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 | 652,927,953 | 9,759 | sundy-li | what about this query
```
INSERT INTO distributed_01099_b SELECT count(1) from distributed_01099_a;
``` | 2020-07-02T10:33:58 | 2020-07-02T10:33:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,999,038 | 12,041 | abyss7 | I'm rewriting all the contribs not to use old fuzzy and error-prone heuristics from `find/*.cmake` files. Instead I'm getting rid of them. Why do you want to bring it back? There is a new and much better way of describing external library linkage. Take a look at replxx for example. | 2020-07-02T13:15:33 | 2020-07-02T13:15:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,269,002 | 10,697 | CurtizJ | From "Yandex synchronization check" report: `cannot find source file: MergeTree/MergeTreeWriteAheadLog.cpp`
But this file is already in `ya.make`. I don't know why it failed. | 2020-07-03T00:15:54 | 2020-07-03T00:15:54 | {} | MEMBER |
ClickHouse | ClickHouse | 653,339,119 | 12,096 | gane5hvarma | confused with the wording dataset in this command, i thought it was the path of the file
```
INSERT INTO [db.]table [(c1, c2, c3)] FORMAT format_name data_set
```
sample query:
```
insert into test.omega (id,dt) format CSV
1213,2012-09-12 11:10:02
1214,2012-09-12 11:10:02
1215,2012-09-12 11:10:02
1216,2012-0... | 2020-07-03T04:28:59 | 2020-07-03T04:28:59 | {} | NONE |
ClickHouse | ClickHouse | 652,630,127 | 12,075 | filimonov | Ok, i think i understand what happened. I think it was not between poll and `read_kafka_message` (in that case next() call would just return false, 0 rows world be parsed and it would never enter `if (new_rows)`,
but stop should happened inside `read_kafka_message` itself. First call of `next()` function return the m... | 2020-07-01T20:28:06 | 2020-07-01T20:28:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,667,987 | 11,942 | filimonov | It is because of `ON CLUSTER`, those commands are executed out of the current user / session scope, by a separate thread asychroniously. That thread uses system profile (default if you did't change that) | 2020-07-01T22:02:03 | 2020-07-01T22:02:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,840,399 | 12,075 | azat | >Yes, it seems like logical error
Ok, squashed | 2020-07-02T07:37:35 | 2020-07-02T07:37:35 | {} | MEMBER |
ClickHouse | ClickHouse | 653,456,785 | 8,159 | wbigdata | Hello, have you solved the problem? | 2020-07-03T09:45:31 | 2020-07-03T09:45:31 | {} | NONE |
ClickHouse | ClickHouse | 653,544,593 | 11,913 | CurtizJ | > > codecs...
>
> Various difference.
It's unrelated, because compact parts don't support columns' special codecs. Need to force creation of wide parts in those tests.
| 2020-07-03T13:17:21 | 2020-07-03T13:31:18 | {} | MEMBER |
ClickHouse | ClickHouse | 652,096,126 | 8,090 | Perlovka | Package has been dropped from the main tree.
Here is a simple binary ebuild, if anyone seeking to use Clickhouse on Gentoo:
https://github.com/Perlovka/portage-overlay/tree/master/dev-db/clickhouse-bin | 2020-06-30T23:16:46 | 2020-06-30T23:16:46 | {
"+1": 4
} | NONE |
ClickHouse | ClickHouse | 652,477,016 | 12,072 | den-crane | @markehammons you need to upgrade to 19.15.7.30
19.15.4.10, 2019-10-31 Fixed INSERT into Distributed non local node with MATERIALIZED columns. #7377 (Azat Khuzhin)
| 2020-07-01T15:10:22 | 2020-07-01T15:10:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,926,273 | 12,041 | alesapin | Now CI is able to test docker images updates in pull requests. Let's check it. | 2020-07-02T10:30:13 | 2020-07-02T10:30:13 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 653,140,903 | 12,094 | nvartolomei | https://github.com/ClickHouse/ClickHouse/blob/8513e1ec7498ca5cb2dc162ebd642cadfc708c6e/src/Interpreters/DDLWorker.cpp#L323-L346 | 2020-07-02T17:46:02 | 2020-07-02T17:46:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,488,067 | 12,107 | alesapin | No builds :( | 2020-07-03T10:57:04 | 2020-07-03T10:57:04 | {} | MEMBER |
ClickHouse | ClickHouse | 652,075,203 | 10,994 | den-crane | @vitlibar @alexey-milovidov it's impossible to start CH 20.4+ if one have a table with Engine=Dictionary (old xml dictionary).
to start CH, it needs to remove tables metadata ( `rm /var/lib/clickhouse/metadata/default/node.sql` ) start CH, create tables. | 2020-06-30T22:16:19 | 2020-06-30T22:21:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,575,161 | 3,413 | alexey-milovidov | @blinkov You did not answer. | 2020-07-01T18:20:24 | 2020-07-01T18:20:24 | {} | MEMBER |
ClickHouse | ClickHouse | 652,827,726 | 12,072 | markehammons | @den-crane is there a version that has been pushed to dockerhub? 19.5.7.3 is the latest available on there. | 2020-07-02T07:11:05 | 2020-07-02T07:11:05 | {} | NONE |
ClickHouse | ClickHouse | 652,872,145 | 11,873 | azat | @KochetovNicolai thoughts on showing `max_threads`/`max_distributed_connections`/.. somehow?
i.e.:
```
SET max_distributed_connections = 10;
explain select * from remote('127.{1,2,3}', system.one);
Union
Expression (Projection)
ReadFromStorage x 10 (Read from Distributed)
```
There were some issu... | 2020-07-02T08:38:57 | 2020-07-02T08:38:57 | {} | MEMBER |
ClickHouse | ClickHouse | 652,922,674 | 12,067 | alesapin | `00943_materialize_index` reported in CI-fest | 2020-07-02T10:21:51 | 2020-07-02T10:21:51 | {} | MEMBER |
ClickHouse | ClickHouse | 652,928,573 | 12,056 | alesapin | Perf test failures are not related to changes. | 2020-07-02T10:35:25 | 2020-07-02T10:35:36 | {} | MEMBER |
ClickHouse | ClickHouse | 652,983,822 | 11,781 | alexey-milovidov | So we have 60 000 seconds for `initLimit`, that seems wrong.
Maybe admins just set this value to be large enough from despair. | 2020-07-02T12:44:48 | 2020-07-02T12:44:48 | {} | MEMBER |
ClickHouse | ClickHouse | 653,049,742 | 11,536 | alexey-milovidov | @lemire Yes, I will remove RapidJSON and generic interface. | 2020-07-02T14:46:19 | 2020-07-02T14:46:19 | {} | MEMBER |
ClickHouse | ClickHouse | 653,378,543 | 11,629 | jackpgao | - The crash is not always to happen with crash data dir.
- BTW, using 20.4.3.16, when restart server, it will detected the data crash and detach the dir, while the 20.3.x will not. | 2020-07-03T06:34:05 | 2020-07-03T06:34:05 | {} | NONE |
ClickHouse | ClickHouse | 652,070,898 | 11,201 | den-crane | duplicate https://github.com/ClickHouse/ClickHouse/issues/10994 | 2020-06-30T22:05:19 | 2020-06-30T22:05:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,368,373 | 12,062 | CurtizJ | Also you can use setting `force_primary_key` in tests. | 2020-07-01T11:40:36 | 2020-07-01T11:40:36 | {} | MEMBER |
ClickHouse | ClickHouse | 652,533,055 | 3,302 | blinkov | Feel free to re-open the issue if it is still relevant. | 2020-07-01T16:51:43 | 2020-07-01T16:51:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,967,845 | 12,041 | azat | >Package 'libcurl-dev' has no installation candidate it works. Also, integration tests shouldn't start.
Yes, because there is fallback - https://github.com/ClickHouse/ClickHouse/pull/12041#discussion_r447300261
Will update to `libcurl4-openssl-dev` | 2020-07-02T12:09:13 | 2020-07-02T12:09:13 | {} | MEMBER |
ClickHouse | ClickHouse | 652,982,482 | 11,781 | alexey-milovidov | > initLimit
Yes, it is the actual value, we have this value in our ZK config. | 2020-07-02T12:41:50 | 2020-07-02T12:46:43 | {} | MEMBER |
ClickHouse | ClickHouse | 652,987,683 | 11,781 | alexey-milovidov | ```
velom, [02.07.20 15:48]
[In reply to Alexey Milovidov]
мы когда то подбирали эти параметры, чтобы zk хотя бы работал
``` | 2020-07-02T12:53:13 | 2020-07-02T12:53:13 | {
"laugh": 1
} | MEMBER |
ClickHouse | ClickHouse | 653,288,330 | 9,980 | leoones | > @leoones, do you have any further questions?
When merge is running, the system stop merges table cannot stop quickly。
Is there any way to quickly stop the merge of a table ? | 2020-07-03T01:44:43 | 2020-07-03T01:44:43 | {} | NONE |
ClickHouse | ClickHouse | 652,505,958 | 12,074 | den-crane | How much memory server has?
How long rows in your table?
```
select round(sum(data_compressed_bytes)/1024/1024/1024) GBsize,
sum(rows) nrows,
round(sum(data_uncompressed_bytes)/nrows) avg_row_size_bytes ,
count() parts_count,
uniqExact(partition) partition_cou... | 2020-07-01T16:00:29 | 2020-07-01T16:01:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,717,301 | 11,536 | lemire | Note that if you use the latest simdjson (0.4), then it works without SSE support. We have a fallback kernel. It is runs pretty much everywhere. You could simplify your setup quite a bit. | 2020-07-02T00:54:31 | 2020-07-02T00:54:31 | {} | NONE |
ClickHouse | ClickHouse | 652,848,531 | 11,995 | azat | Any news on this one? (This will conflicts with #11498 ) | 2020-07-02T07:54:03 | 2020-07-02T07:54:03 | {} | MEMBER |
ClickHouse | ClickHouse | 652,855,214 | 12,052 | alesapin | `01193_metadata_loading` reported in CI-fest | 2020-07-02T08:07:03 | 2020-07-02T08:07:03 | {} | MEMBER |
ClickHouse | ClickHouse | 652,857,154 | 12,044 | qoega | ASAN stack on master(20.6)
```
2020.07.02 11:09:21.296142 [ 25921 ] {12259b75-f839-4d94-b2b4-6d6d29fb2417} <Debug> executeQuery: (from [::1]:32982) SELECT sumIfStateResample(0, 20, 1)(number, 1 = 1, number % 20) AS arr FROM numbers(200)
2020.07.02 11:09:21.296558 [ 25921 ] {12259b75-f839-4d94-b2b4-6d6d29fb2417} <Tra... | 2020-07-02T08:10:51 | 2020-07-02T08:10:51 | {} | MEMBER |
ClickHouse | ClickHouse | 652,476,233 | 12,073 | zhang2014 | Maybe the execution of 'ALTER TABLE test_table MODIFY COLUMN nested_column.x FIRST' should be rejected ? | 2020-07-01T15:09:07 | 2020-07-01T15:42:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,650,936 | 12,075 | azat | >It is not. Check that line:
It will go to the first branch https://github.com/azat/ClickHouse/blob/278061f0b20234787b631397a6fdf1e1243e3180/src/Storages/Kafka/KafkaBlockInputStream.cpp#L198 (remember `stalled_status == CONSUMER_STOPPED`, so `buffer->isStalled() == true`), although not sure this is ok in itself | 2020-07-01T21:16:44 | 2020-07-01T21:16:44 | {} | MEMBER |
ClickHouse | ClickHouse | 652,945,962 | 12,075 | azat | Should be backported to 20.5 (#11870 ) | 2020-07-02T11:17:03 | 2020-07-02T11:17:03 | {} | MEMBER |
ClickHouse | ClickHouse | 652,875,563 | 11,873 | KochetovNicolai | > thoughts on showing max_threads/max_distributed_connections/.. somehow?
Nice idea!
I don't know where to show this information though.
Maybe print it before plan, under another one option. @azat what do you think? | 2020-07-02T08:45:44 | 2020-07-02T14:37:33 | {} | MEMBER |
ClickHouse | ClickHouse | 652,581,456 | 12,076 | filimonov | > What IO schedulers support priority class?
CFQ only afaik.
> ionice is set automatically with nice - maybe we don't need a separate setting.
That's why '[experiment]'.
> ionice is set automatically with nice
is true inside 'best_effort' io_priority_class. You can't use 'realtime' & 'idle' by adjust... | 2020-07-01T18:34:41 | 2020-07-01T18:34:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,642,264 | 12,075 | filimonov | > Yes, this should work, but looks messy (too much flags in that reader), how about not checking stop flag in the ReadBufferFromKafkaConsumer::nextImpl?
+1. Lets do it. I had the same thoughts.
> (NOTE: that consumerStopped() is still required, since otherwise it will retry poll)
It is not. Check that line:
... | 2020-07-01T20:56:06 | 2020-07-01T20:56:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,854,612 | 11,815 | alesapin | > test_adaptive_granularity/test.py::test_version_update_two_nodes
>
> https://clickhouse-test-reports.s3.yandex.net/11974/2a1937625b0abf0ed8244c90284af15b57e8814e/integration_tests_(asan).html
Resolved in docker images. | 2020-07-02T08:05:55 | 2020-07-02T08:05:55 | {} | MEMBER |
ClickHouse | ClickHouse | 653,446,332 | 10,668 | filimonov | Fixed in 20.3.9.
Most probably by this PR https://github.com/ClickHouse/ClickHouse/pull/10418/
Will add a test for that. | 2020-07-03T09:21:03 | 2020-07-03T09:21:03 | {
"hooray": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,529,329 | 12,101 | blinkov | @alexey-milovidov no, security issues are not supposed to be published before fixes. | 2020-07-03T12:40:45 | 2020-07-03T12:40:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 651,956,387 | 9,032 | alexey-milovidov | Ok, thank you! | 2020-06-30T18:08:39 | 2020-06-30T18:08:39 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 652,157,451 | 11,887 | den-crane | No any re-balance. Because nothing to re-balance.
Data is not inserted until node is alive.
>Ho much resources does this re-insert requires?
There is no additional work.
>Does this unbalanced shards affects the query performance?
No unbalanced shards. Nowhere to affect.
Probably you still don't under... | 2020-07-01T02:56:18 | 2020-07-01T02:56:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,681,485 | 12,079 | kononencheg | There is a complex case of cost optimization in a cloud - I have:
- table that stored on net filesystem
- one writer clickhouse and multiple readers
And that’s the case - readers don’t “understand” that data is changed until detach/attach will proceed. | 2020-07-01T22:42:52 | 2020-07-01T22:42:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,964,175 | 12,052 | alesapin | Seems ok. Other tests not important. | 2020-07-02T12:01:02 | 2020-07-02T12:01:02 | {} | MEMBER |
ClickHouse | ClickHouse | 653,132,031 | 7,281 | thyn | Problem is not repeating | 2020-07-02T17:25:18 | 2020-07-02T17:25:18 | {} | NONE |
ClickHouse | ClickHouse | 651,951,690 | 12,065 | filimonov | Please use search first
Duplicate of #7690 | 2020-06-30T17:58:41 | 2020-07-01T08:05:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,578,155 | 12,076 | alexey-milovidov | What IO schedulers support priority class? | 2020-07-01T18:27:11 | 2020-07-01T18:27:11 | {} | MEMBER |
ClickHouse | ClickHouse | 652,665,792 | 11,978 | filimonov | No feedback. | 2020-07-01T21:55:37 | 2020-07-01T21:55:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,662,047 | 12,075 | filimonov | Strange, and seems it has no sence. (next f branch is dead in that case).
The intent was different, i can recheck that tomorrow or you can just try to switch the order of if else to check dataunusable first. | 2020-07-01T21:45:03 | 2020-07-01T22:04:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,747,047 | 8,004 | Slach | look like bug have same reason as https://github.com/ClickHouse/ClickHouse/issues/10397 | 2020-07-02T02:43:23 | 2020-07-02T02:43:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,922,582 | 11,815 | alesapin | `00943_materialize_index`
https://clickhouse-test-reports.s3.yandex.net/12067/a541683efa81cd07d2bf6c27cdb65c5a9cc6b82b/functional_stateless_tests_(unbundled).html#fail1 | 2020-07-02T10:21:37 | 2020-07-02T10:21:37 | {} | MEMBER |
ClickHouse | ClickHouse | 652,930,550 | 12,055 | alesapin | `00965_live_view_watch_heartbeat` reported in CI fest | 2020-07-02T10:40:01 | 2020-07-02T10:40:01 | {} | MEMBER |
ClickHouse | ClickHouse | 652,965,886 | 12,041 | alesapin | `Package 'libcurl-dev' has no installation candidate` it works. Also, integration tests shouldn't start. | 2020-07-02T12:05:12 | 2020-07-02T12:05:48 | {} | MEMBER |
ClickHouse | ClickHouse | 652,981,738 | 11,781 | alexey-milovidov | > maxSessionTimeout
It is the maximum value that client may request and the server will accept.
It is Ok to have high `maxSessionTimeout` on server to allow clients to work with high session timeout if they want.
But we request session timeout of 30 seconds by default (you can change it with `session_timeout_ms` i... | 2020-07-02T12:40:13 | 2020-07-02T12:40:13 | {} | MEMBER |
ClickHouse | ClickHouse | 653,528,367 | 12,102 | alexey-milovidov | What issues did it found? You can post all of them here without any worries... | 2020-07-03T12:38:05 | 2020-07-03T12:38:05 | {} | MEMBER |
ClickHouse | ClickHouse | 653,552,213 | 11,913 | alexey-milovidov | > compact parts don't support columns' special codecs
But why not?
It's reasonable to use codecs even for compact parts.
What if I want NONE... or if I know that Delta, LZ4 is better? | 2020-07-03T13:35:45 | 2020-07-03T13:36:41 | {} | MEMBER |
ClickHouse | ClickHouse | 651,960,729 | 12,045 | alexey-milovidov | Ok, though I did not see this test before as flaky. | 2020-06-30T18:16:35 | 2020-06-30T18:16:35 | {} | MEMBER |
ClickHouse | ClickHouse | 652,394,481 | 12,068 | bohutang | It's better to do a ping() and re-connect() if the connection is broken before we use the result here:
https://github.com/ClickHouse/ClickHouse/blob/master/src/Formats/MySQLBlockInputStream.cpp#L97 | 2020-07-01T12:40:40 | 2020-07-01T12:40:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,415,407 | 12,051 | bohutang | Fix #12071 | 2020-07-01T13:22:44 | 2020-07-01T13:22:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,095,462 | 12,091 | alesapin | Now it's possible to test images in PR, so let's try again. | 2020-07-02T16:05:57 | 2020-07-02T16:05:57 | {} | MEMBER |
ClickHouse | ClickHouse | 653,140,418 | 12,094 | nvartolomei | DDL worker tries to resolve addresses like `ch%2Ds1%2Dr1:9000` (from hosts list in ZK task) and find if any of them are bound to local network interfaces to determine if the task is intended for it, looks like those aren't resolving to any local interfaces? | 2020-07-02T17:44:57 | 2020-07-02T17:44:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,378,588 | 12,041 | azat | >I'm rewriting all the contribs not to use old fuzzy and error-prone heuristics from find/*.cmake files. Instead I'm getting rid of them. Why do you want to bring it back?
Didn't know that
>There is a new and much better way of describing external library linkage. Take a look at replxx for example.
Rewritten w... | 2020-07-03T06:34:12 | 2020-07-03T06:34:12 | {} | MEMBER |
ClickHouse | ClickHouse | 653,453,010 | 10,267 | filimonov | Because the fix was included in 20.4.3.16 or newer. | 2020-07-03T09:36:30 | 2020-07-03T09:36:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,488,979 | 8,159 | wbigdata | > @wbigdata解决了！
How to solve it can you share,thank you very much.
| 2020-07-03T10:59:17 | 2020-07-03T10:59:17 | {} | NONE |
ClickHouse | ClickHouse | 652,211,445 | 11,401 | akuzm | Merged muzzy_decay_ms = 10s. | 2020-07-01T06:05:40 | 2020-07-01T06:05:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,402,827 | 10,436 | gglinux | Thanks, It didn't happen in v20.3.8.53-lts | 2020-07-01T12:58:48 | 2020-07-01T12:58:48 | {
"heart": 1
} | NONE |
ClickHouse | ClickHouse | 652,474,491 | 12,072 | den-crane | duplicate https://github.com/ClickHouse/ClickHouse/issues/7365
fixed https://github.com/ClickHouse/ClickHouse/pull/7377
| 2020-07-01T15:06:12 | 2020-07-01T15:06:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,982,668 | 11,781 | alexey-milovidov | The link
http://hadoop.apache.org/zookeeper/docs/current/zookeeperAdmin.html
in documentation is outdated,
@BayoNet need to replace. | 2020-07-02T12:42:14 | 2020-07-02T12:42:14 | {} | MEMBER |
ClickHouse | ClickHouse | 653,381,167 | 12,100 | filimonov | The most typical reason for such error is when you screw up the setting on the sharded system. For example - due to the wrong values of 'macros' shard2 uses the zookeeper path of shard1.
Check the macros settings (system.mactos) & zookeeper paths.
Also you can check the following:
Get the list of parts from ... | 2020-07-03T06:42:09 | 2020-07-03T06:42:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,065,680 | 11,576 | den-crane | It needs to apply TTL to **ALL parts** of all partitions.
It seems TTL drops only one **PART** per merge_with_ttl_timeout. | 2020-07-02T15:12:38 | 2020-07-02T15:12:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,382,136 | 12,094 | thyn | It's working. | 2020-07-03T06:45:08 | 2020-07-03T06:45:08 | {} | NONE |
ClickHouse | ClickHouse | 653,461,970 | 8,159 | gubinjie | @wbigdata solved！ | 2020-07-03T09:57:32 | 2020-07-03T09:57:32 | {} | NONE |
ClickHouse | ClickHouse | 652,074,519 | 10,994 | den-crane | reproduce
```
<dictionaries_lazy_load>false</dictionaries_lazy_load>
cat /etc/clickhouse-server/node.tsv
1 xxxx
cat /etc/clickhouse-server/dict/node.xml
<dictionaries>
<dictionary>
<name>node</name>
<source>
<file>
<path>/etc/clickhouse-server/node.tsv<... | 2020-06-30T22:14:24 | 2020-06-30T22:14:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,453,804 | 10,994 | vitlibar | I'll try to fix it. | 2020-07-01T14:30:57 | 2020-07-01T14:30:57 | {
"+1": 1,
"heart": 3
} | MEMBER |
ClickHouse | ClickHouse | 652,670,566 | 12,079 | filimonov | No but you can do atomic exchange in new database=Atomic engine.
Why do you need that? | 2020-07-01T22:09:25 | 2020-07-01T22:27:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,841,661 | 11,590 | azat | This is an attempt to use lock-free part of the allocator (front-end) up to 4MB allocations | 2020-07-02T07:40:02 | 2020-07-02T07:40:02 | {} | MEMBER |
ClickHouse | ClickHouse | 652,854,365 | 11,815 | alesapin | 01193_metadata_loading
https://clickhouse-test-reports.s3.yandex.net/12052/cf306c5be13e6cf128e656086f8e0319e84c4de3/functional_stateless_tests_(thread).html#fail1 | 2020-07-02T08:05:25 | 2020-07-02T08:05:25 | {} | MEMBER |
ClickHouse | ClickHouse | 653,062,672 | 12,086 | filimonov | In the opposite direction there is an issue with LowCardinality(Nullable(String)):
Works:
```
clickhouse-client -q "select CAST(NULL, 'Nullable(String)') as a FORMAT Parquet" | clickhouse-local --structure 'a Nullable(String)' --input-format Parquet -q 'select a from table'
\N
```
With LowCardinality - not:
... | 2020-07-02T15:07:19 | 2020-07-02T15:07:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,217,841 | 12,096 | hodgesrm | Assuming you are using clickhouse-client the way to load data from CSV is as follows:
`cat omega.csv | clickhouse-client --query='insert into test.omega (id,dt) format CSV'`
| 2020-07-02T20:59:36 | 2020-07-02T20:59:36 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,267,382 | 11,696 | CurtizJ | Looks like we can't completly remove preliminary merge step while reading in order, because of performance degradation in case of reading from many parts. So, I suggest to introduce threshold and if number of parts exceeds it, preliminary merge step will be added. | 2020-07-03T00:08:06 | 2020-07-03T00:08:06 | {} | MEMBER |
ClickHouse | ClickHouse | 653,452,802 | 11,411 | hexiaoting | this issue can be closed. @alexey-milovidov | 2020-07-03T09:36:00 | 2020-07-03T09:36:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 652,639,284 | 12,075 | azat | >Ok, i think i understand what happened. I think it was not between poll and read_kafka_message (in that case next() call would just return false, 0 rows world be parsed and it would never enter if (new_rows),
but stop should happened inside read_kafka_message itself. First call of next() function return the message a... | 2020-07-01T20:48:59 | 2020-07-01T20:48:59 | {} | MEMBER |
ClickHouse | ClickHouse | 652,882,384 | 11,873 | azat | > Maybe print it before plan, under another one optio
I was thinking about showing it in the form of `ReadFromStorage x 10 (Read from Distributed)` (`x10` part), but showing it before, under option looks good too! | 2020-07-02T08:58:33 | 2020-07-02T08:58:33 | {} | MEMBER |
ClickHouse | ClickHouse | 652,988,631 | 11,781 | alexey-milovidov | I've edited the page (english).
@BayoNet @filimonov Could you please update other languages if you like? | 2020-07-02T12:55:08 | 2020-07-02T12:55:08 | {} | MEMBER |
ClickHouse | ClickHouse | 653,112,883 | 12,069 | 4ertus2 | It was the most simple way to integrate multiple JOINs logic in ClickHouse query pipeline: to automatically rewrite query with multiple JOINs into query with subselects each containing one JOIN. Unfortunatelly there're name collisions in this case that are solved via special aliases. Such aliases are expected to be uni... | 2020-07-02T16:41:57 | 2020-07-02T17:50:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,273,933 | 11,696 | alexey-milovidov | Ok. | 2020-07-03T00:39:56 | 2020-07-03T00:39:56 | {} | MEMBER |
ClickHouse | ClickHouse | 653,401,286 | 12,091 | alesapin | Ok, the error reproduces. Let's try to fix it. | 2020-07-03T07:38:04 | 2020-07-03T07:38:04 | {} | MEMBER |
ClickHouse | ClickHouse | 653,424,656 | 11,942 | bobelev | > That thread uses system profile (default if you did't change that)
By system profile you mean `<default_profile>default</default_profile>` in `config.xml`?
Anyway, it should be working before and after a restart in the same manner | 2020-07-03T08:32:15 | 2020-07-03T08:32:15 | {} | NONE |
ClickHouse | ClickHouse | 652,067,425 | 10,177 | oik741 | Faced the same problem after upgrade from 20.1 to 20.3.10.75 | 2020-06-30T21:56:54 | 2020-06-30T21:56:54 | {
"+1": 3
} | NONE |
ClickHouse | ClickHouse | 652,838,075 | 12,075 | filimonov | Yes, it seems like logical error (the behavior of functions isStalled changed during refactorings, and that part of code was not fixed accordingly).
I've pushed a commit to your branch see https://github.com/ClickHouse/ClickHouse/pull/12075/commits/821317e6273ea3e9f790bec4162ac27986c69571 | 2020-07-02T07:32:50 | 2020-07-02T07:32:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,143,615 | 12,094 | thyn | Yes, they are not local. May be would be better also to compare interserver_http_host? Well, ok I will try to use pods directly, thanks | 2020-07-02T17:52:30 | 2020-07-02T17:52:30 | {} | NONE |
ClickHouse | ClickHouse | 653,476,810 | 12,081 | filimonov | Let's also clean those bugs: https://github.com/ClickHouse/ClickHouse/labels/prio-major | 2020-07-03T10:31:11 | 2020-07-03T10:31:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 653,528,035 | 12,101 | alexey-milovidov | > On ClickHouse fork found 175 instances of 7 different CWE (including contrib).
Could you please post a list of them here? | 2020-07-03T12:37:16 | 2020-07-03T12:37:16 | {} | MEMBER |
ClickHouse | ClickHouse | 653,533,151 | 12,101 | alexey-milovidov | @blinkov You can publish all of them here without any worries. | 2020-07-03T12:50:25 | 2020-07-03T12:50:25 | {} | MEMBER |
ClickHouse | ClickHouse | 668,991,218 | 12,390 | filimonov | Someday real (i.e. 'cheap') alters/deletes may be introduced. Setting is required. | 2020-08-05T05:39:36 | 2020-08-05T05:39:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 669,105,658 | 11,334 | alexey-milovidov | This pull request is reviewed but the feature is found to be far from being ready for practical use-cases
(there is no work on privacy budgets; guarantees of algorithms are unclear).
Close for now. | 2020-08-05T10:06:46 | 2020-08-05T10:07:30 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.