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 | 576,583,178 | 8,636 | filimonov | > > What query do you execute? What is the structure of your tables and configuration of the cluster?
>
> we replayed all the query, no crash happened
If you're satisfied with the result 'it crashes only sometimes', you can close the issue.
If you want to be sure that it is fixed and the problem will not retu... | 2020-01-21T09:00:25 | 2020-01-21T09:07:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,603,269 | 8,341 | alesapin | Conflicts in 19.17 :( | 2020-01-21T09:53:12 | 2020-01-21T09:53:12 | {} | MEMBER |
ClickHouse | ClickHouse | 576,612,540 | 8,375 | alesapin | conflicts in 19.17 | 2020-01-21T10:16:01 | 2020-01-21T10:16:01 | {} | MEMBER |
ClickHouse | ClickHouse | 576,622,934 | 8,636 | alexey-milovidov | > If you're satisfied with the result 'it crashes only sometimes', you can close the issue.
I'm not. | 2020-01-21T10:40:19 | 2020-01-21T10:40:19 | {} | MEMBER |
ClickHouse | ClickHouse | 576,624,716 | 8,755 | alexey-milovidov | Build is broken. | 2020-01-21T10:44:52 | 2020-01-21T10:44:52 | {} | MEMBER |
ClickHouse | ClickHouse | 576,639,435 | 8,684 | 4ertus2 | @aslojno, you may try to run queries with `optimize_move_to_prewhere = 0` | 2020-01-21T11:25:32 | 2020-01-21T11:25:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,252,196 | 4,277 | akuzm | > Do we have an issue on this? Have we fixed this already?
Not sure, personally I haven't done anything about this. | 2020-01-22T15:52:47 | 2020-01-22T15:52:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,425,032 | 8,750 | alexey-milovidov | > Got it, I want to have separate default disk path (since default selector uses <path> and I want to use <tmp_path>)
Not sure that I understand this completely, but looks like it's not an issue:
- if `tmp_policy` is not specified, `tmp_path` is used;
- if `tmp_policy` is default, then default policy is used that ... | 2020-01-20T20:44:44 | 2020-01-20T20:44:44 | {} | MEMBER |
ClickHouse | ClickHouse | 576,438,786 | 6,187 | zlobober | I created an issue CLICKHOUSE-4771 in internal task tracker with repro. Please take another look on this. | 2020-01-20T21:43:20 | 2020-01-20T21:43:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,596,900 | 8,526 | alesapin | Contains feature, so cannot be cherry-picked in 19.17. | 2020-01-21T09:37:11 | 2020-01-21T09:37:11 | {} | MEMBER |
ClickHouse | ClickHouse | 576,611,458 | 8,318 | alesapin | Fix cannot be applied to 19.17. | 2020-01-21T10:13:21 | 2020-01-21T10:13:21 | {} | MEMBER |
ClickHouse | ClickHouse | 576,614,799 | 8,419 | alesapin | conflicts in 19.17 | 2020-01-21T10:21:24 | 2020-01-21T10:21:24 | {} | MEMBER |
ClickHouse | ClickHouse | 576,626,617 | 6,187 | 4ertus2 | It's hard to fix without perfomance penalties. | 2020-01-21T10:50:00 | 2020-01-21T10:50:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,241,900 | 8,731 | akuzm | So basically we are constructing a reference-counted memory block cache over an mmap-backed memory allocator that allocates blocks of the same size, with the defining feature being that the allocator can ask the cache to evict some pages. And probably it can also give some hint about the address at which to evict, if i... | 2020-01-22T15:31:12 | 2020-01-22T15:31:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,260,200 | 5,421 | abyss7 | Continued in #5629 | 2020-01-22T16:09:21 | 2020-01-22T16:09:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,315,196 | 8,784 | 4ertus2 | Could you print workarounds by @den-crane from mail too? | 2020-01-22T18:13:02 | 2020-01-22T18:13:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,428,239 | 8,750 | alexey-milovidov | Always using `tmp` subdirectory is Ok. | 2020-01-20T20:58:33 | 2020-01-20T20:58:33 | {} | MEMBER |
ClickHouse | ClickHouse | 576,460,176 | 7,878 | vzakaznikov | Simple test case
```bash
$ cat 01069_materialized_view_alter_target_table.sql
DROP TABLE IF EXISTS mv_source;
DROP TABLE IF EXISTS mv_target;
CREATE TABLE mv_source (`a` UInt64) ENGINE = MergeTree ORDER BY tuple();
CREATE TABLE mv_target (`a` UInt64) ENGINE = MergeTree ORDER BY tuple();
CREATE MATERIALIZED V... | 2020-01-20T23:32:38 | 2020-01-20T23:34:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,537,482 | 8,736 | filimonov | > Detailed description (optional):
> Fix #8568
Seems unrelated, check issue number. | 2020-01-21T06:28:44 | 2020-01-21T06:28:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,602,279 | 8,684 | m1khal3v | The second option is our case, because we have partitioning by toRelativeWeekNum(updated_at). The shard key in all tables with column item_id is (item_id % 3), so the same item_id will be on the same shard. In the logs we have this
```
executeQuery: (from *) SELECT DISTINCT g.item_id FROM (SELECT gs.item_id FROM gm... | 2020-01-21T09:50:38 | 2020-01-21T09:50:38 | {} | NONE |
ClickHouse | ClickHouse | 576,645,652 | 8,573 | filimonov | https://github.com/ClickHouse/ClickHouse/compare/v19.17.5.18-stable...v19.17.6.36-stable
That may be the possible reason: https://github.com/ClickHouse/ClickHouse/pull/8128 (didn't check)
| 2020-01-21T11:45:14 | 2020-01-21T11:46:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,851,390 | 7,612 | stavrolia | > @alexey-milovidov Thanks you!
>
> If you have an Working example of importing a file from S3, can you please give me.
>
> What version of clickhouse has this feature.
You can import a file with `select * from s3('URL', 'format', 'columns')` - where `URL` means URL, `format` means format of your file and `col... | 2020-01-21T19:52:04 | 2020-01-21T19:52:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,900,138 | 7,612 | manojfim | Thanks, @stavrolia !!
Since I am doing insert into click house should I do insert into tablename select * from s3('URL', 'format', 'columns')
where should i give secretkey and accesskey? | 2020-01-21T21:53:08 | 2020-01-21T21:53:08 | {} | NONE |
ClickHouse | ClickHouse | 577,018,574 | 8,773 | byx313 | > duplicate: #7764
> #8548
>
> ```
> select intDiv(toInt64(1579449600000),60000)*60
>
> ┌─multiply(intDiv(toInt64(1579449600000), 60000), 60)─┐
> │ -17118312120 │
> └─────────────────────────────────────────────────────┘
>
>
> select intDiv(toInt64(1579449600000),t... | 2020-01-22T05:46:36 | 2020-01-22T05:46:36 | {} | NONE |
ClickHouse | ClickHouse | 576,432,423 | 8,738 | filimonov | realtes #208 #902 #2825 etc... (See [label:comp-http](https://github.com/ClickHouse/ClickHouse/labels/comp-http) for others)
Possible alternative in 'general' case (for all formats): https://github.com/ClickHouse/ClickHouse/issues/6294 | 2020-01-20T21:16:00 | 2020-01-20T21:16:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,434,439 | 6,294 | filimonov | > Is the preferred workflow to have X+1 requests or only one?
i think one is enough for the start
> For one request -- The problem is that the default web browser implementation of EventSource does not support POST. https://github.com/mpetazzoni/sse.js has been able to work around this issue.
It's ok if that ... | 2020-01-20T21:24:23 | 2020-01-20T21:24:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,448,028 | 8,703 | filimonov | Another direction of optimization is to avoid some internal grouping if 'external' GROUP BY does grouping subset of columns used in for internal grouping, i.e.
```
select sum(x), y from (select sum(x) as x, y, z from table group by y, z ) group by y
```
can be optimized into
```
select sum(x), y from table group... | 2020-01-20T22:24:28 | 2020-01-20T22:24:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,466,564 | 4,754 | alexey-milovidov | Need to check on recent version or to provide more info to reproduce. | 2020-01-21T00:11:09 | 2020-01-21T00:11:09 | {} | MEMBER |
ClickHouse | ClickHouse | 576,503,575 | 8,757 | hczhcz | See also: #8691
Potentially, similar cases may exist somewhere else. | 2020-01-21T03:32:47 | 2020-01-21T03:32:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,602,198 | 8,385 | alesapin | conflicts in 19.17 | 2020-01-21T09:50:25 | 2020-01-21T09:50:25 | {} | MEMBER |
ClickHouse | ClickHouse | 576,603,570 | 8,334 | alesapin | conflicts in 19.17. | 2020-01-21T09:53:59 | 2020-01-21T09:53:59 | {} | MEMBER |
ClickHouse | ClickHouse | 576,911,569 | 2,621 | den-crane | https://t.me/c/1425947904/4246
> Denny Crane (I don't work at Yandex (never did)), [Jan 21, 2020 at 5:00:12 PM]:
> А что написано в CLICKHOUSE-3381 ?
> Зачем оно (Don't deduplicate a block inserted from MV if the source block is Ok) ?
> Что пытались починить?
> Alexey Milovidov, [Jan 21, 2020 at 5:09:42 PM]:
... | 2020-01-21T22:18:50 | 2020-01-21T23:09:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,052,016 | 8,772 | krnaveen14 | Generally if you `PARTITION BY d`, then it would be useless to `ORDER BY d` as the data is already separated by `d`. So practical useful way is to `ORDER BY os`.
I would suggest the following schema for better performance and storage efficiency.
```
CREATE TABLE default.a
(
`d` Date,
`os` LowCardina... | 2020-01-22T07:51:27 | 2020-01-22T07:51:27 | {} | NONE |
ClickHouse | ClickHouse | 577,324,781 | 8,386 | akuzm | Using another test runner for now. | 2020-01-22T18:36:31 | 2020-01-22T18:36:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,597,376 | 8,503 | alesapin | too many conflicts in 19.17. | 2020-01-21T09:38:38 | 2020-01-21T09:38:38 | {} | MEMBER |
ClickHouse | ClickHouse | 576,620,782 | 8,684 | 4ertus2 | OK. And what's the result of the query if you run it manually?
```
SELECT item_id AS item_id_in_items, updated_at AS isua FROM default.item AS is FINAL WHERE isua < toDateTime('2020-01-02 00:00:00')
```
Would it possible to get old value (that should be overriden by FINAL) with some WHERE predicate? | 2020-01-21T10:35:11 | 2020-01-21T10:36:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,625,083 | 8,758 | filimonov | AFAIK you can't put whole users.xml into zookeeper.
You can move there only particular sections.
Like `<profiles from_zk="/clickhouse/users/profiles" replace="replace"></profiles>` inside regular users.xml file.
And
```
<default>
<max_memory_usage>10000000000</max_memory_usage>
... | 2020-01-21T10:45:47 | 2020-01-21T10:55:04 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,667,128 | 6,440 | achulkov2 | It seems that we still don't know the actual number of rows in the last block. The number is counted correctly even when adaptive granularity is used, but it's still an approximate. | 2020-01-21T12:51:33 | 2020-01-21T12:51:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,111,579 | 8,639 | alesapin | `test_storage_kafka` error under address sanitizer seems like a flap, because it works in release build and logs don't contain any sanitizer assertions. tests of multiple volumes are known to be flaky. | 2020-01-22T10:18:50 | 2020-01-22T10:18:50 | {} | MEMBER |
ClickHouse | ClickHouse | 577,250,770 | 8,731 | akuzm | By the way, is the cache size bounded from above? What happens if we don't have more memory for cache, but still have to read the marks or something else? I'm asking because for a bounded cache, you can use fixed-size data structures that avoid the need for synchronization in some cases. E.g. if your map of key->'cach... | 2020-01-22T15:49:53 | 2020-01-22T15:49:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,452,717 | 8,747 | filimonov | There were some related discussions here: https://github.com/ClickHouse/ClickHouse/pull/4542
I think arrays is just the simplest way to archve 'good enough' results.
Other possible ideas (as an alternative to arrays)
1) something similar to ORDER BY WITH FILL (see https://github.com/ClickHouse/ClickHouse/pull/5... | 2020-01-20T22:49:18 | 2020-01-20T22:49:18 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,615,071 | 8,684 | m1khal3v | The logs from the previous comment relate to the node on which the request was made. On the other node we see the following
```
SELECT item_id FROM default.gmc_item AS gs FINAL WHERE (item_id = 122478784742) AND (flag = 1)
Expression
Filter
ReplacingSorted
Expression × 23
MergeTree
SELECT item_id ... | 2020-01-21T10:22:01 | 2020-01-21T10:22:01 | {} | NONE |
ClickHouse | ClickHouse | 576,634,516 | 8,684 | alexey-milovidov | We have to disable PREWHERE optimization if FINAL is present.
PS. Internal discussion:
```
Artem Zuikov, [21.01.20 13:42]
вот тут у нас ожидаемое поведение в последнем запросе (преликат из WHERE применяется раньше FINAL) или баг? https://github.com/ClickHouse/ClickHouse/issues/8684
Alexey Milovidov, [21.01.20 ... | 2020-01-21T11:11:38 | 2020-01-21T11:11:38 | {} | MEMBER |
ClickHouse | ClickHouse | 576,671,466 | 7,612 | alexey-milovidov | I forget that:
- s3 table function is implemented and released;
- support for wildcards in file, hdfs, url is implemented and released;
- support for wildcards in s3 is *not implemented*. | 2020-01-21T13:04:08 | 2020-01-21T13:04:08 | {} | MEMBER |
ClickHouse | ClickHouse | 576,671,718 | 7,612 | alexey-milovidov | Sorry for incomplete info.
This task is assigned to @stavrolia | 2020-01-21T13:04:51 | 2020-01-21T13:04:51 | {} | MEMBER |
ClickHouse | ClickHouse | 577,104,905 | 8,750 | azat | Hm, `tmp` subdirectory is missing, will fix it
>Always using tmp subdirectory is Ok.
@alexey-milovidov so `tmp_path` directive can go away? | 2020-01-22T10:02:17 | 2020-01-22T10:02:17 | {} | MEMBER |
ClickHouse | ClickHouse | 577,109,304 | 7,683 | alesapin | Fails of multiple volumes tests seem to be OK because they are known to be flaky and possible I have chery-picked something wrong. Performance failure is also OK because we don't have some performance optimizations in this branch. Smoke test for the split build is not important. | 2020-01-22T10:13:26 | 2020-01-22T10:13:26 | {} | MEMBER |
ClickHouse | ClickHouse | 576,464,878 | 3,529 | alexey-milovidov | @wawanawna This exception means that connection to ZooKeeper was lost and ClickHouse will reconnect to ZooKeeper. This behaviour is normal and doesn't lead to data loss. | 2020-01-21T00:01:59 | 2020-01-21T00:01:59 | {} | MEMBER |
ClickHouse | ClickHouse | 577,371,274 | 8,784 | tdingus | @den-crane suggested some workaround in https://groups.google.com/forum/#!topic/clickhouse/nHYAl1oGghk:
bad solution:
CREATE TABLE traffic (id Int64, count Int64) Engine=Memory;
CREATE TABLE category (cat_id Int64, min Int64, max Int64, fake Int64 materialized 1) Engine=Memory;
INSERT INTO category (cat_id, m... | 2020-01-22T20:29:09 | 2020-01-22T20:29:09 | {} | NONE |
ClickHouse | ClickHouse | 576,413,883 | 4,263 | alexey-milovidov | Dangling references to Context were fixed.
But the root cause (bad code structure around passing Context) is not fixed. | 2020-01-20T20:04:48 | 2020-01-20T20:04:48 | {} | MEMBER |
ClickHouse | ClickHouse | 576,613,388 | 4,971 | filimonov | https://github.com/ClickHouse/ClickHouse/issues/902 | 2020-01-21T10:17:59 | 2020-01-21T10:17:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,668,197 | 7,612 | manojfim | @alexey-milovidov thanks for the confirmation. Can you please point me towards documentation or an example of inserting all csv files from s3 directory.
| 2020-01-21T12:54:46 | 2020-01-21T12:54:46 | {} | NONE |
ClickHouse | ClickHouse | 576,414,966 | 4,810 | alexey-milovidov | We don't have wrong query result in newer versions.
But we still lack of type conversions for IN subquery. | 2020-01-20T20:08:56 | 2020-01-20T20:08:56 | {} | MEMBER |
ClickHouse | ClickHouse | 576,426,801 | 8,750 | azat | >Probably one issue is that tmp subdirectory will not be created under path.
> Maybe change the meaning of tmp_path so that it will be equal to path by default but tmp subdirectory is always used inside tmp_path?
The issue is that there is `tmp_path` and `path`, so if `storage_policy` will be used instead and no `t... | 2020-01-20T20:52:16 | 2020-01-20T20:52:16 | {} | MEMBER |
ClickHouse | ClickHouse | 576,539,479 | 8,636 | kwai-code | > What query do you execute? What is the structure of your tables and configuration of the cluster?
we replayed all the query, no crash happened | 2020-01-21T06:37:46 | 2020-01-21T06:37:46 | {} | NONE |
ClickHouse | ClickHouse | 576,599,702 | 8,446 | alesapin | conflicts in 19.17 | 2020-01-21T09:44:25 | 2020-01-21T09:44:25 | {} | MEMBER |
ClickHouse | ClickHouse | 576,615,343 | 8,422 | alesapin | conflicts in 19.17 | 2020-01-21T10:22:45 | 2020-01-21T10:22:45 | {} | MEMBER |
ClickHouse | ClickHouse | 576,623,604 | 7,612 | alexey-milovidov | Yes. | 2020-01-21T10:42:00 | 2020-01-21T10:42:00 | {} | MEMBER |
ClickHouse | ClickHouse | 576,675,276 | 7,612 | manojfim | @alexey-milovidov Thanks you!
If you have an Working example of importing a file from S3, can you please give me.
What version of clickhouse has this feature. | 2020-01-21T13:14:30 | 2020-01-21T13:14:30 | {} | NONE |
ClickHouse | ClickHouse | 576,475,558 | 8,736 | dingxiangfei2009 | Sorry for the mistakes. I am inspecting the failed testing. | 2020-01-21T01:05:10 | 2020-01-21T01:05:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,634,009 | 8,636 | filimonov | > > If you're satisfied with the result 'it crashes only sometimes', you can close the issue.
>
> I'm not.
Me neither :) May be @amosbird or @akuzm will be able to deduct what was that looking on the code only, since @kwai-code don't want to help us. | 2020-01-21T11:10:01 | 2020-01-21T11:10:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,679,266 | 8,684 | m1khal3v | We tried to register the setting in the yandex -> profiles -> default block, tried to tie it to the session using SET, but this did not give any result. But also we tried to transfer all the data from the partitioned table to the identical one without partitioning and this helped due to the fact that the data merged. | 2020-01-21T13:23:01 | 2020-01-21T13:23:17 | {} | NONE |
ClickHouse | ClickHouse | 576,747,328 | 8,750 | azat | Looks like the following is failing in upstream too:
- test_ttl_move/test.py::test_double_move_while_select[test_double_move_while_select_positive-1]
- test_multiple_disks/test.py::test_concurrent_alter_move[concurrently_altering_replicated_mt-ReplicatedMergeTree('/clickhouse/concurrently_altering_replicated_mt', '... | 2020-01-21T15:54:30 | 2020-01-21T15:54:30 | {} | MEMBER |
ClickHouse | ClickHouse | 577,252,574 | 5,071 | abyss7 | We decided to close PR and postpone this feature for now. | 2020-01-22T15:53:36 | 2020-01-22T15:53:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,613,521 | 8,357 | alesapin | Conflicts in 19.17 :( | 2020-01-21T10:18:20 | 2020-01-21T10:18:20 | {} | MEMBER |
ClickHouse | ClickHouse | 576,658,831 | 8,573 | filimonov | After #8128 socket timeout start being respected (which were ignored before for TLS).
Can you try to increase default timeouts? | 2020-01-21T12:26:15 | 2020-01-21T12:26:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,554,208 | 2,621 | urykhy | по мотивам a0d06c5ee0a7f4fd37a0bea43fff973c16ad80b0 получилось сделать такой патч:
```diff
diff --git a/dbms/src/DataStreams/PushingToViewsBlockOutputStream.cpp b/dbms/src/DataStreams/PushingToViewsBlockOutputStream.cpp
index 195c5edcb0..52105202c2 100644
--- a/dbms/src/DataStreams/PushingToViewsBlockOutputStream... | 2020-01-21T07:31:59 | 2020-01-21T12:29:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,015,774 | 7,612 | excitoon | > Not sure, probably it works in test releases
>
> like this
>
> ```
> insert into mytable
> select * from s3('http://s3auth:s3auth@host/bucket/test.csv', 'CSV', 'column1 UInt32, column2 UInt32, column3 UInt32')
> ```
That is not correct. One shall use this syntax:
```
select * from s3('http://host/buck... | 2020-01-22T05:34:32 | 2020-01-22T05:34:32 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,287,784 | 7,432 | abraithwaite | > We don't enforce backward compatibility for system tables. But it should be mentioned in changelog and documentation.
Why? We depend on these tables for operations and system management. It is not a nice experience to have those break suddenly upon upgrade. Even in testing, where we discovered it, it requires a... | 2020-01-22T17:08:25 | 2020-01-22T17:09:31 | {} | NONE |
ClickHouse | ClickHouse | 576,466,122 | 4,470 | alexey-milovidov | Need to check on version 20.1 prestable. | 2020-01-21T00:08:23 | 2020-01-21T00:08:23 | {} | MEMBER |
ClickHouse | ClickHouse | 576,902,690 | 8,726 | yurriy | > Can we do the following:
>
> 1. Generate `password_double_sha1_hex` in the same way as `password_sha256_hex` in ClickHouse installation script.
> 2. Enable `mysql_port` by default.
>
> ?
1. Do you mean without openssl tool?
2. Yes, we can. | 2020-01-21T21:59:02 | 2020-01-21T21:59:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,936,975 | 8,769 | alexey-milovidov | Ok. | 2020-01-21T23:34:03 | 2020-01-21T23:34:03 | {} | MEMBER |
ClickHouse | ClickHouse | 577,135,798 | 7,407 | alesapin | Mutations fix doesn't affect any tests except integration tests for mutations. So we can't wait for them. | 2020-01-22T11:21:16 | 2020-01-22T11:21:16 | {} | MEMBER |
ClickHouse | ClickHouse | 576,455,810 | 7,878 | vzakaznikov | The issue is caused by https://github.com/ClickHouse/ClickHouse/blob/c4e2d71b482da3a7df859d8facef4975ac511a00/dbms/src/DataStreams/ConvertingBlockInputStream.cpp#L65.
| 2020-01-20T23:06:04 | 2020-01-20T23:06:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,546,251 | 7,612 | manojfim | Is S3 import/export function released ? | 2020-01-21T07:04:11 | 2020-01-21T07:04:11 | {} | NONE |
ClickHouse | ClickHouse | 576,593,185 | 8,738 | filimonov | I think it's completely ok to serialize non-data events always the same way (JSON sounds like a good choice), and datastream in the desired format.
I.e. you can have something like
```
Content-Type: text/event-stream
event: meta
data: { "query_id": "82be54be-cb77-4c78-b116-8f0a9a4d9cc4", "format": "CSV","st... | 2020-01-21T09:26:16 | 2020-01-21T09:27:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,617,064 | 8,479 | alesapin | Conflicts in 19.17, because it depends on https://github.com/ClickHouse/ClickHouse/pull/8400. | 2020-01-21T10:26:44 | 2020-01-21T10:26:44 | {} | MEMBER |
ClickHouse | ClickHouse | 576,938,071 | 8,726 | alexey-milovidov | > Do you mean without openssl tool?
No, I didn't mean that. I thought we generate `password_sha256_hex` in `postinst` but actually we don't. | 2020-01-21T23:38:36 | 2020-01-21T23:38:36 | {} | MEMBER |
ClickHouse | ClickHouse | 577,003,429 | 8,750 | azat | Looks like test failures are not relevant to this changes, or am I missing something? | 2020-01-22T04:35:45 | 2020-01-22T04:35:45 | {} | MEMBER |
ClickHouse | ClickHouse | 577,179,431 | 8,165 | MakarovYaroslav | Are there any changes? | 2020-01-22T13:23:44 | 2020-01-22T13:23:44 | {} | NONE |
ClickHouse | ClickHouse | 577,266,570 | 8,573 | dstruck | Version 19.17.6.36 and version 20.1.2.4 give the same errors even when increasing the timeouts:
```
<!-- Connection timeout if there are no replicas. default=10 -->
<connect_timeout>60</connect_timeout>
<!-- default=300 -->
<receive_timeout>600</receive_timeout>
<!-- default=300 -->
<send_timeout>600</send_t... | 2020-01-22T16:22:33 | 2020-01-22T16:22:33 | {} | NONE |
ClickHouse | ClickHouse | 576,622,642 | 8,684 | m1khal3v | Yes, it's possible
```
SELECT
item_id AS item_id_in_items,
updated_at AS isua
FROM default.item AS is
FINAL
WHERE (isua < toDateTime('2020-01-02 00:00:00')) AND (item_id = 122478784742)
┌─item_id_in_items─┬────────────────isua─┐
│ 122478784742 │ 2019-12-19 11:11:08 │
└──────────────────┴──────... | 2020-01-21T10:39:40 | 2020-01-21T10:39:40 | {} | NONE |
ClickHouse | ClickHouse | 576,681,722 | 8,764 | 4ertus2 | #8312 | 2020-01-21T13:29:14 | 2020-01-21T13:29:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,015,127 | 8,773 | den-crane | duplicate: https://github.com/ClickHouse/ClickHouse/issues/7764
https://github.com/ClickHouse/ClickHouse/issues/8548
```
select intDiv(toInt64(1579449600000),60000)*60
┌─multiply(intDiv(toInt64(1579449600000), 60000), 60)─┐
│ -17118312120 │
└─────────────────────────────... | 2020-01-22T05:31:42 | 2020-01-22T05:31:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,234,220 | 4,277 | abyss7 | Do we have an issue on this? Have we fixed this already? | 2020-01-22T15:15:58 | 2020-01-22T15:15:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,235,855 | 4,385 | abyss7 | Is it still relevant? BTW your own test fails. | 2020-01-22T15:19:29 | 2020-01-22T15:19:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,325,160 | 7,432 | alexey-milovidov | Ok, we should note all the changes in changelog as "backward incompatible changes".
Regarding the question of "why", we simply do not have practice to keep them backward compatible. Maybe we should. | 2020-01-22T18:37:23 | 2020-01-22T18:37:34 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 576,414,414 | 4,476 | alexey-milovidov | We need to add type conversions for Views. | 2020-01-20T20:06:49 | 2020-01-20T20:06:49 | {} | MEMBER |
ClickHouse | ClickHouse | 576,465,769 | 7,676 | alexey-milovidov | @stavrolia has added support for wildcards and virtual columns. | 2020-01-21T00:06:24 | 2020-01-21T00:06:24 | {} | MEMBER |
ClickHouse | ClickHouse | 576,890,242 | 8,770 | alexey-milovidov | Look at QueryLog.h
We need to add exception code there. | 2020-01-21T21:27:43 | 2020-01-21T21:27:43 | {} | MEMBER |
ClickHouse | ClickHouse | 577,017,816 | 8,568 | huang--yijian | This issue is about failure to upgrade ClickHouse by RPM packages.
We have shutdown ClickHouse server successfully before we start the upgrade.
#8736 is about failure to close ClickHouse thread pool which was not running at all when we started the upgrade. Why is #8736 going to fix this issue of our failed upgrad... | 2020-01-22T05:43:11 | 2020-01-22T05:43:39 | {} | NONE |
ClickHouse | ClickHouse | 577,181,039 | 7,676 | stavrolia | Wildcards and virtual columns are implemented now only in HDFS anf File. The issue for s3 is [here](https://github.com/ClickHouse/ClickHouse/issues/7612). | 2020-01-22T13:27:25 | 2020-01-22T13:28:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 576,414,338 | 4,476 | alexey-milovidov | Original issue was fixed.
And now we have error message instead:
```
milovidov-Pro-P30 :) create view testv(a UInt32) as select number a from numbers(10);
CREATE VIEW testv
(
`a` UInt32
) AS
SELECT number AS a
FROM numbers(10)
Ok.
0 rows in set. Elapsed: 0.016 sec.
milovidov-Pro-P30 :) selec... | 2020-01-20T20:06:32 | 2020-01-20T20:06:32 | {} | MEMBER |
ClickHouse | ClickHouse | 576,421,743 | 8,750 | azat | @alexey-milovidov So, the question is, what is preferable?
I did the same thing for distributed engine (did not submitted yet), by adding another separate `distributed_storage_configuration`, to leave `storage_configuration` for `MergeTree` engine only, but I don't have any objective reasons, | 2020-01-20T20:31:41 | 2020-01-20T20:31:41 | {} | MEMBER |
ClickHouse | ClickHouse | 576,429,342 | 8,750 | azat | Ok | 2020-01-20T21:02:55 | 2020-01-20T21:02:55 | {} | MEMBER |
ClickHouse | ClickHouse | 576,792,815 | 8,763 | alesapin | I'll merge and check in Release PR. | 2020-01-21T17:33:14 | 2020-01-21T17:33:14 | {} | MEMBER |
ClickHouse | ClickHouse | 576,813,047 | 8,750 | azat | >Always using tmp subdirectory is Ok.
Rebased (`<tmp_path>` behavior left unchanged) | 2020-01-21T18:21:49 | 2020-01-21T18:21:49 | {} | MEMBER |
ClickHouse | ClickHouse | 576,913,005 | 8,769 | traceon | > Please mind that some formats have aliases. For example, TabSeparated = TSV.
@alexey-milovidov I thought it would make sense to always report the "canonical" name of the format, but it seemed to me that it would be too intrusive to expose all those `getName()`'s all the way to `executeQuery()`, so this change repo... | 2020-01-21T22:22:27 | 2020-01-21T22:22:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 577,238,164 | 8,685 | mnacharov | Is it valid decision to add systemd as subrepo, and describe which sources to compile in `systemd-cmake/CMakeLists.txt`? | 2020-01-22T15:23:31 | 2020-01-22T15:23:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,707,025 | 8,515 | azat | And this is not related only to `SimpleAggregateFunction` but to `LowCardinality(Nullable())` | 2020-01-03T21:43:05 | 2020-01-03T21:43:05 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.