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 | 875,969,877 | 18,135 | alexey-milovidov | @RonanMorgan @g1rlp0w3r Your data is broken due to faulty hardware.
If you are using Replicated tables (redundancy), the data will be recovered automatically (no action required).
If you are not using Replicated tables, it is possible to continue operation by moving out broken data part manually (move the directo... | 2021-07-07T22:14:37 | 2021-07-07T22:14:37 | {
"+1": 4
} | MEMBER |
ClickHouse | ClickHouse | 875,967,428 | 26,024 | den-crane | When you insert into server1 using HTTP 2021-01-01 00:00:00 -- CH thinks that '2021-01-01 00:00:00' (BERLIN TZ) and stores in a table the value 1609455600 -- 2020-12-31 23:00:00 UTC
When you insert into server2 using HTTP 2021-01-01 00:00:00 -- CH thinks that '2021-01-01 00:00:00' (UTC) and stores in a table the val... | 2021-07-07T22:09:54 | 2021-07-07T23:12:08 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,074,591 | 17,623 | alexey-milovidov | @genzgd Yes, it is added three months ago: #22992 | 2021-07-08T02:37:17 | 2021-07-08T02:37:17 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 876,082,113 | 23,867 | akamensky | I see the same behavior in `21.5.5.12`. But with entire table TTL. Table defined similar to:
```
create table tbl_name
(
ts DateTime,
name String,
value String
)
engine = MergeTree
partition by toStartOfHour(ts)
order by ts
ttl ts + toIntervalHour(48)
settings index_granularity = 8192
```
Then r... | 2021-07-08T02:58:39 | 2021-07-08T03:00:31 | {} | NONE |
ClickHouse | ClickHouse | 876,458,394 | 7,752 | xoelop | @PHameete you can probably do that with some array functions, `arrayFill` is what you need. Here's a practical example: https://blog.tinybird.co/2021/05/18/tips-7-filling-null-values-on-clickhouse/ | 2021-07-08T13:53:00 | 2021-07-08T13:53:00 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,219,023 | 25,523 | feihengye | @filimonov
reproduce, config kafka_thread_per_consumer = 1 ...
| 2021-07-07T02:22:03 | 2021-07-07T02:23:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,690,026 | 24,194 | kssenii | @Mergifyio update | 2021-07-07T15:13:36 | 2021-07-07T15:13:36 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 876,221,763 | 21,101 | kitaisreal | @VKoptev thanks for example. Will look into that. | 2021-07-08T08:00:22 | 2021-07-08T08:00:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,552,769 | 24,972 | alesapin | No, still cannot understand... If `_pFile` is `nullptr` then we should reallocate it in `unsafeOpen()`. If `unsafeOpen()` throws then we cannot get into `mustRotate`. So yes, the logger can be left in the empty state, but race condition doesn't look . | 2021-07-08T15:52:38 | 2021-07-08T15:52:38 | {} | MEMBER |
ClickHouse | ClickHouse | 875,186,478 | 19,897 | alexey-milovidov | Nevertheless, I welcome if you write, that Coderrect has been successfully evaluated with ClickHouse :)
You can write it on official website if you want.
Though the chance we will use Coderrect in our CI is low (the reason is high resource usage). | 2021-07-07T00:57:53 | 2021-07-07T00:58:40 | {} | MEMBER |
ClickHouse | ClickHouse | 875,575,667 | 26,033 | alexey-milovidov | Ok. The server processed this request successfully with neglible amount of allocated RAM.
It remains in question why the client allocates so much RAM on the same request.
Does it die instantly or it takes several seconds?
You can stop clickhouse-client with gdb during this several seconds and print stack traces:... | 2021-07-07T12:50:31 | 2021-07-07T12:50:31 | {} | MEMBER |
ClickHouse | ClickHouse | 875,794,732 | 26,045 | Algunenano | > @Algunenano Thank you! Your server is so cool!
Thanks! It's my dev machine but I agree, it's cool :sunglasses: | 2021-07-07T17:33:28 | 2021-07-07T17:33:28 | {} | MEMBER |
ClickHouse | ClickHouse | 875,836,882 | 23,508 | kssenii | @hamsterready It looks like you have some other index with modulo? Because `loadIndex()` in trace does not load partition key, it loads primary key. But It was only fixed for partition key. What is table definition? | 2021-07-07T18:34:46 | 2021-07-07T18:34:46 | {} | MEMBER |
ClickHouse | ClickHouse | 875,861,235 | 25,931 | alexey-milovidov | Ok.
PS. Sometimes I want to emphasize that this is just a *minimal* web UI. | 2021-07-07T19:11:16 | 2021-07-07T19:11:16 | {} | MEMBER |
ClickHouse | ClickHouse | 876,433,244 | 25,957 | alesapin | Failures unrelated to changes. | 2021-07-08T13:19:27 | 2021-07-08T13:19:27 | {} | MEMBER |
ClickHouse | ClickHouse | 876,240,794 | 26,052 | vdimir | Failed tests:
> AST fuzzer (debug) — Assertion `frame_start <= frame_end' failed
Not related to changes, see https://github.com/ClickHouse/ClickHouse/pull/26072
> Performance — 4 errors, 2 too long, 2 faster
Added cases was too long on current master
> Stress test (thread) — Test script failed
Seems ... | 2021-07-08T08:26:44 | 2021-07-08T14:47:10 | {} | MEMBER |
ClickHouse | ClickHouse | 875,164,000 | 26,042 | alexey-milovidov | Yes, it is possible that the excessive memory usage is related to S3 tiered storage, that is currently in development and is not ready for production. | 2021-07-07T00:06:45 | 2021-07-07T00:06:45 | {} | MEMBER |
ClickHouse | ClickHouse | 875,280,093 | 26,038 | mergify[bot] | **Command `refresh`: success**
> **Pull request refreshed**
>
| 2021-07-07T04:59:26 | 2021-07-07T04:59:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,368,886 | 25,999 | simPod | I'm on 20.12.5.14, sorry for not mentioning it before. I'll try 21.4. | 2021-07-07T07:42:15 | 2021-07-07T07:42:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,219,213 | 26,071 | shadowDy | > Use replicated tables / zookeeper
Why do distributed tables have this problem
I understand that if local is written first and then distributed to other nodes or replicas in the form of files, the short-term data will be delayed (inconsistent), but it should be the same in the end. | 2021-07-08T07:56:29 | 2021-07-08T07:56:29 | {} | NONE |
ClickHouse | ClickHouse | 875,959,256 | 17,623 | alexey-milovidov | @awslabspl
> modify changelog as soon as new release is made
For a few past releases we prepared changelog in advance - in a few days before the release.
> Make log messages meaningfull.
We make our log messages meaningful and understandable. I always keep an eye on log messages, so there are no overwhelmi... | 2021-07-07T21:52:30 | 2021-07-07T21:57:59 | {} | MEMBER |
ClickHouse | ClickHouse | 876,197,020 | 26,071 | filimonov | Use replicated tables / zookeeper | 2021-07-08T07:20:57 | 2021-07-08T07:20:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,406,701 | 25,956 | alesapin | Failures are not related to changes. | 2021-07-08T12:42:33 | 2021-07-08T12:42:33 | {} | MEMBER |
ClickHouse | ClickHouse | 876,410,195 | 25,589 | vdimir | I temporary removed queries with syntax error from test to merge this PR. When I figure out how not to cause fuzzer error with such queries, I'll enable it. | 2021-07-08T12:47:53 | 2021-07-08T12:47:53 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 875,185,721 | 19,897 | alexey-milovidov | I've read all 111 entries and found all of them false positive. | 2021-07-07T00:55:54 | 2021-07-07T00:55:54 | {} | MEMBER |
ClickHouse | ClickHouse | 875,373,442 | 25,743 | alesapin | No related failures. | 2021-07-07T07:49:11 | 2021-07-07T07:49:11 | {} | MEMBER |
ClickHouse | ClickHouse | 875,424,511 | 25,990 | EgorKurito | @den-crane We use block devices connected over the network. a typical scheme for any cloud provider. moreover, we have mongaDB on the same disks, for example.
Also, 5 MB seems to be not so small. example:
https://forums.tomshardware.com/threads/what-is-a-good-or-normal-read-write-speed-for-a-hdd.2822841/ | 2021-07-07T08:59:16 | 2021-07-07T08:59:16 | {} | NONE |
ClickHouse | ClickHouse | 875,966,371 | 17,623 | alexey-milovidov | @melin This task is in queue after "Provide alternative for ZooKeeper" and "Simplify replication". It is not in the roadmap for 2021, because the chance to get it in 2021 is lower than 50%.
| 2021-07-07T22:07:46 | 2021-07-07T22:07:57 | {} | MEMBER |
ClickHouse | ClickHouse | 876,570,961 | 26,088 | kssenii | > new connections to CH seem to trigger new connections to PostgreSQL
It is possible in case of concurrent access to tables, but onlyy in scope of one pool size.
> Multiple Engine=PostgreSQL() tables
btw, how many? just interested.
> After an update from CH 21.4.6.55 to CH 21.6.5.37
Actually a little str... | 2021-07-08T16:17:08 | 2021-07-08T16:22:08 | {} | MEMBER |
ClickHouse | ClickHouse | 875,185,416 | 19,897 | alexey-milovidov | ```
__attribute__((constructor)) static void __auxv_init()
{
size_t i;
for (i = 0; __environ[i]; i++);
__auxv = (unsigned long *) (__environ + i + 1);
size_t secure_idx = __find_auxv(AT_SECURE);
if (secure_idx != ((size_t) -1))
__auxv_secure = __auxv[secure_idx];
}
```
The fun... | 2021-07-07T00:55:10 | 2021-07-07T00:55:10 | {} | MEMBER |
ClickHouse | ClickHouse | 875,373,163 | 25,884 | alesapin | no related failures. | 2021-07-07T07:48:44 | 2021-07-07T07:48:44 | {} | MEMBER |
ClickHouse | ClickHouse | 875,749,319 | 12,576 | alexey-milovidov | It should be added in `ISerialization` similar to Text, JSON, CSV, etc.
Default implementation will call `deserializeTextEscaped`, then check for end of line and throw if it is not at end of line.
Implementation for String-like types will be overloaded and use something similar to `readEscapedStringUntilEOL`. | 2021-07-07T16:29:09 | 2021-07-07T16:29:09 | {} | MEMBER |
ClickHouse | ClickHouse | 875,797,867 | 23,508 | hamsterready | I still see this issue in 21.6.6.51:
```
96798c2dfb5e :) alter table report_by_utms ATTACH part '202004_0_5_1';
ALTER TABLE report_by_utms
ATTACH PART '202004_0_5_1'
Query id: ad53dd80-3a26-42c8-8e62-6b1601146c1e
0 rows in set. Elapsed: 0.004 sec.
Received exception from server (version 21.6.6):... | 2021-07-07T17:38:32 | 2021-07-07T17:38:32 | {} | NONE |
ClickHouse | ClickHouse | 875,904,729 | 26,048 | CurtizJ | Integration tests failures are not related to changes. | 2021-07-07T20:19:30 | 2021-07-07T20:19:30 | {} | MEMBER |
ClickHouse | ClickHouse | 876,017,858 | 26,046 | zhicwu | Sure, I'll look into this today.
The timestamps are different in above queries so I guess the major differences are in the where clause. I haven't enabled SQL rewrite in JDBC driver yet but it's possible related to SQL parsing. | 2021-07-08T00:14:03 | 2021-07-08T00:14:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,543,730 | 26,086 | nvartolomei | https://github.com/ClickHouse/ClickHouse/blob/9f5cd35a6963cc556a51218b46b0754dcac7306a/src/Access/SettingsProfilesCache.cpp#L176-L180
Looks like the profile is skipped here. | 2021-07-08T15:40:22 | 2021-07-08T15:40:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,313,089 | 23,020 | baofengzhouyu | I have the same error on version 20.8.18.32 lts | 2021-07-07T06:09:09 | 2021-07-07T06:09:09 | {} | NONE |
ClickHouse | ClickHouse | 875,394,040 | 26,037 | vpanfilov | Duplicate of https://github.com/ClickHouse/ClickHouse/issues/22029 | 2021-07-07T08:17:14 | 2021-07-07T08:17:14 | {} | NONE |
ClickHouse | ClickHouse | 875,930,889 | 26,024 | depyronick | Imagine two servers
**Server #1** with OS timezone set to _Europe/Berlin_ and clickhouse server timezone set to _Europe/Berlin_ and with clickhouse version of _21.3.2.5_.
**Server #2** with OS timezone set to _Europe/Berlin_ and clickhouse server timezone set to _UTC_ and with clickhouse version of _21.6.5.37_
... | 2021-07-07T21:02:58 | 2021-07-07T21:02:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,070,433 | 26,066 | akamensky | The query plan for sorting by Int64:
```
┌─explain─────────────────────────────────────────────────────────────────────────────┐
│ Expression (Projection) │
│ MergingSorted (Merge sorted streams for ORDER BY) │
│ Me... | 2021-07-08T02:26:36 | 2021-07-08T02:26:36 | {} | NONE |
ClickHouse | ClickHouse | 876,325,705 | 25,853 | mkocztorz | Yes it's partitioned by event_at, but before inserting its grouped to 15 minutes.
Currently we are recreating the cluster. When it's back I will repeat the test and share the changed settings and logs. But as far as I know we were not changing many defaults. | 2021-07-08T10:30:46 | 2021-07-08T10:30:46 | {} | NONE |
ClickHouse | ClickHouse | 876,406,739 | 26,016 | hssrs | **Update:**
After trying to look at `image.json` in the docker directory, I found an image called `yandex/clickhouse-stateless-test`, so I tried to run the functional test this way:
```
# function_test_helper.sh
export PATH=$PATH:/clickhouse/build/programs
"/clickhouse/build/programs/clickhouse" --server -C /... | 2021-07-08T12:42:37 | 2021-07-08T12:46:15 | {} | NONE |
ClickHouse | ClickHouse | 875,165,016 | 26,042 | alexey-milovidov | Run the following query to check from where the memory is allocated:
```
SELECT
count(),
sum(size) AS bytes,
arrayStringConcat(arrayMap(x -> concat(demangle(addressToSymbol(x)), '\n ', addressToLine(x)), trace), '\n') AS sym
FROM system.trace_log
WHERE event_date >= today() - 5 AND trace_type = ... | 2021-07-07T00:08:53 | 2021-07-07T00:09:36 | {} | MEMBER |
ClickHouse | ClickHouse | 875,183,021 | 25,966 | alexey-milovidov | @azat We have misterious rare query hangs in performance test*. It is blocking release 21.7.
This PR is needed to eliminate one possible cause - Python ClickHouse driver.
\* most likely it is described and fixed here: #25968 | 2021-07-07T00:50:22 | 2021-07-07T00:50:47 | {} | MEMBER |
ClickHouse | ClickHouse | 875,183,549 | 25,966 | alexey-milovidov | Unfortunately, the parameters of `subprocess.run` method in Python are too much dependent on Python version. | 2021-07-07T00:51:27 | 2021-07-07T00:51:27 | {} | MEMBER |
ClickHouse | ClickHouse | 875,279,973 | 26,038 | amosbird | @Mergifyio refresh | 2021-07-07T04:59:07 | 2021-07-07T04:59:07 | {
"+1": 2
} | COLLABORATOR |
ClickHouse | ClickHouse | 875,281,499 | 26,038 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-07-07T05:02:45 | 2021-07-07T05:02:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,366,078 | 25,589 | TszKitLo40 | Why the AST fuzzer test failed? @vdimir This exception should be thrown only `distinct on` and `limit by` exists together.
| 2021-07-07T07:38:04 | 2021-07-07T07:39:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,290,862 | 19,575 | vitlibar | I'll finish the work in https://github.com/ClickHouse/ClickHouse/pull/26075 | 2021-07-08T09:37:27 | 2021-07-08T09:37:27 | {} | MEMBER |
ClickHouse | ClickHouse | 876,470,453 | 26,067 | alexey-milovidov | Ok. I'm also not sure about loading js from dagrejs.github.io. But I don't know what is the most trustworthy CDN. | 2021-07-08T14:07:21 | 2021-07-08T14:07:21 | {} | MEMBER |
ClickHouse | ClickHouse | 875,989,472 | 17,623 | genzgd | Will clickhouse-keeper support TLS in the initial release? | 2021-07-07T23:03:09 | 2021-07-07T23:03:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,141,839 | 21,101 | VKoptev | @kitaisreal
```
$ docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
30f6a73644bc yandex/clickhouse-server:21.1.7.1 "/entrypoint.sh" 25 minutes ago Up 7 minutes 8123/tcp, 9000/tcp, 9009/tcp clickhouse_1
... | 2021-07-08T05:35:09 | 2021-07-08T05:35:09 | {} | NONE |
ClickHouse | ClickHouse | 876,519,383 | 26,046 | den-crane | I was unable to reproduce it with ch server + curl .
```
21.8.1.7290
create table test_table(Ca Int64, Cb Int64, Cc String, TIMESTAMP Int64) Engine=Memory;
for i in `seq 1 5000`; do curl -s -F 'L0=@L0' -F 'L1=@L0' -F 'L2=@L0' -F 'L3=@L0' -F 'L4=@L0' -F 'L5=@L0' -F 'L6=@L0' -F 'L7=@L0' -F 'L8=@L0' -F 'L9=@L0' ... | 2021-07-08T15:08:37 | 2021-07-08T15:08:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,170,389 | 24,416 | alexey-milovidov | Perf test Ok (the column has been renamed and the new name does not exist in old version). | 2021-07-07T00:21:44 | 2021-07-07T00:21:44 | {} | MEMBER |
ClickHouse | ClickHouse | 875,467,193 | 26,038 | Algunenano | I've tested it locally and I can confirm #26020 is fixed | 2021-07-07T09:58:07 | 2021-07-07T09:58:07 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 875,882,403 | 26,046 | kitaisreal | @zhicwu can you take a look at it ? | 2021-07-07T19:45:14 | 2021-07-07T19:45:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,240,270 | 24,172 | taylor12805 | I try to reproduce the test case from website: https://oracle-base.com/articles/misc/rollup-cube-grouping-functions-and-grouping-sets
`
CREATE TABLE default.dimension_tab (
`fact_1_id` Int32,
`fact_2_id` Int32,
`fact_3_id` Int32,
`fact_4_id` Int32,
`sales_value` Decimal(9, 2)
) ENGINE = MergeTree
O... | 2021-07-08T08:26:01 | 2021-07-08T08:26:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,282,086 | 25,958 | alesapin | no related failures | 2021-07-08T09:24:24 | 2021-07-08T09:24:24 | {} | MEMBER |
ClickHouse | ClickHouse | 876,296,278 | 25,960 | BoloniniD | Tried to reproduce the problem with
> ./clickhouse local --stacktrace -q "select * from file(0, 'TSV', 'key String');" <<<foo
As it turns out `ITableFunctionFileLike::parseArguments` fails to represent 0 in the arguments for the file as a String. If I use '0' instead, then `ITableFunctionFileLike::parseArguments... | 2021-07-08T09:45:47 | 2021-07-08T09:45:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,400,295 | 26,070 | den-crane | 21.6 will not be LTS.
LTS will be 21.8 or 21.9.
The experimental mark will be removed from Projections someday in the future, nobody know when. | 2021-07-08T12:32:47 | 2021-07-08T12:32:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,403,727 | 26,071 | den-crane | >Comparing the data of two replicates, there are some differences.
Distributed table does an insert-retry in case of an error.
MergeTree table does not have a deduplication feature.
There is a scenario possible:
1. Distributed table inserts into the shard
2. Shard successfully applies the insert.
3. Distribut... | 2021-07-08T12:38:04 | 2021-07-08T12:51:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,331,373 | 25,589 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-07-07T06:42:34 | 2021-07-07T06:42:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,419,452 | 25,990 | EgorKurito | **Update**
Sorry for slow answer,
One of solution
If I search by partition, like this:
```SQL
SELECT ts, raw
FROM logs
WHERE device_id = '0e38b9b9-df38-4560-8886-8645a32616a7'
AND ts <= toDateTime64('2021-06-10T14:23:57.912', 6)
AND week = toMonday(toDate('2021-06-10'))
ORDER BY ts DESC
LIMIT 10;
```
... | 2021-07-07T08:52:18 | 2021-07-07T08:52:54 | {} | NONE |
ClickHouse | ClickHouse | 875,866,438 | 26,050 | alexey-milovidov | It was already implemented but not documented.
#26055 | 2021-07-07T19:18:52 | 2021-07-07T19:19:02 | {} | MEMBER |
ClickHouse | ClickHouse | 875,963,943 | 17,623 | alexey-milovidov | @zouyonghao
> Fuzzing of cluster configurations
The idea is to run our tests but replace all tables with Distributed tables with some numbers of shards and replicas. Maybe adding cases when Distributed table look at other Distributed tables.
The first step is to ensure that there are no failed assertions, no ... | 2021-07-07T22:02:53 | 2021-07-07T22:04:52 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 876,343,101 | 25,994 | electrical | @alexey-milovidov Yeah, we found that out as well. It's something I realised a bit late :-( We bumped it to 500k which has solved our issue.
I'll have a look at the linked issue as well for anything I might have missed.
Thank you all so much for your help! | 2021-07-08T11:00:50 | 2021-07-08T11:00:50 | {} | NONE |
ClickHouse | ClickHouse | 876,431,936 | 25,959 | alesapin | failures unrelated to changes. | 2021-07-08T13:17:47 | 2021-07-08T13:17:47 | {} | MEMBER |
ClickHouse | ClickHouse | 876,453,723 | 26,046 | zhicwu | @Zark1999, I was not able to reproduce the issue on both 20.8 and 21.3. Does it has to be ReplicatedMergeTree? Are you saying the exact same case on 20.8 worked but not on 21.3? And the two environments are similar?
It sounds like a server side issue, but can you add `log4j.logger.org.apache.http=DEBUG` into your lo... | 2021-07-08T13:47:33 | 2021-07-08T13:47:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,456,790 | 24,997 | evillique | TODO:
- [x] Fix unbundled build
- [ ] Change unordered_map in plain synonyms extensions for something better
- [ ] Test Wordnet synonyms extensions for different languages (currently only english thesaurus was tested)
- [ ] Add optional parameter POS to function synonyms()
- [ ] Improve documentation | 2021-07-07T09:43:05 | 2021-07-08T14:27:01 | {} | MEMBER |
ClickHouse | ClickHouse | 876,572,381 | 26,066 | akamensky | @kitaisreal
## Reproducible example
I've put the simple tool to fill the data here: https://github.com/akamensky/ch_fill
You can also download the linux 64bit compiled binary there in Releases.
For the schema (also available in the code comment for copy-paste):
```
CREATE TABLE default.test
(
`ts... | 2021-07-08T16:19:11 | 2021-07-08T16:28:12 | {} | NONE |
ClickHouse | ClickHouse | 875,168,658 | 15,765 | alexey-milovidov | Why we do
SSL\_**CTX**\_set_cert_cb
but inside the callback we do
SSL\_**(no CTX)**use_certificate
?
It looks inconsistent.
See
SSL_set_cert_cb
vs
SSL_CTX_set_cert_cb | 2021-07-07T00:17:41 | 2021-07-07T00:18:35 | {} | MEMBER |
ClickHouse | ClickHouse | 875,281,443 | 26,038 | amosbird | @Mergifyio update | 2021-07-07T05:02:38 | 2021-07-07T05:02:38 | {
"+1": 1
} | COLLABORATOR |
ClickHouse | ClickHouse | 875,488,059 | 25,595 | nikitamikhaylov | @Mergifyio update | 2021-07-07T10:27:10 | 2021-07-07T10:27:10 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 875,488,240 | 25,595 | mergify[bot] | **Command `update`: success**
> **Branch has been successfully updated**
>
| 2021-07-07T10:27:27 | 2021-07-07T10:27:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,692,561 | 24,194 | mergify[bot] | **Command `update`: failure**
> **Base branch update has failed**
> merge conflict between base and head
> err-code: DE59E
| 2021-07-07T15:16:38 | 2021-07-07T15:16:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,169,507 | 23,867 | akamensky | See also closed #10004, looks like it was not fixed after all? | 2021-07-08T06:33:03 | 2021-07-08T06:33:03 | {} | NONE |
ClickHouse | ClickHouse | 876,308,674 | 26,076 | cadet354 | Move question into clickhouse-jdbc https://github.com/ClickHouse/clickhouse-jdbc/issues/681 | 2021-07-08T10:04:44 | 2021-07-08T10:04:44 | {} | NONE |
ClickHouse | ClickHouse | 875,167,761 | 15,765 | alexey-milovidov | The function `setCertificate` is called whenever the server needs certificate.
It's in question - is the server's certificate the only case when SSL certificate is used...
I hope there are no cases with client certificates. Looks ok. | 2021-07-07T00:15:40 | 2021-07-07T00:15:40 | {} | MEMBER |
ClickHouse | ClickHouse | 875,349,500 | 26,042 | kitaisreal | @nikitamikhaylov can you take a look at it ? | 2021-07-07T07:12:41 | 2021-07-07T07:12:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,518,168 | 24,997 | alesapin | > TODO:
>
> * [ ] Fix unbundled build
>
> * [ ] Change unordered_map in plain synonyms extensions for something better
>
> * [ ] Test Wordnet synonyms extensions for different languages (currently only english thesaurus was tested)
>
> * [ ] Add optional parameter POS to function synonym... | 2021-07-07T11:15:50 | 2021-07-07T11:16:10 | {} | MEMBER |
ClickHouse | ClickHouse | 875,524,830 | 26,034 | alesapin | No related failures. | 2021-07-07T11:26:29 | 2021-07-07T11:26:29 | {} | MEMBER |
ClickHouse | ClickHouse | 875,957,896 | 26,041 | azat | >Functional stateless tests (release) — fail: 1, passed: 3202
#26058 | 2021-07-07T21:49:34 | 2021-07-07T21:49:34 | {} | MEMBER |
ClickHouse | ClickHouse | 876,087,638 | 26,066 | akamensky | @den-crane 21.5.5.12 | 2021-07-08T03:10:46 | 2021-07-08T03:10:46 | {} | NONE |
ClickHouse | ClickHouse | 876,068,098 | 17,623 | h6353115 | > @h6353115 We have functional completeness of clickhouse-keeper and we already started evaluating it in our environments.
> And it is already released in advance, e.g. in 21.7-prestable. Currently we are working on documentation and testing on more cases from production. (It is already being tested 24/7 in our CI and... | 2021-07-08T02:20:33 | 2021-07-08T04:00:27 | {} | NONE |
ClickHouse | ClickHouse | 876,422,804 | 25,710 | alesapin | Too complex to backport | 2021-07-08T13:05:50 | 2021-07-08T13:05:50 | {} | MEMBER |
ClickHouse | ClickHouse | 876,434,929 | 25,871 | alesapin | This patch doesn't fix the whole problem (logical error in log), but makes the situation better. | 2021-07-08T13:21:37 | 2021-07-08T13:21:48 | {} | MEMBER |
ClickHouse | ClickHouse | 876,474,501 | 26,067 | alexey-milovidov | Checked in Firefox and Chromium. | 2021-07-08T14:12:32 | 2021-07-08T14:12:32 | {} | MEMBER |
ClickHouse | ClickHouse | 875,280,391 | 26,026 | amosbird | The issue is related to misused query ast in normal projections. I've posted a proper fix in https://github.com/ClickHouse/ClickHouse/pull/26038 . Thank you for the report! | 2021-07-07T05:00:11 | 2021-07-07T05:00:11 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 875,578,645 | 25,966 | alexey-milovidov | It looks like #25968 was the source of trouble and now it is fixed.
This PR is no longer relevant. | 2021-07-07T12:54:53 | 2021-07-07T12:54:53 | {} | MEMBER |
ClickHouse | ClickHouse | 875,780,330 | 26,045 | alexey-milovidov | @Algunenano Thank you! Your server is so cool! | 2021-07-07T17:13:16 | 2021-07-07T17:13:16 | {
"laugh": 1
} | MEMBER |
ClickHouse | ClickHouse | 876,080,542 | 26,066 | den-crane | CH Version?
| 2021-07-08T02:54:06 | 2021-07-08T02:54:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 876,220,902 | 26,066 | kitaisreal | @akamensky can you provide reproducible example with tables configuration ? Just fill necessary columns with random timestamps should be enought. | 2021-07-08T07:59:05 | 2021-07-08T07:59:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 875,328,981 | 25,589 | vdimir | `01268_procfs_metrics` failed, shouldn't be related, I'll try to investigate, but now we can update to master and run checks again.
@mergifyio update | 2021-07-07T06:38:04 | 2021-07-07T06:38:04 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 875,518,988 | 26,033 | w4x-gh | The request:
debian:~# curl 'http://localhost:8123/?user=default&password=somepwd' -d 'SHOW DATABASES'
mychdb
default
system
The logs:
2021.07.07 13:12:07.822396 [ 23062 ] {} <Trace> HTTPHandler-factory: HTTP Request for HTTPHandler-factory. Method: POST, Address: [::1]:46146, User-Agent: curl/7.64.0, Lengt... | 2021-07-07T11:17:11 | 2021-07-07T11:17:11 | {} | NONE |
ClickHouse | ClickHouse | 875,955,896 | 17,623 | alexey-milovidov | @h6353115 We have functional completeness of clickhouse-keeper and we already started evaluating it in our environments.
And it is already released in advance, e.g. in 21.7-prestable. Currently we are working on documentation and testing on more cases from production. (It is already being tested 24/7 in our CI and wit... | 2021-07-07T21:45:35 | 2021-07-07T21:53:45 | {} | MEMBER |
ClickHouse | ClickHouse | 876,227,171 | 23,508 | hamsterready | @kssenii thank you for pointing it out, you are right, table to which I was attaching part is missing primary key definition. | 2021-07-08T08:07:40 | 2021-07-08T08:07:40 | {} | NONE |
ClickHouse | ClickHouse | 848,913,615 | 24,350 | MyroTk | @alexey-milovidov The testflows check is green, should be ready for merge. | 2021-05-26T16:22:24 | 2021-05-26T16:22:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 849,435,878 | 24,462 | KochetovNicolai | Functional stateless tests (release, DatabaseReplicated)
```
2021-05-27 04:42:45 00910_zookeeper_test_alter_compression_codecs_long: [ FAIL ] 31.35 sec. - return code 159
2021-05-27 04:42:45 Received exception from server (version 21.7.1):
2021-05-27 04:42:45 Code: 159. DB::Exception: Received ... | 2021-05-27T08:16:39 | 2021-05-27T08:16:39 | {} | MEMBER |
ClickHouse | ClickHouse | 849,544,651 | 24,081 | weiyongyuan | I also faced this probloem .The ClickHouse'version is 21.3.3.14. | 2021-05-27T11:09:23 | 2021-05-27T11:09:23 | {} | NONE |
ClickHouse | ClickHouse | 849,580,798 | 24,477 | vdimir | Could you show your `cmake` command?
Also provide a bit more about you environment, OS, please | 2021-05-27T12:13:12 | 2021-05-27T12:13:12 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.