 | 2021-03-08T08:36:15 | {} | MEMBER |
ClickHouse | ClickHouse | 793,109,327 | 21,528 | genzgd | This very much depends on your use case.
At Comcast we use ClickHouse for processing several million log events per second with very fast queries. ClickHouse has replaced Splunk and ElasticSearch for most of these logs. However, it is definitely NOT simply a drop in replacement. A lot of additional work has gone ... | 2021-03-08T21:56:27 | 2021-03-08T21:56:27 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,481,800 | 21,504 | 15537261738 | In addition, I would also like to ask whether two machines that are copies of each other can consume Kafka. | 2021-03-09T07:27:44 | 2021-03-09T07:27:44 | {} | NONE |
ClickHouse | ClickHouse | 793,529,661 | 21,528 | nikbarseghyan | Ok. Thanks!! | 2021-03-09T08:17:57 | 2021-03-09T08:17:57 | {} | NONE |
ClickHouse | ClickHouse | 793,927,656 | 19,401 | akuzm | Integration and perf test failures are unrelated. | 2021-03-09T13:53:28 | 2021-03-09T13:53:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,931,959 | 4,959 | stufently | I mean real server for replica with this issue | 2021-03-09T13:57:11 | 2021-03-09T13:57:11 | {} | NONE |
ClickHouse | ClickHouse | 794,028,330 | 21,557 | UnamedRus | Not sure if it related.
But main query stuck on some stage of execution too.
→ Progress: 195.54 million rows, 4.28 GB (12.69 million rows/s., 277.55 MB/s.) █████████████████████████████████████████████████████████████
```
[LAPTOP-319G0GMQ] 2021.03.09 18:12:54.377399 [ 6026 ] {bd6cbc42-9c43-4740-9e82-c076e08753a3... | 2021-03-09T15:13:41 | 2021-03-09T15:15:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,295,057 | 16,240 | Sallery-X | I think need to consider the ddl worker, no need to do the ddl on the all replicas. | 2021-03-09T02:47:37 | 2021-03-09T02:47:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,064,522 | 21,557 | UnamedRus | [stack_trace2.txt](https://github.com/ClickHouse/ClickHouse/files/6109668/stack_trace2.txt)
| 2021-03-09T15:42:09 | 2021-03-09T15:42:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 792,588,222 | 21,112 | khlystov | > But MV in regard to RabbitMQ table are used not in this way:
>
> > CREATE MATERIALIZED VIEW rabbitmq.sbgi_input_mv TO **rabbitmq.sbgi_input** AS
> > select
> > version as objectId,
> > 'employee' as objectType,
> > location as locationStringRaw
> > from **lkdn_profiles.employees**;
>
> But it should be:
>... | 2021-03-08T08:50:52 | 2021-03-08T08:50:52 | {} | NONE |
ClickHouse | ClickHouse | 792,648,444 | 9,404 | alexey-milovidov | Now we have to deal with performance regression. | 2021-03-08T10:15:57 | 2021-03-08T10:15:57 | {} | MEMBER |
ClickHouse | ClickHouse | 793,435,026 | 21,504 | 15537261738 | In order to configure dual replicas, my cluster is configured with 3 units, two by two for redundancy.
This is really a stupid idea.
When I modify it to a single copy, the data remains consistent.
After modifying it into a double copy, I will observe whether it can be reproduced.
Thank you very much for your answe... | 2021-03-09T05:51:46 | 2021-03-09T05:51:46 | {} | NONE |
ClickHouse | ClickHouse | 793,553,402 | 21,408 | alesapin | Cannot be related to changes. Possibly flaky in this branch. | 2021-03-09T08:38:58 | 2021-03-09T08:38:58 | {} | MEMBER |
ClickHouse | ClickHouse | 793,605,082 | 21,468 | alesapin | Do you use a secure zookeeper connection? Which values you had in `Len error`? Slightly more than 1MB or giant almost random numbers? | 2021-03-09T09:21:28 | 2021-03-09T09:21:39 | {} | MEMBER |
ClickHouse | ClickHouse | 792,709,112 | 5,582 | Felixoid | Even after I've inserted the data `INSERT INTO default.graphite_one_insert SELECT * FROM default.graphite_crap_mv`, it still corrupted:
```
ig-fel :) SELECT Value, count() FROM graphite_one_insert WHERE Path = '12' GROUP BY Value
SELECT
Value,
count()
FROM graphite_one_insert
WHERE Path = '12'
GROUP... | 2021-03-08T12:01:00 | 2021-03-08T12:01:42 | {} | MEMBER |
ClickHouse | ClickHouse | 792,875,281 | 21,515 | den-crane | Please share the result of
```
select FQDN() h, length(g) col_num, groupArray(name) g
from clusterAllReplicas('real_3shards_1replicas', system.columns) where table = 'CLSFD_PI_TRAFFIC'
group by h
```
| 2021-03-08T16:25:00 | 2021-03-08T16:26:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,150,392 | 21,505 | Avogar | @anmolarora can you give an example when parsing negative integer is incorrect?
Here we have a chain of casting: `Int64 -> UInt64 -> Int8/Int16/Int32/Int64/UInt8/UInt16/UInt32/UInt64` (the last cast appears in calling `insertValue` in the needed column. And such chain is valid, we don't loose sign, because we cast `UI... | 2021-03-08T23:01:06 | 2021-03-08T23:02:13 | {} | MEMBER |
ClickHouse | ClickHouse | 793,442,687 | 21,538 | ngsiolei | Please have a look on `Kafka / Configuration / Kerberos support` in ClickHouse docs, https://clickhouse.tech/docs/en/engines/table-engines/integrations/kafka/#kafka-kerberos-support | 2021-03-09T06:07:56 | 2021-03-09T06:07:56 | {} | NONE |
ClickHouse | ClickHouse | 793,568,534 | 20,392 | vdimir | > Probably the reason is that we don't have a suitable structure for representing the join tree, and use raw AST for this?
Yes, I think so. This visitor always rewrites COMMA to CROSS replacing kind
https://github.com/ClickHouse/ClickHouse/blob/90d019a2812c62cd2f8d3df74808c224c414ff97/src/Interpreters/CrossToInnerJ... | 2021-03-09T08:50:56 | 2021-03-09T08:54:31 | {} | MEMBER |
ClickHouse | ClickHouse | 793,984,294 | 21,556 | OmarBazaraa | Version 20.12:
```
[ch-01-01-1-g0] 2021.03.09 12:28:32.227474 [ 12501 ] {84394410-e9e8-4bc7-b928-e4f7e4e8f828} <Debug> executeQuery: (from [::ffff:127.0.0.1]:57066) SELECT ProjectId, UserId, SessionId, Url, EntryUrl, ExitUrl, Browser, BrowserVersion, Device, OS, OSVersion, Country FROM clarity.page_data_all WHERE P... | 2021-03-09T14:40:13 | 2021-03-09T14:40:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,996,627 | 21,089 | cw9 | Gentle ping on this, I'm still seeing idle periods where no data replication is happening after hours of replication, I have to restart the new node to get the data replication to run in full speed again.
| 2021-03-09T14:50:09 | 2021-03-09T14:50:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 792,656,510 | 21,385 | alexey-milovidov | Closed in favor of https://github.com/ClickHouse/ClickHouse/pull/21516 and https://github.com/ClickHouse/ClickHouse/pull/21520 | 2021-03-08T10:29:32 | 2021-03-08T10:29:32 | {} | MEMBER |
ClickHouse | ClickHouse | 792,841,772 | 21,502 | phamtai97 | Thank you for your answer.
> but reduce cardinality (of index column) to one day.
The mean this idea?
| 2021-03-08T15:43:26 | 2021-03-08T15:43:26 | {} | NONE |
ClickHouse | ClickHouse | 793,257,995 | 21,518 | peaksnail | ```
SELECT DISTINCT toYYYYMM(eventDate)
FROM op_customer_action_month_summary
WHERE eventDate > (today() - 30)
┌─toYYYYMM(eventDate)─┐
│ 202103 │
└─────────────────────┘
┌─toYYYYMM(eventDate)─┐
│ 202102 │
└─────────────────────┘
``` | 2021-03-09T01:51:50 | 2021-03-09T01:51:50 | {} | NONE |
ClickHouse | ClickHouse | 793,292,306 | 21,515 | Sammion | Sorry, I make a mistake | 2021-03-09T02:43:26 | 2021-03-09T02:43:26 | {} | NONE |
ClickHouse | ClickHouse | 793,376,079 | 21,220 | alexey-milovidov | > it CAN"T be relative href="#../../"
It is related to one of the ugly hacks in the documentation build process.
See `def concatenate(lang, docs_path, single_page_file, nav)` in `single_page.py`
and `def test_single_page(input_path, lang)` in `test.py`.
The original author of this code cannot help us anymore an... | 2021-03-09T04:35:29 | 2021-03-09T04:38:29 | {} | MEMBER |
ClickHouse | ClickHouse | 793,551,722 | 21,490 | alesapin | Failure not related to changes. | 2021-03-09T08:37:34 | 2021-03-09T08:37:34 | {} | MEMBER |
ClickHouse | ClickHouse | 793,598,144 | 16,240 | ianton-ru | > > > How to solve the cache consistency problem of different replicas?
> >
> >
> > In Clickhouse data is "append only". Sometimes entry part can be deleted. So could you please describe where you see a problem?
>
> If multiple replicas share the same storage, how to resolve the share file confliction when the ... | 2021-03-09T09:15:36 | 2021-03-09T09:16:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,718,497 | 17,762 | thinhnp-it | > before version 20.10.2.20
> Yes we use mutations for clean old data, but data in part fresh, will delete after 3 month.
> Ask about scheme , maybe nda
Which version did you upgrade to fix it ? | 2021-03-09T11:00:53 | 2021-03-09T11:00:53 | {} | NONE |
ClickHouse | ClickHouse | 793,762,831 | 21,538 | Lipeng522 | OK,Thank you very much
But if I use this way, the clickhouse config will be changed, The cluster is not only for me, Maybe the owner of the cluster wouldn't like to change. is there anyone method that can be solved by set parameters when using kafka-engine.
Thank you, waiting for your reply. | 2021-03-09T11:40:34 | 2021-03-09T11:40:34 | {} | NONE |
ClickHouse | ClickHouse | 794,040,392 | 21,544 | alesapin | Ok, the code builds well, but we have no tests for `poll` version, so let's merge it before tests. | 2021-03-09T15:23:19 | 2021-03-09T15:23:19 | {} | MEMBER |
ClickHouse | ClickHouse | 792,603,783 | 21,112 | kssenii | MV `into` rabbitmq table is not implemented, you can only make MV `from` rabbitmq table.
So current setup is:
If you wan't to channel data `into` ClickHouse via RabbitMQ table engine, you use materialised view like `as select * from rabbitmq_table`. If you want to channel data `out of` ClickHouse into RabbitMQ, yo... | 2021-03-08T09:11:52 | 2021-03-08T09:11:52 | {} | MEMBER |
ClickHouse | ClickHouse | 792,626,449 | 10,352 | windylcx | > WA:
>
> ```
> insert into fact_product values('2020-02-11',2102982256,179819), ('2020-02-12',2102982256,180663 )
>
> SELECT
> *,
> neighbor(sold_cnt, -1) AS prev_sold_cnt
> FROM
> (
> SELECT *
> FROM
> (
> SELECT
> snapshot_time,
> product_id,
> ... | 2021-03-08T09:43:48 | 2021-03-08T09:43:48 | {} | NONE |
ClickHouse | ClickHouse | 793,572,264 | 21,406 | hexiaoting | @vzakaznikov Now we support:
```
SELECT map(toFixedString('1',1),1) AS m, m[toFixedString('1',1)]];
```
But not support :
```
SELECT map(toFixedString('1',1),1) AS m, m['1']];
```
It seems there is a bug with fixedString equal check:
```
dell123 :) select toFixedString('1',3) == '1';
SELECT toFixedStri... | 2021-03-09T08:53:57 | 2021-03-09T08:53:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,043,080 | 21,089 | den-crane | @cw9
I am confused. What version do you use right now? How did you configure `background_fetches_pool_size` , value? | 2021-03-09T15:25:21 | 2021-03-09T15:25:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,072,307 | 21,526 | azat | > [AST fuzzer (MSan) — Fuzzer failed (137). See the logs.](https://clickhouse-test-reports.s3.yandex.net/21526/a9313c9eb9acb8920e3a99452ec14c0a116922d5/fuzzer_msan/report.html#fail1)
```
==202==ERROR: MemorySanitizer: allocator is out of memory trying to allocate 0x18 bytes
```
>Integration tests (asan) — fail:... | 2021-03-08T20:58:27 | 2021-03-08T20:58:27 | {} | MEMBER |
ClickHouse | ClickHouse | 792,561,426 | 21,112 | khlystov | @kssenii also new error during data inserts
2021.03.08 09:08:31.734606 [ 3465 ] {b5cbe893-ec57-4d6c-92cd-1d565974a680} <Error> DynamicQueryHandler: Code: 530, e.displayText() = DB::Exception: Cannot connect to RabbitMQ host: 134.00.105.54, port: 5672: while write prefix to view rabbitmq.sbgi_input_mv (3c38740e-19f... | 2021-03-08T08:09:31 | 2021-03-09T07:12:33 | {} | NONE |
ClickHouse | ClickHouse | 793,552,298 | 21,487 | alesapin | Not related to changes. | 2021-03-09T08:38:03 | 2021-03-09T08:38:03 | {} | MEMBER |
ClickHouse | ClickHouse | 793,620,180 | 21,398 | alesapin | Too many changes for the LTS branch. | 2021-03-09T09:33:37 | 2021-03-09T09:33:37 | {} | MEMBER |
ClickHouse | ClickHouse | 793,703,069 | 15,090 | alexey-milovidov | Yes. | 2021-03-09T10:47:20 | 2021-03-09T10:47:20 | {} | MEMBER |
ClickHouse | ClickHouse | 793,937,297 | 21,556 | UnamedRus | Processed 10.88 million rows, 3.16 GB
Processed 10.88 million rows, 304.95 MB
SelectedBytes | 305 029 428.00 | 3 155 152 423.00 | -2850122995.00
Seek | 166.00 | 417.00 | -251.00
can you run both queries with ```set send_logs_level='trace';``` ? | 2021-03-09T14:01:31 | 2021-03-09T14:01:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,056,328 | 21,425 | alesapin | No related failures. Let's merge and see what will happen.
fuzzer:
```
$ grep -a 'Fatal' server.log
2021.03.09 12:48:09.071765 [ 82 ] {} <Fatal> Application: Child process was terminated by signal 6.
```
And that is all.... | 2021-03-09T15:35:36 | 2021-03-09T15:37:18 | {} | MEMBER |
ClickHouse | ClickHouse | 794,126,887 | 21,089 | cw9 | ClickHouse server version 20.12.4 revision 54442
```
┌─name─────────────────────────┬─value─┬─changed─┬─description────────────────────────────────────────────────────────────────────────────────────────────────┬─min──┬─max──┬─readonly─┬─type───┐
│ background_fetches_pool_size │ 16 │ 1 │ Number of threads... | 2021-03-09T16:31:23 | 2021-03-09T16:31:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,128,212 | 21,516 | alexey-milovidov | No difference in perf tests.
But we have only limited range of different CPU models in our CI environment. | 2021-03-08T22:26:15 | 2021-03-08T22:26:15 | {} | MEMBER |
ClickHouse | ClickHouse | 793,248,521 | 18,583 | alexey-milovidov | @jart
I'm trying to evaluate various implementations of memcpy and found that the performance tests are not good enough.
We can test them just slightly better:
1. Test on various CPUs: at least several generations of Intel and latest generation of AMD.
2. Don't run a loop with identical size to copy, use random... | 2021-03-09T01:38:00 | 2021-03-09T01:41:16 | {} | MEMBER |
ClickHouse | ClickHouse | 793,278,837 | 21,414 | amosbird | Integration test errors look irrelevant. | 2021-03-09T02:23:25 | 2021-03-09T02:23:25 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 793,916,110 | 21,552 | nikitamikhaylov | What will be if you disable `input_format_parallel_parsing` setting? | 2021-03-09T13:43:32 | 2021-03-09T13:43:32 | {} | MEMBER |
ClickHouse | ClickHouse | 793,987,468 | 21,556 | OmarBazaraa | I noticed the `MergeTreeWhereOptimizer` in the newest version moved less conditions than the older version.
By manually providing the `PREWHERE` clause, the query performance become better. But I am not sure if this is the only problem in the newest version. | 2021-03-09T14:42:50 | 2021-03-09T14:42:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 792,769,312 | 21,504 | den-crane | But your cluster description **remote_servers** is completely wrong.
You have one-shard one-replica at each CH node.
Here I see each server twice:
```
┌─clu─┬─shard_num─┬─shard_weight─┬─replica_num─┬─host_name────────┬─host_address─┬─port─┬─is_local─┬─user────┬─default_database─┬─errors_count─┬─estimated_recove... | 2021-03-08T13:51:02 | 2021-03-08T13:51:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 792,828,129 | 21,502 | den-crane | >But, most of my queries use the time field
Then try to use my first suggestion `ORDER BY (toDate(fromUnixTimestamp64Milli(time)), field1, field2) `
Place time into leading position but reduce cardinality (of index column) to one day.
>I cannot leave all of those fields as primary keys.
Why?
>And I use PK... | 2021-03-08T15:22:57 | 2021-03-08T15:22:57 | {
"heart": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,459,093 | 21,478 | DimasKovas | > You can try dumping out the `column_names` here https://github.com/ClickHouse/ClickHouse/blob/master/src/Storages/StorageMergeTree.cpp#L178ckHouse/blob/master/src/Storages/StorageMergeTree.cpp#L178 . Columns listed in query_info might not be the actual access ones.
`column_names` also contains both columns. Here i... | 2021-03-09T06:43:18 | 2021-03-09T06:43:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 793,582,686 | 21,410 | alesapin | No related failures. | 2021-03-09T09:02:36 | 2021-03-09T09:02:36 | {} | MEMBER |
ClickHouse | ClickHouse | 793,903,257 | 4,959 | stufently | > Any solution for this, guys ? I have same problem in latest release clickhouse-server 21.2.5.5
Try change server, update clickhouse dont help with this issue. | 2021-03-09T13:32:34 | 2021-03-09T13:32:34 | {} | NONE |
ClickHouse | ClickHouse | 800,796,509 | 21,781 | filimonov | Did you consider extracting data from `system.tables`? | 2021-03-17T05:04:08 | 2021-03-17T05:05:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 800,105,781 | 21,773 | Teeeto | The table DDL is
```
CREATE TABLE XXX
(
'SessionId' Int64 CODEC(Delta(8), LZ4HC(0)),
'SUBSCRIBERID' String CODEC(ZSTD(22)),
'LastResponse' Nullable(String) CODEC(ZSTD(22)),
'LastResponseDateTime' Nullable(DateTime64(3)) CODEC(Delta(8), LZ4HC(0)),
'LastRequest' Nullable(String) C... | 2021-03-16T09:38:45 | 2021-03-17T09:44:17 | {} | NONE |
ClickHouse | ClickHouse | 801,077,077 | 21,831 | jorisgio | ```
#0 futex_wait_cancelable (private=0, expected=0, futex_word=0x7f81cfd91e70) at ../sysdeps/unix/sysv/linux/futex-internal.h:88
#1 __pthread_cond_wait_common (abstime=0x0, mutex=0x7f81cfd91e20, cond=0x7f81cfd91e48) at pthread_cond_wait.c:502
#2 __pthread_cond_wait (cond=0x7f81cfd91e48, mutex=0x7f81cfd91e20) at ... | 2021-03-17T13:24:19 | 2021-03-17T13:24:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 801,136,018 | 21,813 | alexey-milovidov | Yes, the `zOrder` function is fully implemented in #6286
The only remaining part is to finish KeyCondition analysis, that is about 50% implemented. | 2021-03-17T14:36:28 | 2021-03-17T14:36:28 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.