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 | 552,428,616 | 6,312 | orofarne | @alexey-milovidov, single node configuration,
one `ALTER TABLE "..." DROP PARTITION ID '...'` query (via tcp)
and more then one `INSERT INTO "..." FORMAT JSONEachRow` queries in parallel (via http)
= deadlock
| 2019-11-11T12:37:19 | 2019-11-11T12:37:19 | {} | NONE |
ClickHouse | ClickHouse | 553,284,863 | 6,107 | slimtom95 | To complement this issue:
My ON CLUSTER DLL statement ends up timeout as configured in the default profile. **But changing and reloading max_execution_time in default profile won't help, the timeout value in exception message is same as before .** Restart CH servers may work, cannot test this now.
19.14.7.15 | 2019-11-13T08:00:46 | 2019-11-13T08:05:47 | {} | NONE |
ClickHouse | ClickHouse | 553,315,465 | 7,729 | dmitriy-myz | I faced with same issue. Resolved without changing cmdline after removing `CMakeCache.txt` file. Or entire `build_docker`` directory. | 2019-11-13T09:26:31 | 2019-11-13T09:26:31 | {
"+1": 14
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,359,670 | 7,586 | alesapin | conflicts in 19.15 | 2019-11-11T09:21:21 | 2019-11-11T09:21:21 | {} | MEMBER |
ClickHouse | ClickHouse | 552,428,487 | 7,702 | alexey-milovidov | Do you want to build different clickhouse variants from source?
Or you want to run multiple clickhouse binaries on a single server?
What kind of tarball do you need? | 2019-11-11T12:36:56 | 2019-11-11T12:36:56 | {} | MEMBER |
ClickHouse | ClickHouse | 552,453,603 | 7,713 | Tasselmi | > Could you please also provide some data from `test` table?
maybe the array is too long ???
when I test with little data, it is ok. But with big amount of data ,it fails.
Too much data, I can not provide so much. and with little amount of data, you can't reproduce it.
```
SELECT max(length(sessions.2))
FRO... | 2019-11-11T13:51:34 | 2019-11-11T13:52:00 | {} | NONE |
ClickHouse | ClickHouse | 552,470,909 | 7,572 | zhang2014 | @ValBaturin can you help me review this pr?
cc: @alexey-milovidov @millb | 2019-11-11T14:38:03 | 2019-11-11T14:38:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,513,902 | 7,721 | 4ertus2 | #7487 | 2019-11-11T16:26:42 | 2019-11-11T16:26:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,693,158 | 7,617 | CurtizJ | This behavior is expected, optimization for Merge tables wasn't planned. However, it's possible to implement optimization for source tables of Merge table, that have MergeTree-family engine, but it requires some refactoring. Now optimization is performed at query analyzation stage, before Merge table asks it's source t... | 2019-11-12T01:32:18 | 2019-11-12T01:32:18 | {} | MEMBER |
ClickHouse | ClickHouse | 552,716,693 | 7,714 | Tasselmi | the same syntax, yesterday it worked well, today it is broken.
```
↗ Progress: 55.93 thousand rows, 88.39 MB (462.26 thousand rows/s., 730.55 MB/s.) 4%Received exception from server (version 19.16.3):
Code: 9. DB::Exception: Received from localhost:9000. DB::Exception: Size of filter doesn't match size of column.... | 2019-11-12T03:21:02 | 2019-11-12T03:21:02 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 552,995,262 | 7,730 | filimonov | Can you just send a PR? | 2019-11-12T17:19:12 | 2019-11-12T17:19:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,895,578 | 7,678 | millb | > What happens if you change the semantic of `parse` function to expect eof instead of introducing a new function?
I did this but I got a lot of errors. As I think it happened because some functions use the property of ignoring the suffix. I don't know how many such functions are exist and decided that it's easier t... | 2019-11-12T13:35:30 | 2019-11-13T09:40:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,360,734 | 7,542 | alesapin | conflicts in 19.15 | 2019-11-11T09:23:58 | 2019-11-11T09:23:58 | {} | MEMBER |
ClickHouse | ClickHouse | 552,435,234 | 7,708 | alesapin | Trace looks like this:
```
2019.11.11 15:56:14.150567 [ 116 ] {7c5a3220-80d7-4495-a171-6d53331fb08e} <Error> executeQuery: Code: 47, e.displayText() = DB::Exception: Missing columns: 'arr' while processing query: 'SELECT arraySplit(x -> (x < 0), arr) AS sp', required columns: 'arr', source columns: 'dummy' (version 1... | 2019-11-11T12:57:08 | 2019-11-11T12:57:08 | {
"-1": 1
} | MEMBER |
ClickHouse | ClickHouse | 552,435,537 | 4,607 | alexey-milovidov | @malsumis
> I've looked through MergeTreeSettings.h and found a possible problem - maybe the parts are too big or some other. There is a setting, that might be blocking the merge process:
That's right.
You can check the exact reason if you `SET optimize_throw_if_noop = 1`.
ClickHouse will throw an exception ... | 2019-11-11T12:58:07 | 2019-11-11T12:58:20 | {} | MEMBER |
ClickHouse | ClickHouse | 552,772,932 | 7,714 | Tasselmi | ```
Stack trace:
0. 0x559cc1cd9f90 StackTrace::StackTrace() ?
1. 0x559cc1cd9d65 DB::Exception::Exception(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, int) ?
2. 0x559cc5231e02 void DB::filterArraysImpl<unsigned char>(DB::PODArray<unsigned char, 4096ul, Allocator<false, fa... | 2019-11-12T07:40:11 | 2019-11-12T07:40:11 | {} | NONE |
ClickHouse | ClickHouse | 553,162,570 | 7,238 | denistarasov | Is it right that a static field `static bool hide_addresses_in_stack_traces` (not sure if the naming is suitable) should be added to `StackTrace.h`, then these lines: https://github.com/ClickHouse/ClickHouse/blob/7bc561dd55b9188fb2cb4948cd4b11ab763a4f29/dbms/src/Common/StackTrace.cpp#L28-L31
should be changed to so... | 2019-11-12T23:18:21 | 2019-11-12T23:20:24 | {} | NONE |
ClickHouse | ClickHouse | 553,320,007 | 7,748 | nvartolomei | Can you try latest version first? | 2019-11-13T09:37:27 | 2019-11-13T09:37:27 | {
"heart": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,274,816 | 2,310 | alexey-milovidov | IN and LIKE are implemented.
And we don't implement LIMIT becase it is applicable only in trivial cases. | 2019-11-11T03:04:48 | 2019-11-11T03:04:48 | {} | MEMBER |
ClickHouse | ClickHouse | 552,346,553 | 5,707 | alexey-milovidov | @onine007 Sorry, why is it closed? | 2019-11-11T08:42:35 | 2019-11-11T08:42:35 | {} | MEMBER |
ClickHouse | ClickHouse | 552,347,974 | 7,706 | alexey-milovidov | @achimbab Does it work correctly for strings with characters out of Basic Multilingual Plane?
https://github.com/ClickHouse/ClickHouse/pull/7651#issuecomment-551179746 | 2019-11-11T08:46:43 | 2019-11-11T08:46:43 | {} | MEMBER |
ClickHouse | ClickHouse | 552,348,903 | 7,678 | alexey-milovidov | What happens if you change the semantic of `parse` function to expect eof instead of introducing a new function? | 2019-11-11T08:49:40 | 2019-11-11T08:49:54 | {} | MEMBER |
ClickHouse | ClickHouse | 552,350,182 | 7,707 | alexey-milovidov | @4ertus2 Looks Ok (simple change) but it's hacky. Now some functions can accept tables, but it was not possible before. Isn't it too disrupting? | 2019-11-11T08:53:53 | 2019-11-11T08:53:53 | {} | MEMBER |
ClickHouse | ClickHouse | 552,437,177 | 7,706 | ghost | @alexey-milovidov
I think that the purpose of `javaHashUTF16LE()` is to obey the `Java's hashCode()` function.
`Java's hashCode()` handles all the encodings as `StringUTF16` except `StringLatin1`.
```Java
public int hashCode() {
int h = hash;
if (h == 0 && value.length > 0) {
hash = h = isLati... | 2019-11-11T13:03:32 | 2019-11-11T13:08:55 | {} | NONE |
ClickHouse | ClickHouse | 552,560,478 | 7,716 | alexey-milovidov | (The reason is that it's only available from 19.17) | 2019-11-11T18:33:44 | 2019-11-11T18:33:44 | {} | MEMBER |
ClickHouse | ClickHouse | 552,785,394 | 6,621 | Tasselmi | how did you solve this problem?
I run across the same isuee.
I am so desperate.
I checked everything, and I am sure I did nothing wrong. fuck | 2019-11-12T08:22:25 | 2019-11-12T08:22:25 | {} | NONE |
ClickHouse | ClickHouse | 552,283,812 | 7,699 | yuzhichang | @amosbird I manged to fix this issue by rewriting sql as following:
```
ALTER TABLE default.cdp_orders UPDATE mid=joinGet('default.cdp_idmapping_archive_delta_kv', 'to_mid', mid), mid_seq=joinGet('default.cdp_idmapping_archive_delta_kv', 'to_mid_seq', mid) WHERE mid IN (SELECT from_mid FROM cdp_idmapping_archive_de... | 2019-11-11T03:57:05 | 2019-11-11T03:57:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,376,390 | 4,263 | azat | Yep, but there is no `PushingToViewsBlockOutputStream` in the stacktrace, either because of the optimizations or lack of debug symbols.
@greenwolf-nsk can you get the stacktrace with the debug symbols installed? (also note that you need to install them before starting clickhouse-server, since stacktrace resolving is... | 2019-11-11T10:05:34 | 2019-11-11T10:05:34 | {} | MEMBER |
ClickHouse | ClickHouse | 552,821,941 | 7,218 | KochetovNicolai | Yes, when I remove point (108.199173,-6.234169) I have correct result:
```
SELECT pointInPolygon((106.784534, -6.301819), [(106.396667, -6.97889), (106.391403, -6.90002), (106.418701, -6.881078), (106.438408, -6.81564), (106.487457, -6.794578), (106.511497, -6.765959), (106.429398, -6.69578), (106.4319, -6.58653), (1... | 2019-11-12T10:00:31 | 2019-11-12T10:00:31 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 553,081,274 | 7,702 | den-crane | It's much much easier to unpack .deb or .rpm and take clickhouse binary.
If you really want to build it here is the doc https://clickhouse.yandex/docs/en/development/build/ | 2019-11-12T19:40:52 | 2019-11-12T19:41:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 553,095,449 | 7,733 | mdonkers | Did not think about mentioning this. But yes, I had found that query on another issue thread and had executed it. Also checking now, we don't have parts bigger than the threshold that cannot get merged.
Perhaps not sufficiently emphasized, but as mentioned above, we found even when the mutation was 'stuck', the new ... | 2019-11-12T20:15:58 | 2019-11-12T20:15:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,314,685 | 7,241 | filimonov | #7377 | 2019-11-11T06:43:19 | 2019-11-11T06:43:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,429,439 | 7,709 | alesapin | It doesn't affect performance => can be merged. | 2019-11-11T12:39:50 | 2019-11-11T12:39:50 | {} | MEMBER |
ClickHouse | ClickHouse | 552,609,119 | 7,724 | 4ertus2 | #6296 | 2019-11-11T20:48:35 | 2019-11-11T20:48:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,722,431 | 7,714 | Tasselmi | this works fine.
```
SELECT *
FROM test
LIMIT 5
```
but........
```
SELECT count()
FROM test
→ Progress: 55.93 thousand rows, 84.97 MB (479.51 thousand rows/s., 728.52 MB/s.) 4%Received exception from server (version 19.16.3):
Code: 9. DB::Exception: Received from localhost:9000. DB::Exception: Size of ... | 2019-11-12T03:51:03 | 2019-11-12T03:51:39 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 552,882,751 | 5,767 | nikitamikhaylov | If you run "SET use_index_for_in_with_subqueries = 0", then the error won't occur. But it is a bug, and we will fix it. | 2019-11-12T12:59:18 | 2019-11-12T12:59:18 | {} | MEMBER |
ClickHouse | ClickHouse | 552,922,270 | 7,736 | silviucpp | I can confirm without [that patch](https://github.com/ClickHouse-Extras/poco/compare/ea2516be366a73a02a82b499ed4a7db1d40037e0...7a2d304c21549427460428c9039009ef4bbfd899#diff-4115d3841de351d4ca12d8813cfb9026) works fine or if we patch poco to not poll in case MSG_DONTWAIT is set :
```
if (_recvTimeout.totalMicroseco... | 2019-11-12T14:41:44 | 2019-11-12T17:21:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 553,083,168 | 7,484 | filimonov | Hm. Indeed. It looks like i messed something. Sorry, guys :)
```
SELECT substring(description, - 3) AS end
FROM system.merge_tree_settings
┌─end─┐
│ ue. │
│ e). │
│ e). │
│ ue. │
│ ue. │
│ es. │
│ ts" │
│ ts. │
│ es. │
│ le. │
│ on. │
│ on. │
│ on. │
│ d). │
│ me. │
│ ca. │
│ ng. │
│ es. │
... | 2019-11-12T19:45:24 | 2019-11-12T19:45:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,274,577 | 2,284 | alexey-milovidov | Improved in https://github.com/ClickHouse/ClickHouse/pull/2758 | 2019-11-11T03:03:22 | 2019-11-11T03:03:22 | {} | MEMBER |
ClickHouse | ClickHouse | 552,505,703 | 7,617 | cw9 | Hi, is there any update on this issue? | 2019-11-11T16:06:19 | 2019-11-11T16:06:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,317,503 | 7,699 | yuzhichang | With #7707，following statement works:
```
ALTER TABLE cdp_orders UPDATE mid=joinGet(cdp_idmapping_archive_delta_kv, 'to_mid', mid), mid_seq=joinGet(cdp_idmapping_archive_delta_kv, 'to_mid_seq', mid) WHERE mid IN (SELECT from_mid FROM cdp_idmapping_archive_delta_kv);
``` | 2019-11-11T06:55:50 | 2019-11-11T06:55:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,358,074 | 7,664 | alesapin | Conflicts in 19.15 :( | 2019-11-11T09:16:47 | 2019-11-11T09:16:47 | {} | MEMBER |
ClickHouse | ClickHouse | 552,926,747 | 7,733 | den-crane | Check that all parts are not frozen (not RO) (https://github.com/ClickHouse/ClickHouse/issues/7152)
They should have write flags `drwxr-x---` (/var/lib/clickhouse/data/{database}/{table}
Sometimes CH leaves some parts with RO `dr--r-x---` without obvious reason (accidental bug).
Change flags manually using chmo... | 2019-11-12T14:52:06 | 2019-11-12T14:55:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,349,742 | 7,708 | Tasselmi | > Provide trace from clickhouse-server.log file, please.
```
WITH arrayFilter(x -> (x > 3), [0, 1, 2]) AS arr
SELECT arraySplit(x -> (x < 0), arr) AS sp
↓ Progress: 0.00 rows, 0.00 B (0.00 rows/s., 0.00 B/s.)
Exception on client:
Code: 271. DB::Exception: Data compressed with different methods, given method b... | 2019-11-11T08:52:26 | 2019-11-11T08:53:08 | {} | NONE |
ClickHouse | ClickHouse | 552,402,091 | 7,432 | rochoa | Understood. Thanks for the clarification. | 2019-11-11T11:17:07 | 2019-11-11T11:17:07 | {} | NONE |
ClickHouse | ClickHouse | 552,563,266 | 7,713 | alexey-milovidov | > maybe the array is too long ???
Nevertheless, this should not be the reason for that error message.
> Too much data, I can not provide so much. and with little amount of data, you can't reproduce it.
Maybe you can create a table and generate artificial arrays (with `range` function) of enough length. | 2019-11-11T18:41:40 | 2019-11-11T18:41:40 | {} | MEMBER |
ClickHouse | ClickHouse | 552,834,797 | 6,168 | nikitamikhaylov | Why do you modify your test in last commit? If you are worried about red performance test - that is our problem and we will fix it ASAP. | 2019-11-12T10:33:40 | 2019-11-12T10:33:40 | {} | MEMBER |
ClickHouse | ClickHouse | 552,877,176 | 6,312 | orofarne | @alexey-milovidov, ticket #157356245795793 | 2019-11-12T12:42:05 | 2019-11-12T12:42:05 | {} | NONE |
ClickHouse | ClickHouse | 552,884,549 | 7,572 | ValBaturin | @zhang2014 sure | 2019-11-12T13:04:41 | 2019-11-12T13:04:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,919,435 | 7,697 | filimonov | And message 'coordinates' in Kafka, i.e. topic / partition / offset | 2019-11-12T14:35:16 | 2019-11-12T14:35:41 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 553,007,155 | 7,708 | filimonov | Bit simpler example
```
echo 'SELECT arraySplit(x -> (x < 0), emptyArrayUInt8() ) AS sp' | curl 'http://localhost:8123/' --data-binary @-
``` | 2019-11-12T17:43:19 | 2019-11-12T17:43:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,346,832 | 7,494 | alesapin | Fix https://github.com/ClickHouse/ClickHouse/pull/7709. It's not critical. | 2019-11-11T08:43:27 | 2019-11-11T08:43:27 | {} | MEMBER |
ClickHouse | ClickHouse | 552,566,844 | 7,706 | alexey-milovidov | > In my environment the results are the same with Java`s hashCode.
Perfect! | 2019-11-11T18:51:45 | 2019-11-11T18:51:45 | {} | MEMBER |
ClickHouse | ClickHouse | 552,607,240 | 7,238 | alexey-milovidov | Now this task is unassigned. | 2019-11-11T20:43:24 | 2019-11-11T20:43:24 | {} | MEMBER |
ClickHouse | ClickHouse | 552,801,449 | 7,729 | tlightsky | to resolve:
```
cmake -G Ninja /server -DCMAKE_C_COMPILER=`which gcc-9` -DCMAKE_CXX_COMPILER=`which g++-9` \
-DCMAKE_BUILD_WITH_INSTALL_RPATH=ON
``` | 2019-11-12T09:08:21 | 2019-11-12T09:08:21 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,353,013 | 7,707 | amosbird | btw, It also makes sense for dictGet functions to accept tables (as we now have dictionary tables) | 2019-11-11T09:01:57 | 2019-11-11T09:01:57 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 552,614,923 | 7,726 | alexey-milovidov | Notes:
- it's almost impossible to provide a test;
- the error message is lost and it's considered Ok;
- TODO: also need to update libcpuid just in case. | 2019-11-11T21:04:03 | 2019-11-11T21:04:31 | {} | MEMBER |
ClickHouse | ClickHouse | 552,758,171 | 6,312 | orofarne | typical situation:
https://gist.github.com/orofarne/bf0c64ccdc76d715e439ff74022663d9 | 2019-11-12T06:44:19 | 2019-11-12T06:44:19 | {} | NONE |
ClickHouse | ClickHouse | 552,842,813 | 7,731 | jsding | I checked the kafka log, clickhouse kafka consumer group only worked for 5 seconds and left, Is there option to config this? Thanks
```
[2019-11-12 10:58:54,060] INFO [GroupCoordinator 1002]: Preparing to rebalance group ck_event_prod_v3 in state PreparingRebalance with old generation 2728 (__consumer_offsets-9) (r... | 2019-11-12T10:54:57 | 2019-11-12T11:06:11 | {} | NONE |
ClickHouse | ClickHouse | 552,849,525 | 5,318 | BayoNet | As a base functionality I propose to use the following idea:
```
std::string get_askii_string(int string_len)
{
//Generates numbers in the [32, 127] range uniformly and adds them as symbols to the result string.
std::string resulting_string("");
std::default_random_engine generator(time(0));
... | 2019-11-12T11:13:44 | 2019-11-12T11:14:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,275,408 | 2,312 | alexey-milovidov | 133363e8e36b83e6ee7450a44d90b05b43dd8f62 | 2019-11-11T03:08:07 | 2019-11-11T03:08:07 | {} | MEMBER |
ClickHouse | ClickHouse | 552,359,942 | 4,263 | alexey-milovidov | This looks relevant to https://github.com/ClickHouse/ClickHouse/pull/7601
@azat FYI | 2019-11-11T09:22:01 | 2019-11-11T09:22:01 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 552,406,794 | 5,162 | alexey-milovidov | Ok, it's enough. | 2019-11-11T11:31:12 | 2019-11-11T11:31:12 | {} | MEMBER |
ClickHouse | ClickHouse | 552,575,935 | 7,218 | KochetovNicolai | I've found out that your polygon has self-intersection.
My [geojson](https://gist.github.com/KochetovNicolai/f5d22a03064d009e607a6f81a015e9de). | 2019-11-11T19:16:22 | 2019-11-11T19:16:22 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 552,951,448 | 7,733 | mdonkers | Thanks,
I've checked all nodes, and while I do see differences with the column-files within the parts folders (some have `-r--r-----` but others have `-rw-r-----`), all parts folders seem to have the correct rights: all `drwxr-x---`.
Not sure if by killing and continuing this eventually resolved itself? But at le... | 2019-11-12T15:46:18 | 2019-11-12T15:46:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 553,008,930 | 7,715 | filimonov | Closing in advance to https://github.com/ClickHouse/ClickHouse/issues/7697 | 2019-11-12T17:47:19 | 2019-11-12T17:47:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,334,156 | 2,099 | VladRassokhin | Great! Thanks for fixing! | 2019-11-11T08:01:09 | 2019-11-11T08:01:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,338,276 | 7,700 | caueteixeira | Thanks guys! | 2019-11-11T08:15:44 | 2019-11-11T08:15:44 | {} | NONE |
ClickHouse | ClickHouse | 552,361,216 | 5,707 | CurtizJ | It should be closed, because it's fixed in #6225. | 2019-11-11T09:25:11 | 2019-11-11T09:25:11 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 552,426,799 | 6,312 | alexey-milovidov | The bug in 18.14 is related to https://github.com/ClickHouse/ClickHouse/pull/6764 (fixed)
But I don't know what is the bug in 19.14.
@orofarne Is it specific to ON CLUSTER queries? | 2019-11-11T12:31:55 | 2019-11-11T12:31:55 | {} | MEMBER |
ClickHouse | ClickHouse | 552,561,653 | 7,717 | alexey-milovidov | That's very interesting feature!
@yurriy will help with review. | 2019-11-11T18:37:06 | 2019-11-11T18:37:06 | {} | MEMBER |
ClickHouse | ClickHouse | 552,585,576 | 7,218 | alexey-milovidov | @KochetovNicolai Does it work as expected if you remove self-intersection?
PS. Can we throw an exception if self-intersection is found? | 2019-11-11T19:41:59 | 2019-11-11T19:41:59 | {} | MEMBER |
ClickHouse | ClickHouse | 552,921,241 | 7,731 | filimonov | Kafka consumer get stopped in case on any error, leading to rebalance.
After 5 sec it started again and tries one more time...
And yes, it is bad. | 2019-11-12T14:39:21 | 2019-11-12T14:39:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,925,431 | 7,728 | filimonov | Please provide full stacktrace from the copier log, also copier config (if possible) | 2019-11-12T14:49:07 | 2019-11-12T14:49:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,347,542 | 7,708 | alesapin | Provide trace from clickhouse-server.log file, please. | 2019-11-11T08:45:27 | 2019-11-11T08:45:27 | {} | MEMBER |
ClickHouse | ClickHouse | 552,363,249 | 7,350 | alesapin | conflict in 19.15 | 2019-11-11T09:30:34 | 2019-11-11T09:30:34 | {} | MEMBER |
ClickHouse | ClickHouse | 552,735,644 | 7,702 | den-crane | > I want to run multiple clickhouse binaries on a single server.
Then you don't need tar ball. Just install CH version you need using deb or rpm and run using different configs.
clickhouse-server --daemon --pid-file=/var/run/clickhouse-server/clickhouse-server1.pid --config-file=/etc/clickhouse-server/config1.xml... | 2019-11-12T05:04:00 | 2019-11-12T05:04:00 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,738,779 | 7,702 | chu1070y | Sorry, I misunderstood the question.
Could you tell me the way how to build different clickhouse variants?
| 2019-11-12T05:20:57 | 2019-11-12T05:20:57 | {} | NONE |
ClickHouse | ClickHouse | 553,003,506 | 7,708 | filimonov | @alesapin there is no exception on the server side. But it do some wierd stuff.
```
echo 'WITH arrayFilter(x -> (x > 3), [0, 1, 2]) AS arr SELECT arraySplit(x -> (x < 0), arr) AS sp' | curl 'http://localhost:8123/' --data-binary @-
[[208]]
echo 'WITH arrayFilter(x -> (x > 3), [0, 1, 2]) AS arr SELECT arraySplit(x -... | 2019-11-12T17:35:37 | 2019-11-12T17:38:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,411,215 | 3,268 | alexey-milovidov | Our internal users also have faced this issue.
It should be trivial to fix: just skip the check for Distributed tables. | 2019-11-11T11:44:56 | 2019-11-11T11:44:56 | {} | MEMBER |
ClickHouse | ClickHouse | 552,422,327 | 6,312 | orofarne | 19.14 (Yandex.Cloud Managed Service)
alter table drop partition + insert → same problem | 2019-11-11T12:18:41 | 2019-11-11T12:18:41 | {} | NONE |
ClickHouse | ClickHouse | 552,431,754 | 5,185 | alexey-milovidov | The rules of sharding and sampling are different.
For sharding, the value of `cityHash64(sequenceId) % 2 = N` will be used.
For sampling, the value of `cityHash64(sequenceId) <= 0xFFFFFFFFFFFFFFFF / K` will be used.
The rule for sampling is made that way to ensure continuous ranges of sampled data if the data is... | 2019-11-11T12:46:26 | 2019-11-11T12:46:26 | {} | MEMBER |
ClickHouse | ClickHouse | 552,441,041 | 7,708 | Tasselmi | > Trace looks like this:
>
> ```
> 2019.11.11 15:56:14.150567 [ 116 ] {7c5a3220-80d7-4495-a171-6d53331fb08e} <Error> executeQuery: Code: 47, e.displayText() = DB::Exception: Missing columns: 'arr' while processing query: 'SELECT arraySplit(x -> (x < 0), arr) AS sp', required columns: 'arr', source columns: 'dummy' ... | 2019-11-11T13:15:59 | 2019-11-11T13:15:59 | {} | NONE |
ClickHouse | ClickHouse | 552,570,030 | 6,312 | alexey-milovidov | ALTER will wait for all INSERT queries to finish. But should not continue to hang after INSERT queries will finish.
What is the state of SHOW PROCESSLIST when the queries hang?
Could you please also ask Yandex.Cloud support with all the details? They will give these details to us and we will investigate further.... | 2019-11-11T19:00:18 | 2019-11-11T19:00:18 | {} | MEMBER |
ClickHouse | ClickHouse | 552,839,014 | 6,168 | YiuRULE | @nikitamikhaylov Concerning the last commit, it was because the query would do an infinite loop as I didn't put a limit during the insertion, it's mot related to the performance test | 2019-11-12T10:44:23 | 2019-11-12T10:44:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,886,876 | 1,784 | NanoBjorn | @xmariachi are you sure that primary / order by keys are set correctly and you checking corresponding uniqueness? | 2019-11-12T13:11:21 | 2019-11-12T13:11:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 553,011,361 | 7,484 | denistarasov | ```
────────────────┐
│
│
│
│
│
ning merges. │
│
│
│
│
│
│
│
│
licated_deduplication_window t... | 2019-11-12T17:52:10 | 2019-11-12T17:52:10 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 553,287,420 | 7,565 | ghost | I added LRUCache to QueryCache.
Even though anyone doesn't reply for this draft I have fixed some codes and reviewed myself.
But, I worry about deadlock or memory leak or any other problems.
I'm going to some more test and implement other things in TODO list.
I would be very happy if someone would note any comments... | 2019-11-13T08:09:25 | 2019-11-13T08:10:47 | {} | NONE |
ClickHouse | ClickHouse | 553,319,878 | 7,729 | tlightsky | @dmitriy-myz thanks, i'll try this way.
maybe also need to upgrade cmake too | 2019-11-13T09:37:08 | 2019-11-13T09:37:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,276,912 | 2,507 | alexey-milovidov | Fixed in master. | 2019-11-11T03:16:37 | 2019-11-11T03:16:37 | {} | MEMBER |
ClickHouse | ClickHouse | 552,374,235 | 7,711 | alexey-milovidov | You have specified four characters: \\, x, 7, F, as a delimiter, because escape sequence is not interpreted by your shell interpreter.
Write
```
clickhouse-client --host dbt20 --port 9000 --format_csv_delimiter=$'\x7F' --query="INSERT INTO dwzc.twb_m_top_organization FORMAT CSV" < /data/test_data/test.txt
```
in... | 2019-11-11T09:59:50 | 2019-11-11T10:01:44 | {} | MEMBER |
ClickHouse | ClickHouse | 552,377,624 | 7,713 | alexey-milovidov | Could you please also provide some data from `test` table? | 2019-11-11T10:08:58 | 2019-11-11T10:08:58 | {} | MEMBER |
ClickHouse | ClickHouse | 552,384,267 | 5,162 | akuzm | hits_100m is already done, link in the docs https://clickhouse.yandex/docs/en/getting_started/example_datasets/metrica/
Only hits_1000m is left. | 2019-11-11T10:26:06 | 2019-11-11T10:26:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,390,957 | 7,711 | xuxudede | > You have specified four characters: \, x, 7, F, as a delimiter, because escape sequence is not interpreted by your shell interpreter.
>
> Write
>
> ```
> clickhouse-client --host dbt20 --port 9000 --format_csv_delimiter=$'\x7F' --query="INSERT INTO dwzc.twb_m_top_organization FORMAT CSV" < /data/test_data/test... | 2019-11-11T10:45:04 | 2019-11-11T10:45:04 | {} | NONE |
ClickHouse | ClickHouse | 552,703,077 | 7,702 | chu1070y | I want to run multiple clickhouse binaries on a single server. | 2019-11-12T02:16:29 | 2019-11-12T02:16:29 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 552,913,883 | 7,736 | silviucpp | Seems at some point ClickHouse did a patch over poco and enabled `POCO_BROKEN_TIMEOUTS` on all platforms.
Looking to `receiveBytes` implementation if `_isBrokenTimeout` is true `MSG_DONTWAIT` is not
taken into account when the `poll` is executed and will wait for `_recvTimeout` ms instead returning immediately.
... | 2019-11-12T14:21:48 | 2019-11-12T14:21:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 552,963,271 | 7,733 | den-crane | Check result of this query
select * from system.parts where bytes_on_disk > (select toInt64(value) from system.merge_tree_settings where name = 'max_bytes_to_merge_at_max_space_in_pool')
It could be another reason why mutations stuck -- size of part is bigger than max_bytes_to_merge_at_max_space_in_pool | 2019-11-12T16:11:28 | 2019-11-12T16:11:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 553,297,186 | 7,730 | tlightsky | yes, i need to wait for my first build done(in docker),
it have been about 24h | 2019-11-13T08:38:43 | 2019-11-13T08:38:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 553,327,942 | 7,660 | filimonov | Hi @ishirav
We were not able to reproduce the problem when parts were not removed from the original location in synthetic tests. :\ So if it happens in your case - maybe it is caused by some specific queries, or by some kind of race conditions.
Can you help us to create some reproducible example?
What can hel... | 2019-11-13T09:57:09 | 2019-11-13T09:57:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 546,947,747 | 7,491 | alexey-milovidov | The purpose of integration tests is to test as many integrations as possible. | 2019-10-28T13:33:49 | 2019-10-28T13:33:49 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.