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 | 504,312,483 | 5,705 | blinkov | cc @zhang2014 @sundy-li | 2019-06-21T06:48:01 | 2019-06-21T06:48:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,487,720 | 5,013 | alexey-milovidov | We've published an article describing all these optimizations in great detail:
https://news.ycombinator.com/item?id=20274733
Hope you'll enjoy! | 2019-06-25T15:07:49 | 2019-06-25T15:22:50 | {
"+1": 3
} | MEMBER |
ClickHouse | ClickHouse | 505,907,058 | 5,676 | champtar | #5754 | 2019-06-26T14:43:47 | 2019-06-26T14:43:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 506,082,268 | 5,761 | den-crane | Compression rate highly depends of ORDER BY
```
ORDER BY (cid, keyword, ts)
VS
ORDER BY (toDate(ts), keyword)
```
Probably [ORDER BY (cid, keyword, ts)] places similar values in all columns very close that's why compression is better. It's known feature of all columnar databases.
| 2019-06-26T23:35:25 | 2019-06-26T23:35:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,461,898 | 4,538 | stavrolia | Please, fix other whitespaces in code too. You could use clang format (7th version is better) for it, our styleguide differs only in lambas. | 2019-06-21T15:12:01 | 2019-06-21T15:12:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,778,149 | 5,723 | alex-zaitsev | This is Mac specific issue. Please add at the beginning of `IPv6ToBinary.cpp`:
```#include <cstring>```
It fixed the build issue for me. | 2019-06-23T19:05:44 | 2019-06-23T19:05:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,208,000 | 4,637 | den-crane | but it gives incorrect results for tables without sampling ?
https://github.com/yandex/ClickHouse/issues/5733 | 2019-06-24T22:40:18 | 2019-06-24T22:41:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,497,113 | 5,743 | alexey-milovidov | I've changed the title of this bug accordingly:
```
SELECT
toStartOfInterval (now (), toIntervalHour (12), 'UTC') - (3600 * 3) AS t1, toTypeName(t1),
toStartOfInterval (now (), toIntervalHour (12), 'UTC') AS t2, toTypeName(t2)
``` | 2019-06-25T15:29:39 | 2019-06-25T15:29:39 | {} | MEMBER |
ClickHouse | ClickHouse | 504,333,243 | 5,030 | YiuRULE | @proller I create a simple python script who set the header of the gperf file
```
%language=C++
%define lookup-function-name is_valid
%define class-name tldLookupHash
%readonly-tables
%includes
%compare-strncmp
%{
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wimplicit-fallthrough"
#pragma G... | 2019-06-21T08:04:35 | 2019-06-21T08:04:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,340,918 | 5,582 | Felixoid | Well, `age=0` is a totally correct value according to the documentation. And that is defined everywhere https://clickhouse.yandex/docs/en/operations/table_engines/graphitemergetree/#configuration-example | 2019-06-21T08:29:47 | 2019-06-21T08:29:47 | {} | MEMBER |
ClickHouse | ClickHouse | 504,634,351 | 5,596 | excitoon | For the next steps I shall move tests to the proper folder and after that implement authorization things. | 2019-06-22T06:02:49 | 2019-06-22T06:02:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,750,029 | 5,692 | filimonov | #5717 | 2019-06-23T13:04:56 | 2019-06-23T13:04:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,196,925 | 4,786 | nvartolomei | #4918 | 2019-06-24T21:56:10 | 2019-06-24T21:56:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,394,255 | 4,199 | xmariachi | Hi @igor-sh8 , could you please give more detail on how you solved it using tables? Also - did you have to do anything special / did you find any limitation by using `Replicated*` tables? | 2019-06-25T10:58:22 | 2019-06-25T10:58:22 | {} | NONE |
ClickHouse | ClickHouse | 505,913,727 | 5,670 | alesapin | Will try to fix here https://github.com/yandex/ClickHouse/pull/5757. | 2019-06-26T14:58:56 | 2019-06-26T14:58:56 | {} | MEMBER |
ClickHouse | ClickHouse | 505,925,545 | 5,756 | alesapin | Thanks for the report. Can you install version with address sanitizer? It can be downloaded from this page https://clickhouse-builds.s3.yandex.net/5626/2f0c7d094aad8bbf8453e6fd90d5851222dfa79a/report.html (row where third column == 'asan'). Sanitizer reports much more convenient than segfault traces.
P.S. asan print... | 2019-06-26T15:27:33 | 2019-06-26T15:27:33 | {} | MEMBER |
ClickHouse | ClickHouse | 504,321,337 | 5,705 | zhang2014 | LGTM | 2019-06-21T07:21:42 | 2019-06-21T07:21:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,365,666 | 5,707 | onine007 |
select count()
FROM coach_news.dw_news_summary_fact_formal
WHERE (ymd = '2019-06-20') AND (hour > 16) AND (entrance IN (1, 5)) AND (client_id = 1) AND (rq_action IN (1, 2))
return 51072678
| 2019-06-21T09:48:23 | 2019-06-21T09:48:23 | {} | NONE |
ClickHouse | ClickHouse | 504,601,822 | 5,708 | den-crane | 19.10.1.669
$ clickhouse-client -q "create table dist_test ( num Int64, str String) engine =Log;"
$ clickhouse-client -q "insert into dist_test select toInt64(number/1000)number, 'TEST' from system.numbers limit 1500000"
$ clickhouse-client -q "select distinct num from dist_test where str in (select 'TEST') forma... | 2019-06-21T23:20:19 | 2019-06-21T23:22:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,648,459 | 5,596 | alexey-milovidov | > I just thought that it may be useful to continue uploading until we get an error from server.
Yes.
> For the next steps I shall move tests to the proper folder and after that implement authorization things.
Great! | 2019-06-22T09:22:45 | 2019-06-22T09:22:45 | {} | MEMBER |
ClickHouse | ClickHouse | 505,554,015 | 3,783 | simPod | Same here. Thanks! | 2019-06-25T17:56:33 | 2019-06-25T17:56:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,779,182 | 5,742 | alexey-milovidov | PS. What's in stress test? | 2019-06-26T08:38:32 | 2019-06-26T08:38:32 | {} | MEMBER |
ClickHouse | ClickHouse | 504,444,576 | 5,709 | nvartolomei | Related #1226. | 2019-06-21T14:23:45 | 2019-06-21T14:23:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,961,007 | 5,699 | spyros87 | @abyss7 A newer version has been released. See here: https://github.com/edenhill/librdkafka/releases/tag/v1.1.0 | 2019-06-24T10:56:14 | 2019-06-24T10:56:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,807,614 | 5,678 | andyyzh | nice! | 2019-06-26T09:58:39 | 2019-06-26T09:58:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,865,083 | 5,755 | 4ertus2 | #5689 | 2019-06-26T13:02:33 | 2019-06-26T13:02:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,463,989 | 5,742 | 4ertus2 | I've compressed some real data DateTime column with different variants (one month data, not in key)
```
lz4 833569732
zstd 515097904
t64(byte)+lz4 407128098
t64(bit)+lz4 472855596
t64(byte)+zstd 345330458
t64(bit)+zstd 318123008
delta+lz4 407962488
delta+zstd 2415321... | 2019-06-25T14:14:02 | 2019-06-26T19:13:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 506,024,855 | 5,759 | alexey-milovidov | Please add a test.
It will be sh test with 2>&1 and grep. | 2019-06-26T20:05:36 | 2019-06-26T20:07:31 | {} | MEMBER |
ClickHouse | ClickHouse | 506,079,180 | 3,925 | vzakaznikov | Hi @Vxider, the LIVE VIEW table implementation does use incremental computation as much as possible. The implementation uses the same approach as a distributed table. In the case of the query against a distributed table, the result is computed partially on each server and then merged on the
server where the query orig... | 2019-06-26T23:19:34 | 2019-06-26T23:19:34 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 506,083,655 | 5,761 | ramazanpolat | Thank you. I will try to create identical engine settings for both installation to check this. | 2019-06-26T23:42:45 | 2019-06-26T23:42:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,689,153 | 5,582 | blind-oracle | Да, используя скрипт минут через 30 начало мержить криво.
```
┌─Value─┬────toDateTime(Time)─┐
│ 1 │ 2019-06-22 17:05:00 │
│ 2 │ 2019-06-22 17:06:00 │
│ 2 │ 2019-06-22 17:07:00 │
...
│ 2 │ 2019-06-22 17:39:00 │
│ 2 │ 2019-06-22 17:40:00 │
│ 1 │ 2019-06-22 17:41:00 │
│ 1 │ 2019-06-22... | 2019-06-22T18:39:28 | 2019-06-22T18:39:28 | {} | NONE |
ClickHouse | ClickHouse | 504,776,305 | 5,582 | blind-oracle | На проде тоже ошибки вылезли, но очень мало.
```
│ 1192 │ 2019-06-23 19:57:00 │
│ 64 │ 2019-06-23 19:57:00 │
```
У меня такое ощущение что КХ делает два параллельных мержа на одной таблице и в результате получается два значения усредненных до начала минуты... Но по идее он не должен делать два мержа параллел... | 2019-06-23T18:37:14 | 2019-06-23T18:37:14 | {} | NONE |
ClickHouse | ClickHouse | 504,874,763 | 5,212 | andyyzh | @vitlibar
All the review issues have been fixed except toStartOfWeekInterval() function was not added.
Please review again. | 2019-06-24T06:23:38 | 2019-06-24T06:23:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,990,334 | 5,590 | alesapin | Also error messages are unclear:
```
SELECT arrayReduce('sumResample(11, 10, 1)', [10], [0])
Received exception from server (version 19.9.1):
Code: 36. DB::Exception: Received from localhost:9000, ::1. DB::Exception: The range given in function sumResample contains too many elements.
``` | 2019-06-24T12:34:16 | 2019-06-24T12:34:16 | {} | MEMBER |
ClickHouse | ClickHouse | 504,414,713 | 5,703 | BayoNet | En docs were fixed 7 days ago in the https://github.com/yandex/ClickHouse/commit/4f408cef1281b7eb86b6be3edc9322c1d4e18227#diff-5a70d14804c5e8dd00569e461bdca911 commit.
Ru docs I fixed right now in the #5706.
Examples from docs should work:
```
$ clickhouse-client
ClickHouse client version 19.10.1.683 (officia... | 2019-06-21T12:53:22 | 2019-06-21T12:53:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,971,573 | 5,590 | alesapin | `select arrayReduce('sumResample(1, 1, 0)', [10, 11, 12, 13, 14, 15, 16, 17, 18, 19], [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]);` leads to Floating point exception and server crash https://pastebin.com/3UzexYKW. | 2019-06-24T11:33:27 | 2019-06-24T11:33:27 | {} | MEMBER |
ClickHouse | ClickHouse | 458,146,674 | 4,157 | nvartolomei | @KochetovNicolai how would you reshard data if date is used as partition key and partitions are present on both servers? From the following comment my understanding is that it is not possible. Let me know if I did misinterpret it.
```
NOTE: In spite of this section is optional (if it is not specified, all partition... | 2019-01-28T14:13:34 | 2019-06-24T22:16:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,806,594 | 5,742 | 4ertus2 | > PS. What's in stress test?
Code: 202. DB::Exception: Received from localhost:9000, ::1. DB::Exception: Too many simultaneous queries. Maximum: 100. | 2019-06-26T09:55:52 | 2019-06-26T09:55:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,935,766 | 3,609 | crakjie | system.table do not contain partition_key, sorting_key, primary_key and sampling_key for MaterilizedView ( even when build with MergeTree Engine )
ClickHouse server version 19.4.3 revision 54416. | 2019-06-26T15:52:54 | 2019-06-26T15:52:54 | {} | NONE |
ClickHouse | ClickHouse | 506,036,678 | 5,758 | zlobober | > I'd like to rename every getMask... to check...
Sure, no problem. | 2019-06-26T20:41:39 | 2019-06-26T20:41:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 506,076,542 | 5,756 | 4ertus2 | I've seen same errors on other codec when getMaxCompressedDataSize() returned wrong value. It probably should returns
```
+ 2 * sizeof(T)
``` | 2019-06-26T23:05:43 | 2019-06-26T23:08:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,559,545 | 5,582 | blind-oracle | @den-crane
> No. Age=120 still has the discrepancies.
For now on my test server with a copy of production metrics (~30k/sec) I don't see it. Before the errors were all over the place. But I'll monitor for a couple of days.
> It may be rolls incorrectly in different merges when one minute placed in two different... | 2019-06-21T20:16:52 | 2019-06-21T20:16:52 | {} | NONE |
ClickHouse | ClickHouse | 504,776,526 | 5,724 | alexey-milovidov | Я не понял: что такое "грибеного-ата", какие источники. | 2019-06-23T18:40:44 | 2019-06-23T18:41:48 | {} | MEMBER |
ClickHouse | ClickHouse | 504,989,504 | 5,590 | alesapin | What is the main idea of the negative step and intervals? I don't understand how they work:
```
SELECT arrayReduce('sumResample(0, -1, -1)', [10], [0])
┌─arrayReduce('sumResample(0, -1, -1)', [10], [0])─┐
│ [] │
└──────────────────────────────────────────────────┘
S... | 2019-06-24T12:31:41 | 2019-06-24T12:31:41 | {} | MEMBER |
ClickHouse | ClickHouse | 505,522,584 | 5,735 | den-crane | >/dev/sda5 92G 28G 60G 32% /
free only 28GB. It could be eaten in a second by inactive parts.
use [ df -h /var/lib/clickhouse/ ] to check available space. | 2019-06-25T16:33:08 | 2019-06-25T16:34:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 506,017,663 | 5,756 | blind-oracle | @alesapin
Here's the ASAN dump with llvm-symbolizer applied:
```
#0 0x133724d1 in void unalignedStore<int>(void*, int const&) /build/obj-x86_64-linux-gnu/../libs/libcommon/include/common/unaligned.h:17:5
#1 0x133724d1 in void DB::(anonymous namespace)::compressDataForType<unsigned short>(char const*, uns... | 2019-06-26T19:42:55 | 2019-06-26T19:52:54 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 504,373,613 | 5,703 | 4ertus2 | I've looked in code. There's IPv4CIDRToRange function there (with upper cased T).
There's an error in docs. Russian version also contains wrong name IPv4CIDRtoIPv4Range. | 2019-06-21T10:15:12 | 2019-06-21T10:15:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,535,227 | 5,710 | filimonov | I'll finish the PR at the beginning July. I would be appreciate for CR in the meanwhile. | 2019-06-21T18:54:31 | 2019-06-21T18:56:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,542,196 | 5,582 | den-crane | No. Age=120 still has the discrepancies.
```
│ 2 │ 2019-06-21 01:14:00 │
│ 1 │ 2019-06-21 01:15:00 │
select avg(Value) a, toStartOfMinute(Time) t, Path from graphite_crap where Time >= '2019-06-20 22:55:00' and Time <= '2019-06-21 02:24:00' group by Path, t having a <> 2
3780000 rows in set.
select... | 2019-06-21T19:15:22 | 2019-06-21T19:18:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,387,209 | 5,720 | alesapin | Good investigation and fix. Thank you. | 2019-06-25T10:34:50 | 2019-06-25T10:34:50 | {} | MEMBER |
ClickHouse | ClickHouse | 505,690,721 | 1,851 | Aricliang | mark | 2019-06-26T02:21:35 | 2019-06-26T02:21:35 | {} | NONE |
ClickHouse | ClickHouse | 505,858,196 | 3,174 | den-crane | > Does that mean, that if the sizes are bigger than the maximum part size, they will never be merged?
Yes.
>If so, and If max value is increased once it happened, will they eventually be merged then?
Yes. But the merge of too large parts could be unreasonable because it will take a lot of CPU and IO resources ... | 2019-06-26T12:42:48 | 2019-06-26T12:42:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,342,102 | 5,582 | blind-oracle | @Felixoid Yep, that's why I consider this behavior a bug and it's kinda strange that nobody noticed it :-/
Either nobody uses `age=0` (which I doubt) or nobody checks the numbers... because this happens sooner or later under any load. | 2019-06-21T08:33:45 | 2019-06-21T08:35:02 | {} | NONE |
ClickHouse | ClickHouse | 504,366,174 | 5,707 | onine007 | 
| 2019-06-21T09:50:04 | 2019-06-21T09:50:04 | {} | NONE |
ClickHouse | ClickHouse | 504,967,552 | 5,719 | proller | It is possible to place new files to utils/package/arch/ ? | 2019-06-24T11:19:08 | 2019-06-24T11:19:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,147,364 | 5,731 | 4ertus2 | No profit on not sorted data. Test with all compressed columns outside of key:
```
4710836 ./trades
4168280 ./trades_t64
4169004 ./trades_t64_dx
4100384 ./trades_t64_zstd
4100556 ./trades_t64_dx_zstd
```
The same with ordered data. It just doesn't work. | 2019-06-24T19:29:45 | 2019-06-24T20:18:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,258,378 | 5,732 | zhang2014 | I see the ClickHouse team has a similar plan. Maybe we can look forward to that:
https://github.com/yandex/clickhouse-presentations/blob/master/percona2019/index.html#L567
https://gist.github.com/alexey-milovidov/d62d73222d83b9319dc519cbb13aeff6
cc @alexey-milovidov
BTW: max_execution_speed and max_execution_... | 2019-06-25T02:50:20 | 2019-06-25T02:55:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,707,579 | 3,925 | Vxider | Hi, I think Live View provide the abilities of streaming computing, is the live view use an incremental calculation? For example, performing a word count job on a table, does it recalculate the entire table each time data is inserted ? | 2019-06-26T03:55:33 | 2019-06-26T03:55:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,221,682 | 5,596 | excitoon | @alexey-milovidov There is a [funny limitation in multipart API](https://docs.aws.amazon.com/AmazonS3/latest/API/mpUploadUploadPart.html), number of parts can not be greater than 10,000.
> Part numbers can be any number from 1 to 10,000, inclusive. A part number uniquely identifies a part and also defines its positi... | 2019-06-20T22:26:17 | 2019-06-20T22:26:34 | {
"eyes": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,234,622 | 5,582 | den-crane | Seems age 120 do the trick. Maybe avg works over incomplete data.
```
<retention>
<age>120</age>
<precision>60</precision>
</retention>
``` | 2019-06-20T23:29:39 | 2019-06-20T23:29:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,645,018 | 5,316 | excitoon | This is a duplicate of #5169 | 2019-06-22T08:43:52 | 2019-06-22T08:43:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,666,675 | 5,708 | den-crane | it probably fixed with 19.5.4.22 19.6.2.11 https://github.com/yandex/ClickHouse/pull/5001 | 2019-06-22T13:39:53 | 2019-06-22T13:39:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,860,022 | 5,678 | zhangcl6066 | COOL | 2019-06-24T05:11:31 | 2019-06-24T05:11:31 | {} | NONE |
ClickHouse | ClickHouse | 505,203,120 | 3,885 | den-crane | ```
SELECT k, v, d, i FROM
( SELECT t.1 AS k, t.2 AS v,
runningDifference(v) AS d,
runningDifference(xxHash32(t.1)) AS i
FROM
( SELECT arrayJoin([('a', 1), ('a', 2), ('a', 3), ('b', 11), ('b', 13), ('b', 15)]) AS t))
WHERE i = 0
WRONG RESULT
enable_optimize_predicate_expression=... | 2019-06-24T22:19:22 | 2019-06-24T22:19:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,543,614 | 5,736 | 4ertus2 | There's only equi-join supported. It means ON section could have only equality between columns. And it's an algorithm restriction. There're no plans to support some special algorithm for not equal ON sections. You could rewrite such the joins with combination of CROSS JOINs and equi-JOINs. In some cases (if you have at... | 2019-06-25T17:29:03 | 2019-06-25T17:29:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,720,264 | 5,735 | shivacr | Filesystem Size Used Avail Use% Mounted on
/dev/sda5 92G 28G 60G 32% /
Available is 60G not 28G.
What should be the ideal free space for clickhouse to run ?
There are no insertions going on now, so the space seems bigger
root@rdserv-1~# df -h /var/lib/clickhouse/
Filesystem Size Used A... | 2019-06-26T05:08:22 | 2019-06-26T05:09:49 | {} | NONE |
ClickHouse | ClickHouse | 505,867,808 | 5,746 | den-crane | >version : 18.16.1
Is not supported already.
Check ZK logs for errors at the same period time. | 2019-06-26T13:10:24 | 2019-06-26T13:10:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,237,954 | 5,596 | alexey-milovidov | Let's make part size configurable (by `Settings.h`) and 512 MB by default. We will collect every part in a buffer in memory before sending. We should not allocate all 512 MB for the first part, let's grow buffer from 1 MB to 512 MB exponentially.
If 10 000 parts are not enough, throw an exception. | 2019-06-20T23:46:58 | 2019-06-20T23:47:25 | {} | MEMBER |
ClickHouse | ClickHouse | 504,338,345 | 5,582 | blind-oracle | Yes, increasing the `age` works, but we should either fix this or deny setting `age` to <= `precision` :) | 2019-06-21T08:21:20 | 2019-06-21T08:21:20 | {} | NONE |
ClickHouse | ClickHouse | 504,366,009 | 5,707 | onine007 |
| 2019-06-21T09:49:32 | 2019-06-21T09:49:32 | {} | NONE |
ClickHouse | ClickHouse | 504,596,223 | 5,582 | den-crane | https://t.me/c/1425947904/1706
```
Denny Crane, [Jun 21, 2019 at 7:39:33 PM (2019-06-21, 7:55:42 PM)]:
попробовал снова и вот MV которое сморит на graphite_rollup (AVG , age = 120 , precision = 60)
graphite_crap_mv
│ 1 │ 2019-06-21 21:13:28 │
│ 3 │ 2019-06-21 21:13:58 │
│ 1 │ 2019-06-21 21:14:28 │... | 2019-06-21T22:47:20 | 2019-06-21T22:56:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,643,214 | 5,582 | blind-oracle | У меня пока ошибок мержа не наблюдается под продуктивной нагрузкой с `age=120` `precision=60`, буду мониторить дальше и скрипт погоняю еще тоже с новыми настройками. | 2019-06-22T08:21:43 | 2019-06-22T08:21:43 | {} | NONE |
ClickHouse | ClickHouse | 504,645,839 | 5,717 | excitoon | FYI @kmeaw current AUR build is broken :( | 2019-06-22T08:52:21 | 2019-06-22T08:52:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,859,231 | 5,612 | CAspeling | Would anyone please be able to look into this this week? | 2019-06-24T05:07:16 | 2019-06-24T05:07:16 | {} | NONE |
ClickHouse | ClickHouse | 504,913,629 | 1,980 | ghost | @hwchen It is possible to `COALESCE` nullable values to some "empty" value, then it would work with `groupArray` keeping the same number of items. | 2019-06-24T08:32:49 | 2019-06-24T08:32:49 | {} | NONE |
ClickHouse | ClickHouse | 505,696,815 | 5,736 | sudabon | Hi Artem,
thanks for your quick reply.
I understood an algorithm restriction and no plan to support them.
I try to use comparison operators in WHERE section. | 2019-06-26T02:53:40 | 2019-06-26T02:53:40 | {} | NONE |
ClickHouse | ClickHouse | 505,829,876 | 5,747 | alesapin | No, currently CH doesn't support kerberos. But we have plans to add it. | 2019-06-26T11:09:52 | 2019-06-26T11:09:52 | {} | MEMBER |
ClickHouse | ClickHouse | 504,219,200 | 5,582 | den-crane | oops, too early, it still has the issue.
```
┌─Value─┬────────────────Time─┐
│ 2 │ 2019-06-20 22:02:00 │
│ 2 │ 2019-06-20 22:03:00 │
│ 2 │ 2019-06-20 22:04:00 │
│ 2 │ 2019-06-20 22:05:00 │
│ 2 │ 2019-06-20 22:06:00 │
│ 3 │ 2019-06-20 22:07:00 │
│ 3 │ 2019-06-20 22:08:00 │
│ 3 │... | 2019-06-20T22:15:11 | 2019-06-20T22:15:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,418,824 | 5,485 | vincenzon | I am hitting an issue using runningDifference that may be me misunderstanding how the function is applied, but here is an example riffing on the example from @den-crane .
Here I difference, but there is a jump as the key k changes from 'a' to 'b'.
```
SELECT
t.1 AS k,
t.2 AS v,
runningDifference... | 2019-06-21T13:07:30 | 2019-06-21T13:07:30 | {} | NONE |
ClickHouse | ClickHouse | 504,634,292 | 5,596 | excitoon | >
>
> Let's make part size configurable (by `Settings.h`) and 512 MB by default. We will collect every part in a buffer in memory before sending. We should not allocate all 512 MB for the first part, let's grow buffer from 1 MB to 512 MB exponentially.
>
> If 10 000 parts are not enough, throw an exception.
I... | 2019-06-22T06:01:54 | 2019-06-22T06:01:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,767,488 | 5,298 | TimonKK | Hi, can you create merge request for ClickHouse with this path? | 2019-06-23T16:34:18 | 2019-06-23T16:34:18 | {} | NONE |
ClickHouse | ClickHouse | 504,966,594 | 5,681 | MeteHanC | This would be good for pagination too, limit&offset has very poor performance if you have millions of rows. Would be good if we could use auto incremented indexed fields to make the pagination. | 2019-06-24T11:15:40 | 2019-06-24T11:15:40 | {
"+1": 6
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,976,637 | 5,719 | excitoon | Okay!
On Mon, Jun 24, 2019, 14:19 proller <notifications@github.com> wrote:
> It is possible to place new files to utils/package/arch/ ?
>
> —
> You are receiving this because you authored the thread.
> Reply to this email directly, view it on GitHub
> <https://github.com/yandex/ClickHouse/pull/5719?email_source=noti... | 2019-06-24T11:51:06 | 2019-06-24T11:51:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,136,773 | 5,731 | 4ertus2 | related #5557 | 2019-06-24T18:58:43 | 2019-06-24T18:58:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,207,777 | 5,485 | vincenzon | Thanks @den-crane. | 2019-06-24T22:39:19 | 2019-06-24T22:39:19 | {} | NONE |
ClickHouse | ClickHouse | 505,264,997 | 5,689 | Aricliang | mark | 2019-06-25T03:19:33 | 2019-06-25T03:19:33 | {} | NONE |
ClickHouse | ClickHouse | 505,365,996 | 4,631 | k0st1an | Looks like #5680 . Is there a deadline for the decision? This is a critical problem. | 2019-06-25T09:31:58 | 2019-06-25T09:31:58 | {} | NONE |
ClickHouse | ClickHouse | 505,546,275 | 3,783 | 4ertus2 | No reproduction. Probably fixed. | 2019-06-25T17:35:58 | 2019-06-25T17:35:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,546,865 | 5,699 | abyss7 | > @abyss7 A newer version has been released. See here: https://github.com/edenhill/librdkafka/releases/tag/v1.1.0
Thanks for mention it. Will update a little bit later. | 2019-06-25T17:37:36 | 2019-06-25T17:37:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,778,738 | 3,174 | xmariachi | > @filimonov If there are more than one part per partition and their sizes are smaller than maximum part size - that means they will be chosen for merge someday. The smaller part is the bigger chance that it will be merged soon.
Does that mean, that if the sizes are bigger than the maximum part size, they will never... | 2019-06-26T08:37:21 | 2019-06-26T08:37:21 | {} | NONE |
ClickHouse | ClickHouse | 504,644,925 | 5,716 | excitoon | This is a duplicate of #5316 | 2019-06-22T08:42:47 | 2019-06-22T08:42:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,645,763 | 5,169 | excitoon | FYI @kmeaw | 2019-06-22T08:51:33 | 2019-06-22T08:51:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 504,775,851 | 5,724 | MironAtHome | Может кто думает, что русский язык святым духом живет, но вообще - то истоки, не на русском языке, это просто позор.
А я думал здесь нормальные... | 2019-06-23T18:31:15 | 2019-06-23T18:31:15 | {} | NONE |
ClickHouse | ClickHouse | 505,621,387 | 5,725 | alesapin | Integration test failure is not associated with changes. | 2019-06-25T21:02:09 | 2019-06-25T21:02:09 | {} | MEMBER |
ClickHouse | ClickHouse | 505,779,103 | 5,742 | alexey-milovidov | You can also compare with different levels of ZSTD, because level 1 is used by default.
When comparing by compressed size, you should also mention compression and decompression speed. | 2019-06-26T08:38:19 | 2019-06-26T08:38:19 | {} | MEMBER |
ClickHouse | ClickHouse | 505,856,030 | 5,735 | den-crane | >What should be the ideal free space for clickhouse to run ?
Depends of your requirements and environment.
You can move DB to /home from /var using symlinks or using CH settings
```
<!-- Path to data directory, with trailing slash. -->
<path>/var/lib/clickhouse/</path>
<!-- Path to temporary data ... | 2019-06-26T12:36:34 | 2019-06-26T12:37:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,891,617 | 5,755 | 4ertus2 | There're places where we check isColumnNullable() and then static_cast it to ColumnNullable. It's totally wrong cause no one column could have isColumnNullable() returning true. Interface method is used as type checker. | 2019-06-26T14:08:20 | 2019-06-26T14:08:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,201,979 | 5,485 | den-crane | @vincenzon enable_optimize_predicate_expression bug
19.4.3.11
```
┌─k─┬──v─┬─d─┬─i─┐
│ a │ 1 │ 0 │ 0 │
│ a │ 2 │ 1 │ 0 │
│ a │ 3 │ 1 │ 0 │
│ b │ 13 │ 2 │ 0 │
│ b │ 15 │ 2 │ 0 │
└───┴────┴───┴───┘
```
19.10.1.669
```
┌─k─┬──v─┬──d─┬─i─┐
│ a │ 1 │ 0 │ 0 │
│ a │ 2 │ 1 │ 0 │
│ a │ 3 │ 1 │ 0 │
│ ... | 2019-06-24T22:14:50 | 2019-06-24T22:14:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 505,228,755 | 5,734 | den-crane | https://www.altinity.com/blog/clickhouse-in-the-storm-part-1
https://www.altinity.com/blog/clickhouse-in-the-storm-part-2
https://www.altinity.com/blog/tag/Time+Series | 2019-06-25T00:20:39 | 2019-06-25T00:20:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,467,336 | 6,439 | alesapin | Maybe now we don't need to enable this setting by default? | 2019-08-12T15:13:50 | 2019-08-12T15:13:50 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.