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 | 545,536,245 | 7,453 | ghost | pthread_cond_wait and pthread_cond_destroy were called on the same conditional variable. The main thread was unwinding stack and called ~ZooKeeperImpl, the receiveThread got exception and called finalize.
Stacktrace:
```
Thread 6 (Thread 0x7fe39cd03bc0 (LWP 629694)):
#0 0x00007fe39c4d1404 in pthread_cond_wait@... | 2019-10-23T16:50:52 | 2019-10-23T16:51:33 | {} | NONE |
ClickHouse | ClickHouse | 545,669,510 | 4,119 | filimonov | Implemented long ago
https://clickhouse.yandex/docs/en/operations/table_engines/mergetree/#table_engine-mergetree-ttl | 2019-10-23T22:54:03 | 2019-10-23T22:54:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,712,910 | 7,449 | alexey-milovidov | Can you provide full build scenario (all modifications to code and all CMake options) to reproduce this issue? | 2019-10-24T02:22:30 | 2019-10-24T02:22:30 | {} | MEMBER |
ClickHouse | ClickHouse | 545,817,995 | 7,434 | excitoon | @alesapin
> It's very easy to test with the integration test. Just move one part with `ALTER TABLE MOVE PART` and then try to move partition.
3debdc211904eb0b5fe6cca1a4af99d8db635892 | 2019-10-24T08:55:15 | 2019-10-24T08:55:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,913,350 | 7,327 | akuzm | Can we add some kind of test for this code? | 2019-10-24T13:18:05 | 2019-10-24T13:18:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,942,916 | 7,433 | alexey-milovidov | We still need a test. | 2019-10-28T13:21:59 | 2019-10-28T13:21:59 | {} | MEMBER |
ClickHouse | ClickHouse | 541,256,500 | 7,287 | tmartin-gh | Attaching example program that can reproduce the issue
[bug_test.cpp.txt](https://github.com/ClickHouse/ClickHouse/files/3719906/bug_test.cpp.txt)
caused:
```
2019.10.11 23:31:50.123645 [ 29 ] {} <Error> void DB::BackgroundProcessingPool::threadFunction(): Code: 49, e.displayText() = DB::Exception: Tagging alre... | 2019-10-11T23:35:30 | 2019-10-23T22:37:40 | {} | NONE |
ClickHouse | ClickHouse | 545,945,953 | 7,469 | den-crane | duplicate https://github.com/ClickHouse/ClickHouse/issues/5131 | 2019-10-24T14:31:00 | 2019-10-24T14:31:00 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,121,576 | 7,470 | Felixoid | So, I've checked `ALTER .. DETACH|ATTACH|DROP PARTITION ID` and it works.
Looks like this one fixes #4735 | 2019-10-24T22:03:11 | 2019-10-24T22:03:11 | {} | MEMBER |
ClickHouse | ClickHouse | 546,173,285 | 7,417 | yedy99 | @tutuhuagong can show the detail of how to build on a mac? thanks | 2019-10-25T02:24:13 | 2019-10-25T02:24:13 | {} | NONE |
ClickHouse | ClickHouse | 546,301,402 | 2,667 | develar | Doesn't work.
```
default> alter table report modify column product Enum8('IU' = 1, 'WS' = 2, 'PS' = 3)
[2019-10-25 12:29:44] Code: 44, e.displayText() = DB::Exception: Trying to ALTER key column product (version 19.15.2.2 (official build))
```
Current type: `Enum8('IU' = 1, 'WS' = 2)`
Table engine:
`engin... | 2019-10-25T10:36:56 | 2019-10-25T10:38:01 | {} | NONE |
ClickHouse | ClickHouse | 546,314,450 | 7,380 | akuzm | Let's add an explanation of this function to `docs/en/query_language/functions/array_functions.md`. | 2019-10-25T11:19:55 | 2019-10-25T11:19:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,333,356 | 7,488 | traceon | ```
clickhouse-client --param_x='Null-not-am-I'
Code: 457. DB::Exception: Value Null-not-am-I cannot be parsed as Nullable(String) for query parameter 'x'
``` | 2019-10-25T12:26:40 | 2019-10-25T12:26:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,604,637 | 7,380 | hczhcz | Performance tests on different filling patterns
```sql
SELECT arraySlice(arrayFill(x -> ((x % 2) >= 0), range(100000000)), 1, 10)
┌─arraySlice(arrayFill(lambda(tuple(x), greaterOrEquals(modulo(x, 2), 0)), range(100000000)), 1, 10)─┐
│ [0,1,2,3,4,5,6,7,8,9] ... | 2019-10-26T13:53:43 | 2019-10-26T13:54:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,809,882 | 7,499 | hczhcz | The expression `col1 = 'CH'` will yield a "condition column" of boolean values. Physically, it is one or more memory blocks containing UInt8 values.
Then, we can perform a for-loop within the range of the row numbers. Let the row number to be "i", the i-th element in the condition column will indicate whether the i-th... | 2019-10-28T06:30:46 | 2019-10-28T06:30:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,842,571 | 7,491 | alex-zaitsev | > Should we add an integration test for this?
We can, but with MySQL the problem with SQL_FLOAT was not reproducible, because it reports different data types. So extending MySQL integration test would not help. We've faced this issue with MSSQL. I am not sure it makes sense to have MSSQL in integration tests just fo... | 2019-10-28T08:30:25 | 2019-10-28T08:30:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,867,904 | 7,502 | filimonov | ClickHouse can decrease a number of working threads below `max_threads` automatically if selected datasize is small enough. | 2019-10-28T09:41:52 | 2019-10-28T09:41:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,712,599 | 7,449 | alexey-milovidov | Note that shared linking is not tested and not recommended for production usage. | 2019-10-24T02:20:57 | 2019-10-24T02:20:57 | {} | MEMBER |
ClickHouse | ClickHouse | 545,921,169 | 7,292 | akuzm | String hash map improvements were merged to #6243, and the work on iterators continues in #7459. | 2019-10-24T13:36:04 | 2019-10-24T13:36:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,082,126 | 6,836 | 4ertus2 | Need to ckeck columns in subselect results. Related #4567 | 2019-10-24T20:09:41 | 2019-10-24T20:09:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,425,300 | 6,710 | jcdang | @akuzm We can close this issue. The company I work for has decided to maintain a forked version of clickhouse. | 2019-10-23T12:43:53 | 2019-10-23T12:43:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,649,452 | 7,450 | filimonov | New row appeared in system.storage_policies and test checking that failed... | 2019-10-23T21:44:38 | 2019-10-23T21:44:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,091,322 | 7,473 | den-crane | Imagine you've created a table
`create table T(A Int, B Int) order by A`
`insert into T values(1,1) ` --- first part no skip index
`alter table add index min_max(B);`
`insert into T values(1,2)` --- second part with skip index file with B_min_max=(2,2)
And you run a query `select * from T where B>10000000`
... | 2019-10-24T20:34:15 | 2019-10-24T20:47:00 | {
"+1": 5
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,145,138 | 3,529 | stale[bot] | This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
| 2019-10-24T23:50:11 | 2019-10-24T23:50:11 | {} | NONE |
ClickHouse | ClickHouse | 546,295,999 | 7,482 | alexey-milovidov | The behaviour is inconsistent and also we don't process the case with constant columns correctly.
```
milovidov@milovidov-nix:~$ clickhouse local --query "SELECT * FROM system.collations"
af
am
ar
as
az
be
bg
bn
bo
bs
bs_Cyrl
ca
chr
cs ... | 2019-10-25T10:19:14 | 2019-10-25T10:19:14 | {} | MEMBER |
ClickHouse | ClickHouse | 546,721,810 | 7,498 | zhang2014 | This script is based on the assumption that documents in other languages are fully synchronized with the en document at a commit. | 2019-10-27T18:40:12 | 2019-10-27T18:40:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,493,857 | 7,455 | alexey-milovidov | BTW, I'm Ok if query_log, query_thread_log, trace_log will be reused for background processes: merges, fetches, etc. | 2019-10-23T15:16:12 | 2019-10-23T15:16:12 | {} | MEMBER |
ClickHouse | ClickHouse | 545,520,087 | 7,453 | alexey-milovidov | Could you please describe the scenario that reproduces the issue?
The fix looks too heavyweight. We need to understand the root cause of the issue. | 2019-10-23T16:11:27 | 2019-10-23T16:11:27 | {} | MEMBER |
ClickHouse | ClickHouse | 545,693,340 | 7,392 | amosbird | `enable_scalar_subquery_optimization` is set to `true` by default in order to get the perf test result. So there is a backward incompatibility and one might update the settings profile and set it to false so that a cluster installation can be upgraded without downtime. | 2019-10-24T00:44:22 | 2019-10-24T00:44:22 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 545,837,168 | 7,358 | akuzm | `00094_union_race_conditions_5` fails, is this relevant? | 2019-10-24T09:40:49 | 2019-10-24T09:40:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,710,154 | 7,449 | tencupofkaiwater | 2019.10.23 17:37:42.936693 [ 134 ] {} <Fatal> BaseDaemon: ########################################
2019.10.23 17:37:42.936785 [ 134 ] {} <Fatal> BaseDaemon: (version 19.15.2.2) (from thread 101) Received signal Broken pipe (13).
2019.10.23 17:37:42.936828 [ 134 ] {} <Fatal> BaseDaemon:
2019.10.23 17:37:42.936866 [ ... | 2019-10-24T02:08:43 | 2019-10-24T02:08:43 | {} | NONE |
ClickHouse | ClickHouse | 545,780,374 | 7,463 | filimonov | Checked on official build 19.15.1.1374. | 2019-10-24T07:14:52 | 2019-10-24T07:14:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,830,681 | 7,459 | akuzm | Do you have a benchmark for which this change has the most effect? I'll work on the more precise performance comparison in the CI, so that we can see the performance effect clearly.
The code still looks somewhat complicated to me, but if the performance improvement is noticeable, we can improve that. | 2019-10-24T09:25:17 | 2019-10-24T09:25:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,831,134 | 7,471 | filimonov | > I want to remove the element [2,5]
```
WITH [0, 1, 2, 3, 2, 4, 5, 6] AS arr
SELECT
arr,
arrayFilter((element, index) -> index NOT IN (2, 5), arr, arrayEnumerate(arr))
```
> suggest removing any nth element in an array.
removing every 5th (4,9,14,19):
```
WITH range(20) AS arr
SELECT
ar... | 2019-10-24T09:26:20 | 2019-10-24T09:26:20 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,867,879 | 7,240 | BayoNet | Do we need to document this? | 2019-10-24T11:07:58 | 2019-10-24T11:08:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,857,240 | 6,828 | Tigro | Hi bzed.
Sorry, but our repo has non standart Debian pool structure to save old versions of clickhouse software.
So "deb http://debian.foo/debian stretch main" will not work one of these fine days.
To mirror your can use debmirror (with many options to override standart behavior), but the best way to use rsync
... | 2019-10-24T10:35:40 | 2019-10-24T13:29:03 | {} | NONE |
ClickHouse | ClickHouse | 545,993,117 | 7,470 | alexey-milovidov | Is it also applicable also to all ALTER PARTITION queries? | 2019-10-24T16:16:21 | 2019-10-24T16:16:21 | {} | MEMBER |
ClickHouse | ClickHouse | 546,101,415 | 1,287 | touseefrafique | We are also facing this issue. Does anyone knows any plans on the fix | 2019-10-24T21:00:29 | 2019-10-24T21:00:29 | {} | NONE |
ClickHouse | ClickHouse | 545,419,121 | 4,190 | uncletimmy3 | Hi! Dont close issue pls :)
Or may be this has been fixed? | 2019-10-23T12:27:26 | 2019-10-23T12:27:26 | {} | NONE |
ClickHouse | ClickHouse | 546,519,178 | 7,484 | filimonov | Just a silly newline character at the end of each description in system.merge_tree_settings :)
`Select * from system.merge_tree_settings` | 2019-10-25T21:26:33 | 2019-10-25T21:26:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,639,858 | 3,571 | stale[bot] | This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
| 2019-10-26T21:13:27 | 2019-10-26T21:13:27 | {} | NONE |
ClickHouse | ClickHouse | 545,652,578 | 7,449 | filimonov | It look like you're build differs from standard.
Can you show your system.build_options?
Can you try to reproduce the problem with official build?
About problem itself
1) Can you please show how exactly do you insert data (it doesn't use pipes for client-server communication, so bit cryptic).
2) Can you please g... | 2019-10-23T21:54:36 | 2019-10-23T21:56:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,812,905 | 7,471 | slimtom95 | to remove the nth element, how about:
`arrayConcat(arraySlice(arr, 1, n - 1), arraySlice(arr, n + 1, length(arr) - n + 1))`
```
select [1,2,3,4,5,6];
with [1,2,3,4,5,6] as arr select arrayConcat(arraySlice(arr, 1, 3 - 1), arraySlice(arr, 3 + 1, length(arr) - 3 + 1));
```
waiting for better solution... | 2019-10-24T08:43:04 | 2019-10-24T08:46:50 | {} | NONE |
ClickHouse | ClickHouse | 545,879,741 | 7,240 | alexey-milovidov | Yes. | 2019-10-24T11:44:18 | 2019-10-24T11:44:18 | {} | MEMBER |
ClickHouse | ClickHouse | 545,945,108 | 7,477 | alexey-milovidov | Approved.
You should replace thread_id with system thread id everywhere and get rid of the columns `os_thread_id` in system tables. | 2019-10-24T14:29:14 | 2019-10-24T14:29:14 | {} | MEMBER |
ClickHouse | ClickHouse | 546,159,945 | 7,403 | RaulGuo | 日志目录下的error日志里有什么？如果日志目录下什么也没有，说明是日志目录的权限的问题。 | 2019-10-25T01:11:03 | 2019-10-25T01:11:03 | {} | NONE |
ClickHouse | ClickHouse | 546,303,474 | 2,620 | filimonov | replicated_deduplication_window & replicated_deduplication_window_seconds are merge_tree settings, you can adjust them per table using SETTINGS statement, or globally - in <merge_tree> section of config.xml.
insert_deduplicate is session setting, so you can bypass that deduplication logic for one particular insert.
... | 2019-10-25T10:43:57 | 2019-10-25T10:43:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 544,250,669 | 7,369 | vdimir | As I understand ClickHouse have graphite support ([GraphiteWriter](https://github.com/ClickHouse/ClickHouse/blob/09f3d68f6e3adf74bd8106c6bb9532a4b9e3c82d/libs/libdaemon/include/daemon/GraphiteWriter.h#L15), [MetricsTransmitter](https://github.com/ClickHouse/ClickHouse/blob/4952c6bc65a42be52627ae0fdd66b3fa7d400c04/dbms/... | 2019-10-20T13:08:04 | 2019-10-24T12:48:03 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 546,295,317 | 7,482 | alexey-milovidov | Collation names are from ICU library.
Let's check what is the collation name for english language. | 2019-10-25T10:16:59 | 2019-10-25T10:16:59 | {} | MEMBER |
ClickHouse | ClickHouse | 545,398,614 | 7,423 | excitoon | @alexey-milovidov:
> Missing integration test.
https://github.com/ClickHouse/ClickHouse/pull/7450 | 2019-10-23T11:27:50 | 2019-10-23T11:27:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,709,750 | 7,449 | tencupofkaiwater | -O2 -pipe -g | 2019-10-24T02:06:33 | 2019-10-24T02:06:33 | {} | NONE |
ClickHouse | ClickHouse | 546,133,436 | 6,973 | Tri0L | Nope, thanks a lot! | 2019-10-24T22:50:42 | 2019-10-24T22:50:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,828,641 | 6,357 | arsis1 | May I ask, the plan described here (https://clickhouse.yandex/blog/en/concept-cloud-mergetree-tables) has some progress ? or just on todo-list ? Current distributed engine is not the one, and we need auto-rebalancing when add node. | 2019-10-28T07:44:56 | 2019-10-28T07:44:56 | {} | NONE |
ClickHouse | ClickHouse | 546,864,813 | 7,495 | filimonov | /cc @yurriy | 2019-10-28T09:33:24 | 2019-10-28T09:33:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,936,699 | 7,505 | alexey-milovidov | We should also implement the same support for `File`, `URL` and `S3` storages.
Determining the compression by file extension is Ok. But we should provide a setting to more fine control over it. This setting should have the following values: `none`, `gz` (`bz2`, `xz` are possible to also implement later) and `auto`. ... | 2019-10-28T13:06:24 | 2019-10-28T13:06:53 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 545,448,374 | 7,443 | den-crane | @vpanfilov ha, that what I asked for.
uniqMergeState works.
Great. Thank you. | 2019-10-23T13:38:38 | 2019-10-23T13:38:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,626,378 | 7,417 | pelletier | If it helps, it works for me with the following:
```
$ sw_vers
ProductName: Mac OS X
ProductVersion: 10.14.6
BuildVersion: 18G103
$ ninja --version
1.9.0
$ cmake --version
cmake version 3.15.4
$ gcc-8 --version
gcc-8 (Homebrew GCC 8.3.0) 8.3.0
$ git branch
* stable
$ git rev-parse HEAD
ea739d193229df3f... | 2019-10-23T20:40:10 | 2019-10-23T20:42:38 | {} | NONE |
ClickHouse | ClickHouse | 545,654,301 | 7,445 | filimonov | All tables from mergetree family use the same on-disk format. And parts can be moved between different variants of engines. Of course you still need to have requirement column set. | 2019-10-23T22:00:27 | 2019-10-23T22:00:27 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,715,493 | 7,460 | alexey-milovidov | BTW, We don't need to `install` any libraries from contrib at all. | 2019-10-24T02:35:15 | 2019-10-24T02:35:15 | {} | MEMBER |
ClickHouse | ClickHouse | 545,720,516 | 7,449 | alexey-milovidov | `/usr/lib64/libPocoNet.so.60`
It means you are using a library from OS instead of `contrib`. Also it means that you are using shared linking istead of static linking. It can only be enabled with custom CMake options and it is unsupported build variant. (You have manually enabled two sources of trouble.) | 2019-10-24T03:00:11 | 2019-10-24T03:01:21 | {} | MEMBER |
ClickHouse | ClickHouse | 545,999,019 | 7,470 | Felixoid | I didn't check yet. It's said so on the linked page
It works with optimize, though | 2019-10-24T16:31:11 | 2019-10-24T16:31:52 | {} | MEMBER |
ClickHouse | ClickHouse | 546,171,548 | 7,417 | ghost | After a long struggle trying to build ClickHouse on macOS 10.15, finally, this works for me:
```
cd ${YOUR_CLICKHOUSE_SRC_PATH}/docker/builder/;
make build;
```
Docker helps a lot! | 2019-10-25T02:15:50 | 2019-10-25T02:41:14 | {} | NONE |
ClickHouse | ClickHouse | 546,349,420 | 7,489 | den-crane | CAST('Null', 'Nullable(String)') -- it's a string with text Null.
```
SELECT
CAST('Null', 'Nullable(String)'),
NULL
┌─CAST('Null', 'Nullable(String)')─┬─NULL─┐
│ Null │ ᴺᵁᴸᴸ │
└──────────────────────────────────┴──────┘
```
it's how cast works from strings to Int32
... | 2019-10-25T13:16:11 | 2019-10-25T13:16:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,420,725 | 6,710 | akuzm | What I gather from a cursory read of the discussion is that this functionality is controversial and that we should work on #6294 first, and then reconsider.
@filimonov @jcdang is this OK with you? | 2019-10-23T12:31:43 | 2019-10-23T12:31:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 539,961,867 | 7,229 | filimonov | @Tasselmi it sound like you are looking for that:
#1616
#2741
#3483
BTW. @BayoNet looks like `from_env` is not documented. | 2019-10-09T11:35:10 | 2019-10-23T23:15:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,719,779 | 7,449 | tencupofkaiwater | CMake options : -O2 -pipe -g
Don't modify any code.
| 2019-10-24T02:56:37 | 2019-10-24T02:56:37 | {} | NONE |
ClickHouse | ClickHouse | 545,920,756 | 7,267 | akuzm | The work on iterators is continued in #7459. | 2019-10-24T13:35:13 | 2019-10-24T13:35:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,593,394 | 3,140 | stale[bot] | This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
| 2019-10-26T11:13:27 | 2019-10-26T11:13:27 | {} | NONE |
ClickHouse | ClickHouse | 546,823,381 | 7,499 | chu1070y | It`s very helpful. Thanks. | 2019-10-28T07:24:56 | 2019-10-28T07:24:56 | {} | NONE |
ClickHouse | ClickHouse | 546,934,359 | 7,498 | blinkov | > This script is based on the assumption that documents in other languages are fully synchronized with the en document at a commit.
@zhang2014 it's better to add this comment to `--help` output or at least in code comments | 2019-10-28T12:59:44 | 2019-10-28T12:59:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,867,017 | 7,209 | BayoNet | Do we need any docs here? | 2019-10-24T11:05:07 | 2019-10-24T11:05:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,037,099 | 7,360 | alesapin | my own tests failed :(
I have missed config file :( | 2019-10-24T18:09:33 | 2019-10-24T20:09:11 | {} | MEMBER |
ClickHouse | ClickHouse | 546,295,424 | 7,482 | alexey-milovidov | I remember that we have `system.collations` table. | 2019-10-25T10:17:18 | 2019-10-25T10:17:18 | {} | MEMBER |
ClickHouse | ClickHouse | 546,402,392 | 7,360 | alesapin | @alexey-milovidov recommend to merge this PR. | 2019-10-25T15:34:05 | 2019-10-25T15:34:05 | {
"hooray": 1
} | MEMBER |
ClickHouse | ClickHouse | 545,507,251 | 6,661 | romankos | > Should be fixed in v19.17+
@abyss7 But at the moment we have 19.15.2.2, 2019-10-01 version | 2019-10-23T15:43:45 | 2019-10-23T15:44:20 | {} | NONE |
ClickHouse | ClickHouse | 545,658,235 | 7,408 | filimonov | https://github.com/flyway/flyway/pull/2166
https://github.com/filimonov/ClickHouse/wiki/Clickhouse-database-schema-migrations
And use ON CLUSTER queries.
It will not work if you will try to add new replica with incremental schema (to add new replica you need to use on that the latest schema, not incremental ... | 2019-10-23T22:12:57 | 2019-10-23T22:12:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,788,439 | 7,467 | filimonov | About UDF - see #11
For now the only possible way to create custom function is just add the to ClickHouse codebase.
There are currently some UDF-related concepts & ideas, so maybe UDF support will be added in further versions.
`arraySplit` was added in 19.16 (currenlty prestable), see https://github.com/Clic... | 2019-10-24T07:38:21 | 2019-10-24T07:39:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,834,251 | 7,469 | filimonov | Thanks for good report. Confirmed & reproduced:
```
(SelectExecutor): Key condition: (column 0 in [1546297200, +inf)), (column 0 in (-inf, 1577746800]), and
(SelectExecutor): MinMax index condition: (column 0 in [1546297200, +inf)), (column 0 in (-inf, 1577746800]), and
(SelectExecutor): Selected 0 parts by date,... | 2019-10-24T09:33:51 | 2019-10-24T09:35:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,857,113 | 6,610 | ghuname | Yes, I missed the "STEP" clause. Thanks.
Anyway, it would be nice to put arbitrary missing value.
In example above:
SELECT * FROM t ORDER BY date WITH FILL FROM toDate('2019-06-05') TO toDate('2019-06-16')
counter is filled with zero and field with empty string.
It would be nice to be able to fill missing valu... | 2019-10-24T10:35:20 | 2019-10-24T10:35:20 | {} | NONE |
ClickHouse | ClickHouse | 545,907,792 | 5,721 | lzy305 | i upgrade clickhouse. i don't found it. Please close the quetion. | 2019-10-24T13:04:05 | 2019-10-24T13:04:05 | {} | NONE |
ClickHouse | ClickHouse | 545,951,160 | 7,476 | sundy-li | LGTM | 2019-10-24T14:41:51 | 2019-10-24T14:41:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,341,060 | 7,484 | alesapin | ??? | 2019-10-25T12:51:43 | 2019-10-25T12:51:43 | {} | MEMBER |
ClickHouse | ClickHouse | 546,389,581 | 3,776 | vdimir | I have not enough time to work on this issue. But now I returned to this and main problem that I cannot figure out – how to pass one more argument to `PositionImpl`. It already contains four functions for constant/vector needle and constant/vector haystack. If I decide to add parameter it shouldn't be concrete constant... | 2019-10-25T15:00:18 | 2019-10-25T15:00:18 | {} | MEMBER |
ClickHouse | ClickHouse | 545,655,996 | 7,435 | filimonov | Related https://github.com/ClickHouse/ClickHouse/issues/6462 | 2019-10-23T22:05:51 | 2019-10-23T22:05:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,249,370 | 7,247 | BayoNet | Should we document it? | 2019-10-25T08:05:32 | 2019-10-25T08:05:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,722,106 | 7,486 | zhang2014 | https://github.com/ClickHouse/ClickHouse/pull/7498#issuecomment-546721810
Undo cn documents(Need `rebase`, because we need clear the commit record), or keep them synchronized. | 2019-10-27T18:43:59 | 2019-10-27T18:43:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,869,269 | 7,503 | Myshiner | Ok, thanks for your quick reply. | 2019-10-28T09:45:22 | 2019-10-28T09:45:22 | {} | NONE |
ClickHouse | ClickHouse | 546,917,332 | 4,178 | alesapin | Closed in favor of https://github.com/ClickHouse/ClickHouse/pull/7360 | 2019-10-28T12:07:52 | 2019-10-28T12:07:52 | {} | MEMBER |
ClickHouse | ClickHouse | 545,660,445 | 7,400 | filimonov | Please check
https://clickhouse.yandex/tutorial.html
especially section **ClickHouse deployment to cluster** | 2019-10-23T22:20:37 | 2019-10-23T22:20:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,675,801 | 7,412 | 4ertus2 | > Why it's named `default_merge_block_size` instead of `merge_max_block_size`?
Fixed | 2019-10-23T23:19:26 | 2019-10-23T23:19:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 545,729,675 | 7,449 | tencupofkaiwater | I know what you mean, will rebuild, thank you very much!! | 2019-10-24T03:42:04 | 2019-10-24T03:42:04 | {} | NONE |
ClickHouse | ClickHouse | 545,789,581 | 7,468 | filimonov | You are looking for `<merge_tree>` setting called `max_bytes_to_merge_at_max_space_in_pool` | 2019-10-24T07:41:31 | 2019-10-24T07:41:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,170,128 | 7,403 | zhang2014 | 看起来像是权限问题，请确保log、data目录有足够的权限启动。
最好使用英语来描述issues（可以方便其他遇到相似问题的人查找） | 2019-10-25T02:08:09 | 2019-10-25T02:08:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,293,903 | 7,480 | azat | >2019-10-25 09:48:42 /usr/bin/ld.gold: error: contrib/zlib-ng/libz.a(crc32.c.o): multiple definition of 'get_crc_table'
>2019-10-25 09:48:42 /usr/bin/ld.gold: contrib/mariadb-connector-c/libmariadb/libmariadbclient.a(crc32.c.o): previous definition here
Don't see how the patch can introduce this, maybe I need to re... | 2019-10-25T10:12:49 | 2019-10-25T10:12:49 | {} | MEMBER |
ClickHouse | ClickHouse | 546,351,256 | 7,489 | traceon | I see, that makes sense. | 2019-10-25T13:21:13 | 2019-10-25T13:21:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,471,065 | 7,491 | akuzm | Should we add an integration test for this? | 2019-10-25T18:53:47 | 2019-10-25T18:53:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,689,069 | 6,920 | stale[bot] | This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
| 2019-10-27T12:18:57 | 2019-10-27T12:18:57 | {} | NONE |
ClickHouse | ClickHouse | 546,867,026 | 7,503 | filimonov | Sending SIGKILL (9) with `kill` command will work but recommended way is
```
systemctl stop clickhouse-server
```
or
```
/etc/init.d/clickhouse-server stop
``` | 2019-10-28T09:39:33 | 2019-10-28T09:40:02 | {
"+1": 6,
"heart": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,941,961 | 7,465 | alexey-milovidov | We should remove this settings (and make is obsolete).
It was introduced only to overcome misconfiguration of some servers in Yandex Banner System. | 2019-10-28T13:19:27 | 2019-10-28T13:19:27 | {} | MEMBER |
ClickHouse | ClickHouse | 526,472,156 | 6,707 | filimonov | And what if (just for example) the sequence will fail at the very end, when table already exists, some data is already there but it's not full? Or you create final table in single (last) step? (but even it that case it's not 100% atomic operation, and it can fail after 90%) | 2019-08-30T06:13:27 | 2019-08-30T14:02:54 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.