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 | 735,603,558 | 17,144 | troll2016 | > Changelog category (leave one):
>
> * New Feature
>
>
> Changelog entry (a user-readable short description of the changes that goes to CHANGELOG.md):
> add `*.zst` compression/decompression support.It enables using `*.zst` in `file()` function and `Content-encoding: zstd` in http client.This closes [#167... | 2020-11-30T07:23:20 | 2020-11-30T07:23:20 | {} | NONE |
ClickHouse | ClickHouse | 735,690,324 | 17,542 | alexey-milovidov | Yes. | 2020-11-30T10:13:09 | 2020-11-30T10:13:09 | {} | MEMBER |
ClickHouse | ClickHouse | 735,749,295 | 16,937 | CurtizJ | Test failures are not related to changes. | 2020-11-30T12:12:47 | 2020-11-30T12:12:47 | {} | MEMBER |
ClickHouse | ClickHouse | 735,761,193 | 10,637 | martinitus | It seems to work for me, however the created user seems to lack some privileges. When trying to connect with DataGrip I get the following.
``DB::Exception: username: Not enough privileges. To execute this query it's necessary to have the grant SHOW USERS ON *.* (version 20.11.4.13 (official build))`` | 2020-11-30T12:37:38 | 2020-11-30T12:37:38 | {} | NONE |
ClickHouse | ClickHouse | 735,826,686 | 17,387 | DmitriZhukov | я не понимаю почему у меня в клиенте DataGrip время отличается от Москвы на 5 часов
сейчас в МСК 17:41
**select now()**
2020-11-30 12:41:19 | 2020-11-30T14:41:29 | 2020-11-30T14:42:04 | {} | NONE |
ClickHouse | ClickHouse | 735,892,900 | 17,631 | den-crane | And upgrade to 20.10.5.10 | 2020-11-30T16:27:24 | 2020-11-30T16:27:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,915,345 | 17,632 | den-crane | > Gorilla on LowCardinality column is an extremely weird idea.
Earlier CH allowed unusable CODECS. I see the third report about upgrade problem because of that.
It's easy to fix, need to ALTER CODEC before an upgrade.
The real issue that it WAS? impossible to ALTER CODEC for a column with SKIP index. Need to DRO... | 2020-11-30T17:04:37 | 2020-11-30T17:33:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,625,369 | 17,578 | l1t1 | isn't Ryu good enough for clickhouse? | 2020-11-30T08:08:17 | 2020-11-30T08:08:17 | {} | NONE |
ClickHouse | ClickHouse | 735,646,119 | 17,606 | alesapin | hmmm... and these are all places where API differs? | 2020-11-30T08:49:51 | 2020-11-30T08:49:51 | {} | MEMBER |
ClickHouse | ClickHouse | 735,702,402 | 17,632 | alexey-milovidov | This is unlikely to be fixed. | 2020-11-30T10:35:31 | 2020-11-30T10:35:31 | {} | MEMBER |
ClickHouse | ClickHouse | 735,774,169 | 17,632 | siradjev | > This is unlikely to be fixed.
Recreating index on >100TB data is a very long-time consuming operation, making "no-impact" upgrade from 20.3 to 20.8 on live large cluster virtually impossible.
We'll either have to slow down select queries for the duration of skip-index creation, or arrange downtime on cluster to ... | 2020-11-30T13:06:26 | 2020-11-30T13:08:21 | {} | NONE |
ClickHouse | ClickHouse | 735,908,473 | 17,632 | filimonov | Hard to follow the sequence.
Did you get that problem during the update to 20.8 or during `alter table test_table modify column index_column String CODEC(ZSTD(1)) `
The alter was still running during update? Were you doing those alters on 20.3 or on 20.8?
Can you explain the scenario step by step:
1) Initial ... | 2020-11-30T16:53:38 | 2020-11-30T16:55:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,465,157 | 17,596 | hodgesrm | I don't see how to address the missing 'can be tested' label. I don't seem to have access to the build output in the even that this was generated by the build system. | 2020-11-29T22:03:33 | 2020-11-29T22:03:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,469,925 | 17,585 | filimonov | I guess that will end up with single node processing 2 shards data due to prefer localhost replica (expected is to pick another node if it's avaliable). | 2020-11-29T22:45:39 | 2020-11-29T22:46:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,503,893 | 16,995 | sundy-li | The fail tests are irrelevant.
@azat @alexey-milovidov could you help review this PR? | 2020-11-30T02:03:28 | 2020-11-30T02:03:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,506,757 | 17,215 | coreyralph | Those tables are 44M and 25M rows respectively. The initial queries take ~200s. I increased the timeout in the ODBC gateway to allow that. They have several other columns, the dictionary is only fetching the needed subset.
The updates are much smaller. In the past 24h it's 10k on dmn and 380k on rpm. A mixture of in... | 2020-11-30T02:16:41 | 2020-11-30T02:16:41 | {} | NONE |
ClickHouse | ClickHouse | 735,640,748 | 17,612 | ilejn | > So, we have to build multiple hash tables to support OR?
Exactly.
One per disjunct. | 2020-11-30T08:38:42 | 2020-11-30T08:50:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,687,318 | 17,542 | nvartolomei | Should this be backported to 20.8 lts as well? | 2020-11-30T10:07:46 | 2020-11-30T10:07:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,820,615 | 17,603 | vitlibar | Cherry-picked manually | 2020-11-30T14:31:18 | 2020-11-30T14:31:18 | {} | MEMBER |
ClickHouse | ClickHouse | 735,895,390 | 17,638 | DmitriZhukov | date,Date,"","","","",""
datetime,DateTime,"","","","",""
timestamp,Int64,"","","","",""
id,String,"","","","",""
pair,String,"","","","",""
status,String,"","","","",""
direction,String,"","","","",""
maker_order,Nullable(String),"","","","",""
taker_order,Nullable(String),"","","","",""
maker_trader,Nullable... | 2020-11-30T16:31:25 | 2020-11-30T16:31:25 | {} | NONE |
ClickHouse | ClickHouse | 735,467,749 | 17,585 | filimonov | So i guess the fix is wrong... | 2020-11-29T22:26:52 | 2020-11-29T22:26:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,615,546 | 17,532 | alesapin | `test_https_replication/test.py::test_replication_after_partition` something new | 2020-11-30T07:49:29 | 2020-11-30T07:49:29 | {} | MEMBER |
ClickHouse | ClickHouse | 735,645,599 | 17,612 | alexey-milovidov | Ok. | 2020-11-30T08:48:44 | 2020-11-30T08:48:44 | {} | MEMBER |
ClickHouse | ClickHouse | 735,694,821 | 17,623 | l1t1 | nice job! **Window functions** | 2020-11-30T10:21:36 | 2020-11-30T10:21:36 | {
"+1": 24
} | NONE |
ClickHouse | ClickHouse | 735,784,164 | 17,250 | alexey-milovidov | @nautaa Yes, it will be very helpful! Implementation will be similar to `writeXMLString` but with more characters to encode, because we will encode not only for text elements but also for attributes, so we need to encode `"` and `'` additionally. | 2020-11-30T13:26:06 | 2020-11-30T13:26:26 | {} | MEMBER |
ClickHouse | ClickHouse | 735,858,398 | 17,638 | den-crane | Maybe you have changed "VM options" ?
| 2020-11-30T15:31:46 | 2020-11-30T15:31:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,859,280 | 17,387 | den-crane | а что в `select now(), toString(now())` | 2020-11-30T15:33:11 | 2020-11-30T15:33:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,937,000 | 17,154 | vitlibar | The conflict doesn't seem to be easy to resolve. | 2020-11-30T17:43:18 | 2020-11-30T17:43:18 | {} | MEMBER |
ClickHouse | ClickHouse | 735,433,979 | 17,557 | alexey-milovidov | Functional tests Ok. | 2020-11-29T18:15:55 | 2020-11-29T18:15:55 | {} | MEMBER |
ClickHouse | ClickHouse | 735,583,007 | 17,547 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/pull/17597 | 2020-11-30T06:37:34 | 2020-11-30T06:37:34 | {} | MEMBER |
ClickHouse | ClickHouse | 735,654,884 | 17,446 | RonanMorgan | Hello,
I made a mistake in my description, it seems that there are errors server side for those "Broken Pipe" requests but it's written
"The Transfer-Encoding is not chunked and there is no Content-Length header for POST"
I didn't make the link with my previous [issue](https://github.com/ClickHouse/ClickHouse/i... | 2020-11-30T09:06:21 | 2020-11-30T09:06:21 | {} | NONE |
ClickHouse | ClickHouse | 735,892,830 | 17,605 | vitlibar | Cherry-picked manually | 2020-11-30T16:27:16 | 2020-11-30T16:27:16 | {} | MEMBER |
ClickHouse | ClickHouse | 735,542,017 | 16,457 | moweonlee | @den-crane
Oh I guess arrayJoin( users ) still need similar workload with source table because uniq() still need to process every row like what uniq() do in the source table .
But can it be said that uniq(arrayJoin( users ) ) need less memory than uniq(users) in source table which doesn't me merged at all. ? | 2020-11-30T04:47:41 | 2020-11-30T04:47:55 | {} | NONE |
ClickHouse | ClickHouse | 735,557,925 | 11,714 | ucasfl | I found I totally misunderstand this task, I thought we need to use two variables to track total bytes and total rows of `Committed` parts in each MergeTree table, so when invoke `totalRows` and `totalBytes`, just return these two variables value, like [https://github.com/ClickHouse/ClickHouse/blob/master/src/Storages/... | 2020-11-30T05:36:28 | 2020-11-30T05:36:28 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 735,579,386 | 17,569 | alexey-milovidov | Shouldn't be any drawbacks. The possible drawback is extra cache consumption. But we process data in tight loops, so possible cache misses will be "amortized" across the loop. | 2020-11-30T06:29:25 | 2020-11-30T06:29:25 | {} | MEMBER |
ClickHouse | ClickHouse | 735,582,079 | 17,542 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/pull/17597 | 2020-11-30T06:35:29 | 2020-11-30T06:35:29 | {} | MEMBER |
ClickHouse | ClickHouse | 735,582,719 | 17,544 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/pull/17597 | 2020-11-30T06:36:55 | 2020-11-30T06:36:55 | {} | MEMBER |
ClickHouse | ClickHouse | 735,466,125 | 17,585 | azat | >Also, replica is considered non-local, if it has default database set
> (only reason is to avoid query rewrite).
Ok, I see
- subqueries are affected too
- Distributed can accept empty database (didn't knew that), which means that use `default_database` from the cluster definition
But seems that rewrite can ... | 2020-11-29T22:12:23 | 2020-11-29T22:13:01 | {} | MEMBER |
ClickHouse | ClickHouse | 735,582,858 | 17,546 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/pull/17597 | 2020-11-30T06:37:15 | 2020-11-30T06:37:15 | {} | MEMBER |
ClickHouse | ClickHouse | 735,603,568 | 1,562 | 1332498780 | hi @CrystalCat , i encounter the similar problem, i resolved with the ways of @filimonov . but i find this way on the bellow also can resolved this problem .
```
select
case when (total<>0) then round( ((total - b) / case when total <> 0 then total else 1 end)*100, 2) else 0 end per
from demo
;
```
result... | 2020-11-30T07:23:22 | 2020-11-30T07:23:22 | {} | NONE |
ClickHouse | ClickHouse | 735,627,430 | 17,578 | alexey-milovidov | It's good enough, but in ClickHouse there is a principle to choose the best available option for performance. | 2020-11-30T08:11:52 | 2020-11-30T08:11:52 | {} | MEMBER |
ClickHouse | ClickHouse | 735,656,476 | 8,939 | KochetovNicolai | I did not change anything in StorageMerge | 2020-11-30T09:09:16 | 2020-11-30T09:09:16 | {} | MEMBER |
ClickHouse | ClickHouse | 735,656,499 | 17,613 | thyn | is it possible to apply that fix to 20.11? | 2020-11-30T09:09:19 | 2020-11-30T09:09:19 | {} | NONE |
ClickHouse | ClickHouse | 735,852,912 | 17,510 | vitlibar | The conflict doesn't seem to be easy to resolve. | 2020-11-30T15:23:49 | 2020-11-30T15:23:49 | {} | MEMBER |
ClickHouse | ClickHouse | 735,814,727 | 17,623 | filimonov | * smth about multilevel Merge-tree partitioning
* dedup on MergeTree (block level)
etcd vs embedded Raft - questionable if it's worth to have both (too many different stuff to support/maintain, I would prefer to go to embedded Raft only + may be some tooling around instead - i.e. a way to access the Raft state even... | 2020-11-30T14:21:25 | 2020-11-30T15:46:54 | {
"+1": 8
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,627,292 | 17,612 | ilejn | I am working on it.
Started from Hash joins.
Suggestions are highly appreciated. | 2020-11-30T08:11:36 | 2020-11-30T08:11:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,678,771 | 17,585 | alexey-milovidov | Preferring localhost replicas and interaction with localhost replicas without IPC are two different things. | 2020-11-30T09:52:15 | 2020-11-30T09:52:15 | {} | MEMBER |
ClickHouse | ClickHouse | 735,808,270 | 16,947 | akuzm | Is this problem related to the fact that we match remote columns by name? This is not the first one, we probably should match them by index, like for `UNION`. | 2020-11-30T14:09:55 | 2020-11-30T14:09:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,433,934 | 17,541 | alexey-milovidov | Functional tests Ok. | 2020-11-29T18:15:32 | 2020-11-29T18:15:32 | {} | MEMBER |
ClickHouse | ClickHouse | 735,434,153 | 17,562 | alexey-milovidov | Functional tests Ok. | 2020-11-29T18:17:17 | 2020-11-29T18:17:17 | {} | MEMBER |
ClickHouse | ClickHouse | 735,525,348 | 16,176 | hexiaoting | @alexey-milovidov
I'm trying to implement this feature on Interpreterxxx part, But I find a small problem :
I cannot judge wether this is a direct SELECT query or a subquery in other Queries, So right now If I add LimitTransform directly in InterpterSelectWithUnionQuery, it will make the result wrong.
Like :
```
... | 2020-11-30T03:35:22 | 2020-11-30T06:09:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,620,918 | 17,403 | audricbizouarn | ```
time curl 'http://user:password@srvA:8123/?query=SELECT%201'
1
real 0m5.775s
user 0m0.013s
sys 0m0.000s
```
```
2020.11.30 08:35:18.565320 [ 24498 ] {} <Trace> HTTPHandler-factory: HTTP Request for HTTPHandler-factory. Method: GET, Address: [::ffff:ip]:49338, User-Agent: curl/7.35.0, Content Type: , Tra... | 2020-11-30T08:00:04 | 2020-11-30T08:00:04 | {} | NONE |
ClickHouse | ClickHouse | 735,647,369 | 17,606 | alexey-milovidov | Also there are a few missing required functions in `FunctionsAES` - it will be rewritten separately.
And there are no more incompatible places. | 2020-11-30T08:52:17 | 2020-11-30T08:52:17 | {} | MEMBER |
ClickHouse | ClickHouse | 735,672,345 | 17,613 | alexey-milovidov | Yes, it will be backported. | 2020-11-30T09:39:36 | 2020-11-30T09:39:36 | {} | MEMBER |
ClickHouse | ClickHouse | 735,674,398 | 17,585 | filimonov | > That's true, but we can also disable `prefer_localhost_replica`.
But if you say that the perf difference is caused by lack of those optimizations for localhost, then disabling 'prefer_localhost_replica' will also end up with same perf degradation? Or no?
BTW it's better to recheck the problem on cluster, as ori... | 2020-11-30T09:43:34 | 2020-11-30T09:43:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,693,506 | 17,109 | nvartolomei | Slightly off-topic: Why/when would you use a SimpleAggregateFunction as part of the primary key? | 2020-11-30T10:19:14 | 2020-11-30T10:19:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,856,588 | 17,474 | vitlibar | The conflict doesn't seem to be easy to resolve. | 2020-11-30T15:29:11 | 2020-11-30T15:29:11 | {} | MEMBER |
ClickHouse | ClickHouse | 735,536,094 | 14,150 | excitoon | @azat I switched default number of threads back to 1 but it is not very clear for me why in some cases we can use separate threads, for example here: https://github.com/ClickHouse/ClickHouse/blob/b94cc5c4e53c3d275d287e0b6b71e37d9421d339/src/Storages/MergeTree/MergeTreeDataSelectExecutor.cpp#L685 . Can you also clarify ... | 2020-11-30T04:22:46 | 2020-11-30T04:23:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,574,170 | 17,585 | alexey-milovidov | @azat FYI, here is the first commit:
https://github.com/ClickHouse/ClickHouse/commit/de71157b10199020d725bca94ec99fcf40a400d8#diff-749f8d00fc58896f2215a5d08e450a768ecee040eb36e6dbbeb379c5366c6d0dR103
You can use `git log -S` to find it. Or, better: run `clickhouse-git-import`. | 2020-11-30T06:17:32 | 2020-11-30T06:17:32 | {} | MEMBER |
ClickHouse | ClickHouse | 735,575,545 | 17,585 | alexey-milovidov | > I guess that will end up with single node processing 2 shards data due to prefer localhost replica (expected is to pick another node if it's avaliable).
That's true, but we can also disable `prefer_localhost_replica`. | 2020-11-30T06:20:43 | 2020-11-30T06:20:43 | {} | MEMBER |
ClickHouse | ClickHouse | 735,630,518 | 17,612 | alexey-milovidov | So, we have to build multiple hash tables to support OR? | 2020-11-30T08:17:49 | 2020-11-30T08:17:49 | {} | MEMBER |
ClickHouse | ClickHouse | 735,706,043 | 17,512 | vitlibar | The conflict doesn't seem to be easy to resolve. | 2020-11-30T10:42:32 | 2020-11-30T10:42:32 | {} | MEMBER |
ClickHouse | ClickHouse | 735,883,580 | 17,638 | DmitriZhukov | No I dont have this setting
```
select datetime, toStartOfMinute(datetime) as minuteTime,
from organic_trades where
datetime >= '2020-11-30 00:00:00' and
datetime <= '2020-11-30 15:38:00' an... | 2020-11-30T16:12:17 | 2020-11-30T16:22:32 | {} | NONE |
ClickHouse | ClickHouse | 735,892,556 | 17,631 | den-crane | it's another parameters, server parameters (config.xml) **max_server_memory_usage_to_ram_ratio** **max_server_memory_usage**
https://clickhouse.tech/docs/en/operations/server-configuration-parameters/settings/#max_server_memory_usage
https://clickhouse.tech/docs/en/operations/server-configuration-parameters/setti... | 2020-11-30T16:26:47 | 2020-11-30T16:26:47 | {
"+1": 8
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,434,046 | 17,561 | alexey-milovidov | Functional tests Ok. | 2020-11-29T18:16:35 | 2020-11-29T18:16:35 | {} | MEMBER |
ClickHouse | ClickHouse | 735,571,641 | 16,176 | alexey-milovidov | @hexiaoting Let's check if it's ASTSelectQuery or ASTSelectWithUnionQuery. In the ASTSelectWithUnionQuery, if it contains a single SELECT without WITH TIES, WITH FILL, adjust limits there, otherwise add transform over the result in InterpreterSelectWithUnionQuery. | 2020-11-30T06:11:15 | 2020-11-30T06:11:15 | {} | MEMBER |
ClickHouse | ClickHouse | 735,582,456 | 17,545 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/pull/17597 | 2020-11-30T06:36:18 | 2020-11-30T06:36:18 | {} | MEMBER |
ClickHouse | ClickHouse | 735,607,248 | 17,535 | alesapin | `toomanyrequests: You have reached your pull rate limit. You may increase the limit by authenticating and upgrading: https://www.docker.com/increase-rate-limit` | 2020-11-30T07:31:34 | 2020-11-30T07:31:34 | {} | MEMBER |
ClickHouse | ClickHouse | 735,643,127 | 17,613 | alexey-milovidov | The bug was introduced here: https://github.com/ClickHouse/ClickHouse/pull/16555/files#diff-60ac1b8c4ea5fc12319121f5bc970ed065b4cffd2ffd16e81f815ae80a75386eR1108
CC @tavplubix | 2020-11-30T08:43:51 | 2020-11-30T08:43:51 | {} | MEMBER |
ClickHouse | ClickHouse | 735,580,710 | 17,566 | alexey-milovidov | > 01455_opentelemetry_distributed
\- flaky test, @akuzm | 2020-11-30T06:32:21 | 2020-11-30T06:32:21 | {} | MEMBER |
ClickHouse | ClickHouse | 735,584,775 | 17,525 | alexey-milovidov | > 01455_opentelemetry_distributed
\- bad test, @akuzm | 2020-11-30T06:41:20 | 2020-11-30T06:41:20 | {} | MEMBER |
ClickHouse | ClickHouse | 735,609,883 | 17,533 | alesapin | https://gist.github.com/alesapin/bb456c1babd38696f6b3bd3ba2712736 RocksDB crashed again... | 2020-11-30T07:37:32 | 2020-11-30T07:37:32 | {} | MEMBER |
ClickHouse | ClickHouse | 735,616,089 | 17,499 | alesapin | `01317_no_password_in_command_line` flaky
`test_reload_zookeeper/test.py::test_reload_zookeeper` fixed in master. | 2020-11-30T07:50:32 | 2020-11-30T07:50:32 | {} | MEMBER |
ClickHouse | ClickHouse | 735,468,076 | 17,585 | azat | >Single clickhouse instanse serving 2 shards data.
Yes, this is also an option.
But everything should work if query will be rewritten (and only for the local shard) with taking default_database into account. | 2020-11-29T22:29:40 | 2020-11-29T22:29:40 | {} | MEMBER |
ClickHouse | ClickHouse | 735,543,416 | 16,457 | moweonlee | > Third time I am looking at this issue and I don't want to retype pictures. Why did you post pictures?
Oh I guess picture would be more easy to understand.
I promise there will be paste-able text from next time. | 2020-11-30T04:53:18 | 2020-11-30T04:53:55 | {} | NONE |
ClickHouse | ClickHouse | 735,570,100 | 11,714 | alexey-milovidov | That's the most of this task. Then we should add metrics on total data volume into AsynchronousMetrics.
I think you have understand everything right. | 2020-11-30T06:07:28 | 2020-11-30T06:07:28 | {} | MEMBER |
ClickHouse | ClickHouse | 735,582,293 | 17,548 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/pull/17597 | 2020-11-30T06:35:56 | 2020-11-30T06:35:56 | {} | MEMBER |
ClickHouse | ClickHouse | 735,608,289 | 17,534 | alesapin | docker: Error response from daemon: toomanyrequests: You have reached your pull rate limit. You may increase the limit by authenticating and upgrading: https://www.docker.com/increase-rate-limit. | 2020-11-30T07:33:52 | 2020-11-30T07:33:52 | {} | MEMBER |
ClickHouse | ClickHouse | 735,641,697 | 17,614 | TCeason | @tavplubix Please add a `can be tested` label. | 2020-11-30T08:40:44 | 2020-11-30T08:40:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,898,876 | 17,438 | spongedu | @akuzm Thanks for your review. I've made some updates according to our discussion. PTAL. | 2020-11-30T16:37:22 | 2020-11-30T16:37:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,913,880 | 17,537 | filimonov | @elfring may be you can collaborate and just send a PR with fixes? | 2020-11-30T17:02:22 | 2020-11-30T17:02:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,920,366 | 17,559 | vitlibar | Depends on https://github.com/ClickHouse/ClickHouse/pull/15074 which hasn't merged to `20.10`. | 2020-11-30T17:13:11 | 2020-11-30T17:19:41 | {} | MEMBER |
ClickHouse | ClickHouse | 735,933,662 | 17,630 | alexey-milovidov | We also need something to suppress TSan. | 2020-11-30T17:37:01 | 2020-11-30T17:37:01 | {} | MEMBER |
ClickHouse | ClickHouse | 735,467,559 | 17,585 | filimonov | > when multiple servers may run on the same host they will have different port,
That shound you misunderstood the idea of the way how that 'cross-replication' were designed (i personally not a big fan of that).
Single clickhouse instanse serving 2 shards data.
I.e.
node1 has databases: shard1, shard2
node2 ... | 2020-11-29T22:25:22 | 2020-11-29T22:25:22 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,468,173 | 14,099 | filimonov | AFAIR it was originally observed on a cluster with prefer_localhost=0, and definitely not on localhost-only queries (5 shards, 2 replicas arair).
So i'm not sure it it's the correct reason, but i had no chances to recheck it by my own.
The patch looks wrong (removed logic was there by intent). | 2020-11-29T22:30:19 | 2020-11-29T22:31:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,669,890 | 17,622 | TCeason | @tavplubix Please add a can be tested label. | 2020-11-30T09:34:50 | 2020-11-30T09:34:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,704,992 | 17,630 | alexey-milovidov | Simply revert this commit: https://github.com/ClickHouse/ClickHouse/pull/8344/commits/bdefa9248c2beeb4527cc2e1bcc0948824bf53cb | 2020-11-30T10:40:29 | 2020-11-30T10:40:29 | {} | MEMBER |
ClickHouse | ClickHouse | 735,731,414 | 17,591 | CurtizJ | docker: Error response from daemon: toomanyrequests: You have reached your pull rate limit. You may increase the limit by authenticating and upgrading: https://www.docker.com/increase-rate-limit. | 2020-11-30T11:33:25 | 2020-11-30T11:33:25 | {} | MEMBER |
ClickHouse | ClickHouse | 735,734,828 | 17,250 | nautaa | hello,i want to complete encodeXMLComponent and decodeXMLComponent. Are they a function format similar to function writeXMLString() in WriteHelpers.h? Put the string escape result into the writerbuffer? | 2020-11-30T11:41:22 | 2020-11-30T11:41:22 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,918,165 | 17,632 | siradjev | >The real issue that WAS
Issue is you have to modify index on 20.3. | 2020-11-30T17:09:16 | 2020-11-30T17:09:16 | {} | NONE |
ClickHouse | ClickHouse | 735,485,196 | 17,568 | azat | All ttl move tests are flaky, requires more generic approach | 2020-11-30T00:33:59 | 2020-11-30T00:33:59 | {} | MEMBER |
ClickHouse | ClickHouse | 735,610,086 | 17,533 | alesapin | `test_storage_rabbitmq/test.py::test_rabbitmq_csv_with_delimiter` flaky test, I'll cherry-pick fix for it. | 2020-11-30T07:37:57 | 2020-11-30T07:38:16 | {} | MEMBER |
ClickHouse | ClickHouse | 735,807,085 | 17,624 | filimonov | 1) llvm is an optional component, you can disable it with CMake flag
2) you're trying to compile an obsolete version. Use 20.8 at least. | 2020-11-30T14:07:46 | 2020-11-30T14:07:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,885,663 | 17,638 | den-crane | please show the output `desc table organic_trades` | 2020-11-30T16:15:32 | 2020-11-30T16:15:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 735,444,116 | 14,099 | azat | >Now it depends on how localhost optimization is performed. If it takes both host and port into account - everything is ok.
Yes, host and port and AFAICS from the git history it was from beginning (1c86e58), well it was definitely there when default_database has been added.
Anyway, here is a patch (#17585), let's t... | 2020-11-29T19:44:47 | 2020-11-29T19:44:47 | {} | MEMBER |
ClickHouse | ClickHouse | 735,466,773 | 17,568 | azat | >Integration tests flaky check (asan) — Flaky tests found: fail: 1, passed: 56, error: 0
This is different - `test_ttl_move/test.py::test_moves_to_volume_work` | 2020-11-29T22:18:05 | 2020-11-29T22:18:05 | {} | MEMBER |
ClickHouse | ClickHouse | 735,917,044 | 17,632 | siradjev | >alter table test_table modify column index_column String CODEC(ZSTD(1))
This statement on 20.3 gives error.
> Initial schema on 20.3 (was it working?) - Gorilla on LowCardinality column is an extremely weird idea.
It was initial design on 19.x was working. Was selected after test on all possible combinatio... | 2020-11-30T17:07:22 | 2020-11-30T17:07:51 | {} | NONE |
ClickHouse | ClickHouse | 724,883,684 | 15,847 | tavplubix | 245 (0xf5) is for `MYSQL_TYPE_JSON`, it's not supported | 2020-11-10T18:27:51 | 2020-11-10T18:27:51 | {
"+1": 3
} | MEMBER |
ClickHouse | ClickHouse | 725,072,862 | 16,833 | wonsaign | > ```
> 2020.11.10 19:14:42.888949 [ 27085 ] {} MaterializeMySQLSyncThread: Skip MySQL event:
>
> === UpdateRowsEventV2 ===
> Timestamp: 1605006837
> Event Type: 31
> Server ID: 1
> Event Size: 77
> Log Pos: 6031
> Flags: 0
> [DryRun Event]
> ```
>
> This means that the current binlog event modifies data... | 2020-11-11T01:39:02 | 2020-11-11T01:39:02 | {} | NONE |
ClickHouse | ClickHouse | 725,178,329 | 1,469 | kiwimg | when the window function will be supported in Clickhouse?
I hope to see the function early! | 2020-11-11T04:10:38 | 2020-11-11T04:10:38 | {} | NONE |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.