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 | 759,925,461 | 18,935 | fastio | > @fastio Just wonder. Why did you use integration test? You can simply make `.sql & .reference` see https://github.com/ClickHouse/ClickHouse/blob/master/tests/queries/0_stateless/01532_execute_merges_on_single_replica.sq
Thank you for your suggestion. | 2021-01-14T04:56:16 | 2021-01-14T04:56:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 760,100,263 | 19,054 | alexey-milovidov | Thank you!
That's embarrassing. How it was even possible? | 2021-01-14T10:18:03 | 2021-01-14T10:18:23 | {} | MEMBER |
ClickHouse | ClickHouse | 759,258,289 | 18,935 | alesapin | Sorry, didn't notice that the test flaky. | 2021-01-13T07:21:20 | 2021-01-13T07:21:20 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 759,483,563 | 18,856 | tavplubix | I cannot reproduce the issue, could you please provide a full list of queries (including `CREATE`s and `INSERT`s) to reproduce it and output of `select * from system.clusters format TSV` (or cluster description from config.xml)? | 2021-01-13T14:27:23 | 2021-01-13T14:27:23 | {} | MEMBER |
ClickHouse | ClickHouse | 759,715,712 | 18,853 | azat | Here are some numbers, in general, they shows that `CheckingCompressedReadBuffer` has almost zero overhead.
### patched
```sql
create table data (key Int, value String) engine=Null();
create table dist (key Int, value String) engine=Distributed(test_shard_localhost, currentDatabase(), data);
system stop distri... | 2021-01-13T20:19:22 | 2021-01-13T20:20:22 | {} | MEMBER |
ClickHouse | ClickHouse | 759,776,284 | 18,552 | CurtizJ | I will add support of `EXISTS VIEW` for ANTLR parser in separate PR. | 2021-01-13T22:23:28 | 2021-01-13T22:23:28 | {} | MEMBER |
ClickHouse | ClickHouse | 759,819,363 | 18,791 | Avogar | Test failures are not related. | 2021-01-13T23:18:14 | 2021-01-13T23:18:14 | {} | MEMBER |
ClickHouse | ClickHouse | 759,865,925 | 18,996 | chenk818 | > Verify results getMacro results with `select * from system.macros`
system.macros returns the same value | 2021-01-14T01:33:42 | 2021-01-14T01:33:42 | {} | NONE |
ClickHouse | ClickHouse | 760,119,266 | 18,982 | nikitamikhaylov | > BTW why it was added? (here https://github.com/ClickHouse/ClickHouse/pull/6375/files , so - quite long ago and was not affecting users before )
>
> UPD (nikitamikhaylov): Moved the issue number to the description because Github recognizes it as a PR number. Just not to confuse people
I don't know. We discussed ... | 2021-01-14T10:53:34 | 2021-01-14T10:53:34 | {} | MEMBER |
ClickHouse | ClickHouse | 760,157,193 | 18,954 | alexey-milovidov | Yes, and we can rename it to just `clickhouse`. | 2021-01-14T12:10:14 | 2021-01-14T12:10:14 | {} | MEMBER |
ClickHouse | ClickHouse | 759,350,583 | 15,889 | KochetovNicolai | ```
SELECT count(7 =
(
SELECT number
FROM numbers(0)
ORDER BY number ASC NULLS FIRST
LIMIT 7
))
FROM xp_d
PREWHERE toYYYYMM(A) GLOBAL IN
(
SELECT
NULL =
(
SELECT number
FROM numbers(1)
ORDER BY number DESC... | 2021-01-13T10:17:21 | 2021-01-13T10:17:21 | {} | MEMBER |
ClickHouse | ClickHouse | 759,891,425 | 18,991 | awesomeleo | > Enabling log_bin_use_v1_row_events=1 causes mysqld to write the binary log using Version 1 logging events.
> BTW, Why enable this setting?
Some RDS enables this option. I think they want to compatible with old mysql versions.
Ok. I think this not a problem. Just to verify MytarializeMySQL can support v1.
Thks... | 2021-01-14T02:54:18 | 2021-01-14T02:54:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 757,648,202 | 18,894 | mathfool | something is wrong here, when the alias and the underlying column name is clash but they are actually the same one, it should not throw the exception.
version 20.11.5.18.
JoinToSubqueryTransformVisitor.cpp:477
```
if (got_alias)
{
throw Exception("Alias clashes with qualified column '" + ident->name(... | 2021-01-11T06:19:24 | 2021-01-13T11:06:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,453,494 | 18,910 | alexey-milovidov | Ok. | 2021-01-13T13:38:14 | 2021-01-13T13:38:14 | {} | MEMBER |
ClickHouse | ClickHouse | 759,866,754 | 18,996 | chenk818 | > Macros are initialized on server start. Change of macros in config file does not apply until server restart. Function `getMacro` has probably used old value.
Change of macros in config file will be updated by ConfigReloader,i checked it
otherwise,no one set thats wrong values(RSHARD,RREPLICA) | 2021-01-14T01:36:13 | 2021-01-14T01:36:13 | {} | NONE |
ClickHouse | ClickHouse | 760,049,719 | 18,927 | filimonov | Also uid/gid was changed:
```
### <= 20.10
root@50b8580b79da:/# id clickhouse
uid=101(clickhouse) gid=101(clickhouse) groups=101(clickhouse)
### >= 20.11
root@d49c922a1181:/# id clickhouse
uid=999(clickhouse) gid=999(clickhouse) groups=999(clickhouse)
```
That can require chown on huge data folders dur... | 2021-01-14T09:05:24 | 2021-01-14T09:05:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 760,101,162 | 18,812 | vdimir | As I understand this PR allows to call `getSettings` with not full endpoint, but only with prefix.
@excitoon Could you please provide some extra details? | 2021-01-14T10:19:37 | 2021-01-14T10:19:37 | {} | MEMBER |
ClickHouse | ClickHouse | 760,209,101 | 18,696 | gyuton | @Akazz
Hi!
PR is ready for review. | 2021-01-14T13:51:25 | 2021-01-14T13:51:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,351,343 | 15,889 | KochetovNicolai | ```
WITH concat(addressToLine(arrayJoin(trace) AS addr), '#') AS symbol
SELECT count() > 7
FROM system.trace_log AS t
WHERE (query_id =
(
SELECT
[NULL, NULL, NULL, NULL, 0.00009999999747378752, NULL, NULL, NULL, NULL, NULL],
query_id
FROM system.query_log
WHERE (query LIKE '%test ... | 2021-01-13T10:18:47 | 2021-01-13T10:18:47 | {} | MEMBER |
ClickHouse | ClickHouse | 759,468,125 | 18,542 | alexey-milovidov | Ok, now it's ready for review and merging. | 2021-01-13T14:02:28 | 2021-01-13T14:02:28 | {} | MEMBER |
ClickHouse | ClickHouse | 759,654,774 | 17,955 | den-crane | >whenever I drop the replicated tables, it did not deletes the zookeeper nodes.
It's completely wrong behavior if it's not related to Atomic database.
@BhavyaRajSharma check this https://github.com/ClickHouse/ClickHouse/issues/18382#issuecomment-749854816
| 2021-01-13T19:04:14 | 2021-01-13T19:04:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,936,962 | 10,280 | JJWIT | Same issue! | 2021-01-14T05:32:49 | 2021-01-14T05:32:49 | {} | NONE |
ClickHouse | ClickHouse | 759,973,533 | 19,026 | alexey-milovidov | But we can accept this change for readability. | 2021-01-14T07:07:49 | 2021-01-14T07:07:49 | {} | MEMBER |
ClickHouse | ClickHouse | 759,975,063 | 19,019 | alexey-milovidov | > Integration tests (thread) — fail: 1
@CurtizJ | 2021-01-14T07:11:35 | 2021-01-14T07:11:35 | {} | MEMBER |
ClickHouse | ClickHouse | 760,020,896 | 18,952 | alexey-milovidov | @gonzalezjo Need to fix the "Fast test", see the logs in "Details". | 2021-01-14T08:35:27 | 2021-01-14T08:35:27 | {} | MEMBER |
ClickHouse | ClickHouse | 759,354,220 | 15,889 | KochetovNicolai | ```
WITH (
(
SELECT query_start_time_microseconds
FROM system.query_log
ORDER BY query_start_time DESC
LIMIT 1
) AS time_with_microseconds,
(
SELECT
inf,
query_start_time
FROM system.query_log
ORDER BY query_start_... | 2021-01-13T10:23:54 | 2021-01-13T10:23:54 | {} | MEMBER |
ClickHouse | ClickHouse | 759,414,205 | 10,710 | hhkbble | +1 | 2021-01-13T12:23:01 | 2021-01-13T12:23:01 | {} | NONE |
ClickHouse | ClickHouse | 759,425,688 | 19,007 | KochetovNicolai | Ok, it fails for debug build. | 2021-01-13T12:45:55 | 2021-01-13T12:45:55 | {} | MEMBER |
ClickHouse | ClickHouse | 759,465,704 | 18,542 | nautaa | The problem has been addressed. Thanks for your advice! @alexey-milovidov | 2021-01-13T13:58:38 | 2021-01-13T13:58:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,815,482 | 15,450 | CurtizJ | Maybe need a test for backward compatibility. | 2021-01-13T23:08:01 | 2021-01-13T23:08:01 | {} | MEMBER |
ClickHouse | ClickHouse | 759,905,514 | 18,991 | awesomeleo | @BohuTANG what's the difference of binlog version and binlog row event version? MySQLReplication.cpp#L54 means binlog version. But log_bin_use_v1_row_events controls the bin log row event version. It may be a bug.
ref to https://dev.mysql.com/doc/refman/5.7/en/replication-options-binary-log.html#sysvar_log_bin_us... | 2021-01-14T03:42:10 | 2021-01-14T03:42:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,975,535 | 18,981 | alexey-milovidov | > test_ttl_replicated/test.py::test_ttl_empty_parts
@CurtizJ | 2021-01-14T07:12:45 | 2021-01-14T07:12:45 | {} | MEMBER |
ClickHouse | ClickHouse | 760,070,475 | 19,027 | jfk9w | I've faced the same error upon executing DSL query via Clickhouse JDBC driver (create table). Seems quite inconsistent.
Version 20.12.5.14. | 2021-01-14T09:24:18 | 2021-01-14T09:25:03 | {} | NONE |
ClickHouse | ClickHouse | 760,088,725 | 18,959 | Veryfirefly | > Windows is not supported.
>
> Probably fixed here #17664. Never backported to 20.8.
> Check with 20.12.5.14
>
> Or try to disable Atomic databases.
>
> ```
> # service clickhouse-server stop
>
> # cat /etc/clickhouse-server/conf.d/d_d.xml
> <?xml version="1.0"?>
> <yandex>
> <profiles>
> ... | 2021-01-14T09:57:14 | 2021-01-14T09:57:14 | {} | NONE |
ClickHouse | ClickHouse | 758,510,292 | 18,958 | ElapsedSoul | I have done this:
```shell
git clone https://github.com/llvm/llvm-project.git
cd llvm-project
mkdir build (in-tree build is not supported)
cd build
cmake -DLLVM_ENABLE_PROJECTS=clang -G "Unix Makefiles" ../llvm
make -j8
make install
cd llvm-project
mkdir build-compiler-rt
cd build-compiler-rt
cmake ../com... | 2021-01-12T09:00:19 | 2021-01-13T04:38:53 | {
"+1": 5
} | NONE |
ClickHouse | ClickHouse | 753,437,562 | 18,662 | awesomeleo | > MaterializeMySQL underlying engine is ReplacingMergeTree, this case looks not make sense. Can you insert data to ReplacingMergeTree directly and do the test?
@BohuTANG Yes. I have done the test. It's an issue with ReplacingMergeTree.
MaterializeMySQL with final:
localhost :) select count(*) from mmgt;
SELECT... | 2021-01-02T06:13:58 | 2021-01-13T07:08:29 | {
"+1": 2,
"laugh": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,417,913 | 18,910 | amosbird | ANTLR parser doesn't seem to work with column transformers
https://clickhouse-test-reports.s3.yandex.net/18910/90e9f4fc2e33098ad2442f53758489fc697ce431/functional_stateless_tests_(antlr_debug)/test_run.txt.out.log | 2021-01-13T12:30:25 | 2021-01-13T12:30:25 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 759,435,167 | 18,996 | KochetovNicolai | Macros are initialized on server start. Change of macros in config file does not apply until server restart. Function `getMacro` has probably used old value. | 2021-01-13T13:04:00 | 2021-01-13T13:04:00 | {} | MEMBER |
ClickHouse | ClickHouse | 759,974,153 | 19,026 | alexey-milovidov | But I cannot merge it with this commit message :) | 2021-01-14T07:09:14 | 2021-01-14T07:09:14 | {} | MEMBER |
ClickHouse | ClickHouse | 759,317,768 | 18,999 | filimonov | Same as https://github.com/ClickHouse/ClickHouse/issues/17933 | 2021-01-13T09:19:49 | 2021-01-13T09:19:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,590,628 | 18,924 | snar | > @snar Will you check build from https://clickhouse-builds.s3.yandex.net/18924/af2fa524366a029793de56f2fa38a6ddc532d1c7/clickhouse_special_build_check/report.html ?
This is Linux build, i guess ? Then it makes a little sense to test it as it will neither run on FreeBSD (even within
emulator, already tried that) no... | 2021-01-13T17:09:53 | 2021-01-13T17:09:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,980,139 | 18,723 | alexey-milovidov | #19029 | 2021-01-14T07:23:13 | 2021-01-14T07:23:13 | {} | MEMBER |
ClickHouse | ClickHouse | 760,166,544 | 17,993 | alexey-milovidov | LGTM. I removed `or COLUMNS` fragment because it can confuse reader. | 2021-01-14T12:30:11 | 2021-01-14T12:30:11 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 759,435,661 | 7,192 | djrobby | Does `DROP PARTITION` propagate changes into the `MATERIALIZED VIEW`s? | 2021-01-13T13:04:54 | 2021-01-13T13:04:54 | {} | NONE |
ClickHouse | ClickHouse | 759,921,283 | 18,991 | bohutang | Thanks for the information, will have a look. | 2021-01-14T04:41:47 | 2021-01-14T04:41:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,939,031 | 10,280 | sundy-li | ```
:) SELECT bitmapToArray(bitmapBuild([695788478213916450,693475830172484980,693566989590856286,693439459143715197,693457216849577977])) AS res
SELECT bitmapToArray(bitmapBuild([695788478213916450, 693475830172484980, 693566989590856286, 693439459143715197, 693457216849577977])) AS res
Query id: 0eaea06b-4498... | 2021-01-14T05:39:11 | 2021-01-14T05:39:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 760,026,671 | 18,979 | alexey-milovidov | Unit test has to be updated. | 2021-01-14T08:46:37 | 2021-01-14T08:46:37 | {} | MEMBER |
ClickHouse | ClickHouse | 759,352,492 | 15,889 | KochetovNicolai | ```
WITH addressToSymbol(arrayJoin(trace)) AS symbol
SELECT count() > 0
FROM system.trace_log AS t
WHERE greaterOrEquals(event_date, ignore(ignore(ignore(NULL, ''))), yesterday()) AND (trace_type = 'Memory') AND (query_id =
(
SELECT
ignore(ignore(ignore(ignore(65536)), ignore(65537), ignore(2)), '')... | 2021-01-13T10:20:58 | 2021-01-13T10:20:58 | {} | MEMBER |
ClickHouse | ClickHouse | 759,429,833 | 18,910 | amosbird | > I have a feeling that `WithNames` meaning is non-obvious, it's not clear whether it is
>
> * normalizing names as well (1);
> * keeping names (2).
>
> (in fact it is 2)
Yeah, let's switch to `-KeepNames` | 2021-01-13T12:53:41 | 2021-01-13T12:53:41 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 759,465,813 | 18,996 | den-crane | Verify results getMacro results with `select * from system.macros` | 2021-01-13T13:58:47 | 2021-01-13T13:58:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,590,035 | 17,955 | den-crane | >I delete the zookeeper nodes manually.
This step kinda confuses me. Why do you need to do anything manually with ZK?
| 2021-01-13T17:08:49 | 2021-01-13T17:09:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,603,238 | 18,924 | alexey-milovidov | @snar This is FreeBSD build, see `clang-11-freebsd`. Just download the binary and run. | 2021-01-13T17:31:35 | 2021-01-13T17:31:35 | {} | MEMBER |
ClickHouse | ClickHouse | 759,618,947 | 18,975 | filipecaixeta | Solved
My function was returning `return DataTypeUInt8().createColumnConst(elem.column->size(), 1u);`
When I added `return DataTypeUInt8().createColumnConst(elem.column->size(), 1u)->convertToFullColumnIfConst();` it appeared in the `WHERE` clause.
It looks like there is an optimization step running before this `que... | 2021-01-13T17:59:32 | 2021-01-13T17:59:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,720,260 | 18,853 | azat | >And I believe that batching mode overhead can be eliminated by removing extra file reading.
Confirmed, w/o splitting by block header it is 2x faster (since does not requires extra file reading after other patches in this series) | 2021-01-13T20:28:25 | 2021-01-13T20:28:25 | {} | MEMBER |
ClickHouse | ClickHouse | 759,974,944 | 19,015 | alexey-milovidov | > https://clickhouse-test-reports.s3.yandex.net/19015/1d9a3a492a21ca83662e473466c46e5aa1051669/integration_tests_(release).html#fail1
@CurtizJ | 2021-01-14T07:11:18 | 2021-01-14T07:11:18 | {} | MEMBER |
ClickHouse | ClickHouse | 760,020,043 | 18,954 | alexey-milovidov | > P.S. maybe we should start publishing alpine-based images?
@filimonov Are there any reasons to not use it as default and the only option? | 2021-01-14T08:34:03 | 2021-01-14T08:34:50 | {} | MEMBER |
ClickHouse | ClickHouse | 759,196,942 | 18,791 | sundy-li | All tests look pretty good, read for merge cc @Avogar | 2021-01-13T04:26:35 | 2021-01-13T04:26:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,445,289 | 18,999 | KochetovNicolai | `cast` uses fast and not very precise parser from string:
https://github.com/ClickHouse/ClickHouse/blob/d88d3de0d18f4879c8994b03ef7c4b957b6a83cf/src/IO/readFloatText.h#L317
Maybe add setting for using more precise parser? | 2021-01-13T13:23:42 | 2021-01-13T13:23:42 | {} | MEMBER |
ClickHouse | ClickHouse | 759,760,004 | 18,982 | filimonov | BTW why it was added? (here https://github.com/ClickHouse/ClickHouse/pull/6375/files , so - quite long ago and was not affecting users before )
UPD (nikitamikhaylov): Moved the issue number to the description because Github recognizes it as a PR number. Just not to confuse people | 2021-01-13T21:48:37 | 2021-01-13T22:35:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,906,853 | 18,958 | ElapsedSoul | Compile completed. | 2021-01-14T03:47:15 | 2021-01-14T03:47:15 | {} | NONE |
ClickHouse | ClickHouse | 759,973,722 | 19,011 | alesapin | `test_ttl_replicated/test.py::test_ttl_empty_parts` -- not related to changes. | 2021-01-14T07:08:12 | 2021-01-14T07:08:12 | {} | MEMBER |
ClickHouse | ClickHouse | 760,025,026 | 17,951 | alexey-milovidov | We decided to investigate and fix the recently added assertion about marks before the release: #18913 | 2021-01-14T08:43:26 | 2021-01-14T08:43:26 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 760,060,968 | 18,954 | filimonov | Compatibility issues.
1) Some people (may be even some CI tools) can build their own containers based on the official own relying on the ubuntu commands and features which work differently on alpine.
For example imagine they install some odbc drivers using `apt-get` commands.
2) Also (ust came to my mind ri... | 2021-01-14T09:13:29 | 2021-01-14T09:13:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,223,730 | 17,623 | chang0391 | Is grace hash join the same thing as shulffe join of spark? | 2021-01-13T05:55:14 | 2021-01-13T05:55:14 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 759,249,109 | 18,928 | alesapin | Stress test failure was fixed in https://github.com/ClickHouse/ClickHouse/pull/18964 | 2021-01-13T06:59:41 | 2021-01-13T06:59:41 | {} | MEMBER |
ClickHouse | ClickHouse | 759,425,938 | 18,910 | alexey-milovidov | I have a feeling that `WithNames` meaning is non-obvious, it's not clear whether it is
- normalizing names as well (1);
- keeping names (2).
(in fact it is 2) | 2021-01-13T12:46:25 | 2021-01-13T12:46:25 | {} | MEMBER |
ClickHouse | ClickHouse | 759,475,350 | 18,818 | alexey-milovidov | @templarzq Looks like you forgot to commit test files...
This is the only small thing remaining before merge. | 2021-01-13T14:14:18 | 2021-01-13T14:14:18 | {} | MEMBER |
ClickHouse | ClickHouse | 759,540,103 | 7,192 | den-crane | > Does `DROP PARTITION` propagate changes into the `MATERIALIZED VIEW`s?
No. `MATERIALIZED VIEW` is an insert trigger and it watches only inserts into base-table.
https://den-crane.github.io/Everything_you_should_know_about_materialized_views_commented.pdf
https://youtu.be/ckChUkC3Pns?t=9326 | 2021-01-13T15:54:20 | 2021-01-13T15:54:20 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,586,123 | 19,014 | den-crane | You cannot use this method because it does not copy autoincrement values which are used for part block numbers.
There are much simpler way. You can migrate ZK cluster by adding new ZK nodes as followers.
```
Here is a plan for ZK 3.4.9 (no dynamic reconfiguration):
1. Configure the 3 new ZK nodes as a cluster o... | 2021-01-13T17:02:33 | 2021-01-13T17:02:33 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,356,993 | 18,861 | hexiaoting | First, build clickhouse using clang with option `-fno-limit-debug-info`. then we can get the `DB::ColumnArray` type in gdb.
build command:
```
cmake ../src -DCLICKHOUSE_SPLIT_BINARY=1 -DUSE_UNWIND=1 -DUSE_DEBUG_HELPERS=1 -DCMAKE_INSTALL_PREFIX=/usr -DCMAKE_VERBOSE_MAKEFILE=ON -DCMAKE_BUILD_TYPE=Debug -DCMAKE_INSTALL... | 2021-01-13T10:28:48 | 2021-01-13T10:28:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,420,549 | 18,975 | KochetovNicolai | Hi, @filipecaixeta .
Could you please show an example where you want to find those functions?
`SelectQueryInfo` contains different information that may be helpful to execute query. E.g. `SelectQueryInfo::query` is a query AST, which should contain unchanged query with `WHERE` clause. It is used by `MergeTree` in PK... | 2021-01-13T12:35:43 | 2021-01-13T12:35:43 | {} | MEMBER |
ClickHouse | ClickHouse | 759,426,440 | 15,889 | KochetovNicolai | Ok, some tests fails for debug build. | 2021-01-13T12:47:16 | 2021-01-13T12:47:16 | {} | MEMBER |
ClickHouse | ClickHouse | 759,463,092 | 3,700 | erste | > changed Date EXPRESSION cast(changed as timestamp)
@van4oza попробуйте c кавычками:
```
changed Date EXPRESSION 'cast(changed as timestamp)'
``` | 2021-01-13T13:54:29 | 2021-01-13T13:54:29 | {} | NONE |
ClickHouse | ClickHouse | 759,508,570 | 19,014 | NotAndD | I see, thank you! I'll try what you suggested with a mix of backups and restores to split my Zookeeper cluster into two. I missed that part of the documentation, I apologize. | 2021-01-13T15:06:29 | 2021-01-13T15:06:29 | {} | NONE |
ClickHouse | ClickHouse | 759,861,134 | 15,085 | UnamedRus | It also would be useful if we have used a -Resample combinator as it produces constant sized bucketed arrays.
```
CREATE TABLE test_tbl_2_mv
(
`datetime` UInt32,
`state` SimpleAggregateFunction(maxForEach, Array(UInt8))
)
ENGINE = AggregatingMergeTree
PARTITION BY tuple()
ORDER BY intDiv(datetime, 30... | 2021-01-14T01:18:42 | 2021-01-14T01:18:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,972,850 | 19,026 | alexey-milovidov | They are not dangling references according to the C++ standard.
Note that we also run functional, integration, unit and stress tests under ASan. | 2021-01-14T07:06:16 | 2021-01-14T07:08:05 | {} | MEMBER |
ClickHouse | ClickHouse | 760,202,867 | 19,066 | den-crane | 1. 20.5.2.7 is out of support.
2. Check system.merges -- long merge can block `drop column`
3. use `system stop merges tablename; drop column; system start merges tablename;` if you need to drop column immediately.
| 2021-01-14T13:40:12 | 2021-01-14T13:40:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,502,216 | 18,913 | alesapin | Got it: https://github.com/ClickHouse/ClickHouse/pull/19011
```
2021.01.13 17:39:45.410267 [ 5781 ] {} <Fatal> : Logical error: 'Incomplete granules are not allowed while blocks are granules size. Mark number 9 (rows 8192), rows written in last mark 0, rows to write in last mark from block 4434 (from row 8192), total... | 2021-01-13T14:57:00 | 2021-01-13T14:57:00 | {} | MEMBER |
ClickHouse | ClickHouse | 759,761,718 | 18,999 | filimonov | Yes, let's do it. There are cases when precision is more important than speed. | 2021-01-13T21:52:20 | 2021-01-13T21:52:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,885,446 | 18,958 | ElapsedSoul | I got this warning error:
```
[105/2077] Building CXX object src/CMakeFiles/dbms.dir/Storages/MergeTree/DataPartsExchange.cpp.o
FAILED: src/CMakeFiles/dbms.dir/Storages/MergeTree/DataPartsExchange.cpp.o
/usr/local/bin/clang++ -DCONFIGDIR=\"\" -DHAVE_CONFIG_H -DLIBSASL_EXPORTS=1 -DLZ4_DISABLE_DEPRECATE_WARNINGS=1 -... | 2021-01-14T02:33:55 | 2021-01-14T02:33:55 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 759,913,908 | 9,157 | zhangjmruc | @alexey-milovidov I had a design idea for the "fallback table". Would you please let me know your opinion on this?
This can be implemented by extending the Poco::Channel class, and adding ConnectionChannel class which maintains a connection with the database server and writes parse errors via logger into the error tab... | 2021-01-14T04:14:52 | 2021-01-14T04:14:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,160,407 | 18,465 | ghost | @alexey-milovidov
Could you please review? | 2021-01-13T02:23:24 | 2021-01-13T02:23:24 | {} | NONE |
ClickHouse | ClickHouse | 759,415,287 | 18,991 | bohutang | MaterializeMySQL EVENT_VERSION_V4 required.
https://github.com/ClickHouse/ClickHouse/blob/master/src/Core/MySQL/MySQLReplication.cpp#L54
| 2021-01-13T12:25:12 | 2021-01-13T12:25:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,425,107 | 18,910 | alexey-milovidov | ANTLR parser is not yet complete.
You can add an exception to skip_list.json | 2021-01-13T12:44:42 | 2021-01-13T12:44:42 | {} | MEMBER |
ClickHouse | ClickHouse | 759,467,486 | 18,986 | den-crane | duplicate https://github.com/ClickHouse/ClickHouse/issues/13274 | 2021-01-13T14:01:26 | 2021-01-13T14:01:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,584,930 | 17,955 | BhavyaRajSharma | @den-crane,
Click house is giving the same problem even with an Ordinary database.
Based on the observations, I am getting this problem when I am trying to drop the replicated tables. After dropping the tables I delete the zookeeper nodes manually. After deleting the zookeeper nodes manually and recreating the repli... | 2021-01-13T17:00:43 | 2021-01-13T17:00:43 | {} | NONE |
ClickHouse | ClickHouse | 759,980,922 | 17,955 | filimonov | For atomic DB - nice practice is to use table uuid in ZooKeeper path.
https://github.com/ClickHouse/ClickHouse/blob/ac989f3a8fe8e2135e25ac7d6e340b048092e906/tests/integration/test_distributed_ddl_on_cross_replication/test.py#L92 | 2021-01-14T07:25:09 | 2021-01-14T07:25:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 760,024,296 | 19,013 | alexey-milovidov | > test_ttl_replicated/test.py::test_ttl_empty_parts
Broken, @CurtizJ @alesapin
> Logical error: 'Incomplete granules are not allowed while blocks are granules size. Mark number 33 (rows 8192), rows written in last mark 0, rows to write in last mark from block 1047 (from row 8
192), total marks currently 34'.
... | 2021-01-14T08:42:01 | 2021-01-14T08:42:01 | {} | MEMBER |
ClickHouse | ClickHouse | 760,150,297 | 18,954 | filimonov | One more thing to change - prepare clickhouse-client image for alpine (BTW - as for me that separate client image is a bit useless, as you can use clickhouse-server for everything). | 2021-01-14T11:55:38 | 2021-01-14T11:55:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,268,970 | 18,991 | awesomeleo | @BohuTANG Could you take a look when you are free? | 2021-01-13T07:45:49 | 2021-01-13T07:46:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,332,863 | 18,887 | KochetovNicolai | Probably fixed by #18964 | 2021-01-13T09:46:13 | 2021-01-13T09:46:13 | {} | MEMBER |
ClickHouse | ClickHouse | 759,398,660 | 18,991 | bohutang | Enabling log_bin_use_v1_row_events=1 causes mysqld to write the binary log using Version 1 logging events.
BTW, Why enable this setting?
| 2021-01-13T11:50:24 | 2021-01-13T11:50:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,451,467 | 18,929 | sevirov | @Akazz Please, review this PR. | 2021-01-13T13:34:36 | 2021-01-13T13:34:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,495,534 | 19,014 | genzgd | This is not recommended at all for ClickHouse: https://github.com/ksprojects/zkcopy/issues/15. The documentation suggests adding additional nodes to your starting zookeeper cluster then breaking it apart.
The problem has to do with sequential nodes, which ClickHouse uses for the replication queue and log data. The... | 2021-01-13T14:46:47 | 2021-01-13T14:46:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 759,607,803 | 17,955 | BhavyaRajSharma | @den-crane,
In the above-mentioned click house version, whenever I drop the replicated tables, it did not deletes the zookeeper nodes. Without dropping the zookeeper nodes if I try to recreate the replicated tables, it shows the node already exists. | 2021-01-13T17:39:37 | 2021-01-13T17:39:37 | {} | NONE |
ClickHouse | ClickHouse | 760,035,324 | 18,723 | alexey-milovidov | I cannot understand why "Docs check" has failed. Just removed the link. | 2021-01-14T08:53:04 | 2021-01-14T08:53:04 | {} | MEMBER |
ClickHouse | ClickHouse | 760,099,146 | 18,954 | alexey-milovidov | > Some people (may be even some CI tools)
We can keep scripts for Ubuntu container but publish Alpine container as the official.
> BTW - maybe we should create clickhouse user in the Dockerfile with fixed uid/gid ahead / before deb installallation
Yes, we can tune `Install.cpp` for it. | 2021-01-14T10:15:58 | 2021-01-14T10:15:58 | {} | MEMBER |
ClickHouse | ClickHouse | 768,895,627 | 19,690 | FavonianKong | @alexey-milovidov Sorry to bother you again， I installed `pretty-printers` as you suggested, and It did work well in the test program:
```
(gdb) r
Starting program: /home/algorithm/testO0_4.out
Breakpoint 1, main () at /home/algorithm/testO0.cpp:6
6 string s= "abc" ;
(gdb) n
8 map<int, strin... | 2021-01-28T08:45:05 | 2021-01-28T08:51:16 | {} | NONE |
ClickHouse | ClickHouse | 768,904,854 | 19,487 | ucasfl | > I have to validate this message manually. It's suspected that it can output two consecutive whitespaces.
```
:) select xyz
SELECT xyz
Query id: 34565bc0-8d25-46bf-a835-fbf0b0031413
0 rows in set. Elapsed: 0.039 sec.
Received exception from server (version 21.2.1):
Code: 47. DB::Exception: Received ... | 2021-01-28T09:00:30 | 2021-01-28T09:00:30 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 769,207,301 | 19,735 | kitaisreal | 20.10 is no longer supported. | 2021-01-28T16:30:45 | 2021-01-28T16:30:45 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.