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 | 479,897,530 | 4,736 | abyss7 | > @abyss7 Is there any estimate/milestone when this bug is going to be fixed?
I will investigate this next week - no estimation on the fix for now. | 2019-04-04T13:30:19 | 2019-04-04T13:30:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,345,448 | 4,921 | alexey-milovidov | The issue has been fixed in #4782 | 2019-04-05T16:49:16 | 2019-04-05T16:49:16 | {} | MEMBER |
ClickHouse | ClickHouse | 480,498,846 | 4,926 | nvartolomei | It works pretty much similar to how MV works on inserted data. The query is applied for each new "block".
From MV docs:
> A materialized view is arranged as follows: when inserting data to the table specified in SELECT, part of the inserted data is converted by this SELECT query, and the result is inserted in the... | 2019-04-06T12:04:36 | 2019-04-06T12:04:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,562,971 | 4,933 | andrea-s | Thank you for your kind response @filimonov!
It it true that in this (somewhat simplified) version of the problem one could use a subquery like this - but doing as you suggest will "pre-rollup" the data, preventing metrics that depend on the "account_id" field to be calculated. Examples of such metrics:
number_of... | 2019-04-07T06:22:59 | 2019-04-07T06:22:59 | {} | NONE |
ClickHouse | ClickHouse | 480,975,434 | 4,942 | filimonov | That is wrong.
```
:) select windowFunnel(3600)(ts, number = 1, number = 2, number = 3, number = 4) from ( select toUInt32(3599*number) as ts, number from numbers(5) );
SELECT windowFunnel(3600)(ts, number = 1, number = 2, number = 3, number = 4)
FROM
(
SELECT
toUInt32(3599 * number) AS ts,
... | 2019-04-08T19:33:35 | 2019-04-08T19:33:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,154,744 | 4,629 | bgranvea | @alexey-milovidov I've rebased the pull request because there was a conflict. It is ready for reviewing. | 2019-04-09T08:28:35 | 2019-04-09T08:28:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,323,353 | 4,886 | alex-zaitsev | @bzon This is exactly what we are planning to do. The core functionality is pretty much complete. We should have 0.2 release with core functionality complete in a few days, and switch to testing and operational documentation after that.
What is your use case? Are you going to deploy many ClickHouse clusters, or a si... | 2019-04-09T16:24:29 | 2019-04-09T16:24:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,412,700 | 4,212 | alesapin | @CurtizJ Your own tests failed. | 2019-04-09T20:00:56 | 2019-04-09T20:00:56 | {} | MEMBER |
ClickHouse | ClickHouse | 481,599,017 | 4,248 | Junus | @turbotankist did you found problem, why this happening? | 2019-04-10T08:45:21 | 2019-04-10T08:45:21 | {} | NONE |
ClickHouse | ClickHouse | 481,620,384 | 4,945 | peaksnail |
I solve it by adding `order by name,category,total` when I read the [link](https://www.percona.com/blog/2017/04/07/non-deterministic-order-select-limit/)
the name is uniq, and the categgory/total will be the same value | 2019-04-10T09:44:05 | 2019-04-10T09:44:05 | {} | NONE |
ClickHouse | ClickHouse | 481,687,079 | 4,879 | 4ertus2 | It looks like wrong name qualification. t2.id and t2.name unexpectedly became t1.id and t1.name.
```
select t2.name from remote('127.0.0.2', 'test.local_table') as t1 global left join test.other_table as t2 on t1.oth_id = t2.id;
Code: 47. DB::Exception: Received from localhost:9000, 127.0.0.1. DB::Exception: Mis... | 2019-04-10T13:18:50 | 2019-04-10T13:18:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,692,880 | 4,959 | free6k | @alesapin what is it? I have default config | 2019-04-10T13:33:57 | 2019-04-10T13:33:57 | {} | NONE |
ClickHouse | ClickHouse | 481,697,167 | 4,959 | free6k | How? dump? | 2019-04-10T13:45:22 | 2019-04-10T13:45:22 | {} | NONE |
ClickHouse | ClickHouse | 480,042,130 | 4,906 | alexey-milovidov | Simple `std::vector` actually can be copied with memcpy/memmove correctly.
Although it is not guaranteed by the standard, we assume that it contains three fields that are sizes or pointers to the single allocated memory region. When we copy these fields and forget about old object, nothing bad can happen.
`std::i... | 2019-04-04T20:00:15 | 2019-04-04T20:04:32 | {} | MEMBER |
ClickHouse | ClickHouse | 481,113,973 | 4,388 | fandyushin | Ok, I reproduced its localy. Apparently the reason is in the unbounded POCO library. Its looks like poco server response destroyed prematurely. | 2019-04-09T06:01:45 | 2019-04-09T06:01:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,652,420 | 4,947 | alesapin | Seems like bug in entrtypoint.sh https://github.com/yandex/ClickHouse/blob/master/docker/server/entrypoint.sh#L36. As workaround you can add following lines to `config.xml`:
```(xml)
<user_files_path>/var/lib/clickhouse/user_files</user_files_path>
<format_schema_path>/var/lib/clickhouse/format_schema</forma... | 2019-04-10T11:30:27 | 2019-04-10T11:30:27 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 481,664,910 | 4,955 | nvartolomei | Likely duplicate of #1344. | 2019-04-10T12:13:02 | 2019-04-10T12:13:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,672,311 | 4,958 | Felixoid | Test with config from #4947
```
┌[14:34:04] felixoid@ig-fel:~/OPT/Felixoid/github/yandex/ClickHouse/docker/server/ [entrypoint_skip_empty|…1]
└>±> docker run -v ${PWD}/config.xml:/etc/clickhouse-server/config.xml yandex/clickhouse-server:19.4.3.11
Poco::Exception. Code: 1000, e.code() = 0, e.displayText() = No... | 2019-04-10T12:35:51 | 2019-04-10T12:37:55 | {} | MEMBER |
ClickHouse | ClickHouse | 479,856,509 | 2,070 | abyss7 | Please, try the same scenario on a recent CH version, and create new issue if the problem persists. | 2019-04-04T11:18:46 | 2019-04-04T11:18:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,282,109 | 4,917 | KochetovNicolai | I have changed the return type from `Tuple(Nullable(Float64) * 2)` to `Tuple(Float64 * 2)`. | 2019-04-05T13:49:04 | 2019-04-05T13:49:04 | {} | MEMBER |
ClickHouse | ClickHouse | 480,445,317 | 4,913 | alexey-milovidov | There is a (logical) race condition in SystemLog.
`SYSTEM FLUSH LOGS` does not guarantee that all data is written, because the data may be still in queue. The method `flush` called by user, should also drain the queue.
The issue may happen in other tests as well. | 2019-04-05T22:47:39 | 2019-04-05T22:48:42 | {} | MEMBER |
ClickHouse | ClickHouse | 479,872,079 | 3,779 | chenxing-xc | @alexey-milovidov which delivery include this fix? I encounter exact the same issue in my env, and want to patch the fix. | 2019-04-04T12:12:51 | 2019-04-04T12:12:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,026,057 | 4,911 | alexey-milovidov | It can be fixed but requires performance testing.
The relevant fragment in code is:
https://github.com/yandex/ClickHouse/blob/master/dbms/src/IO/ReadHelpers.h#L614
The function `readIntTextImpl` has no limit on number length and no overflow check (for performance reasons). We can add another function with addition... | 2019-04-04T19:11:00 | 2019-04-04T19:11:22 | {} | MEMBER |
ClickHouse | ClickHouse | 480,562,152 | 4,933 | filimonov | While your propositions sound interesting, it looks like you can solve your task in more natural way currently
```sql
SELECT
sum(comments),
argMax(fans, day)
FROM
(
SELECT
day,
sum(number_of_fans) AS fans, sum(post_comments) AS comments
FROM table
GROUP BY day
)
``` | 2019-04-07T06:04:50 | 2019-04-07T06:04:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,776,393 | 4,924 | 4ertus2 | I've changed and renamed 00927_asof_join_long.sql
It triggered tests timeouts and out of memory/disk for test environment.
Names of big tests should contain word 'long' cause there's option --no-long for ctest run. | 2019-04-08T10:30:00 | 2019-04-08T10:30:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,853,940 | 4,923 | KochetovNicolai | Thanks! | 2019-04-08T14:25:30 | 2019-04-08T14:25:30 | {} | MEMBER |
ClickHouse | ClickHouse | 481,639,303 | 3,700 | yutrostin | сейчас использую версию КХ `19.4.0`
Воспроизведение проблемы:
1. Поднимаете постгрес, создаете базу данных `postgres_database` и в ней таблицу `users` со следующими столбцами:
* created_at DateTime NOT NULL
* birthday Date
* name Varchar
Создаете в ней записи
2. Настраиваете на сервере КХ DSN так, чтобы можн... | 2019-04-10T10:44:00 | 2019-04-10T10:44:00 | {} | NONE |
ClickHouse | ClickHouse | 481,698,233 | 4,959 | free6k | yes, free6k nickname | 2019-04-10T13:48:10 | 2019-04-10T13:48:10 | {} | NONE |
ClickHouse | ClickHouse | 479,754,750 | 4,902 | kindred77 | May be you can use the "on cluster" option. | 2019-04-04T05:24:56 | 2019-04-04T05:24:56 | {} | NONE |
ClickHouse | ClickHouse | 480,036,577 | 3,779 | alexey-milovidov | I don't remember the exact commit. It was one of the @4ertus2 fixes for JOINs. | 2019-04-04T19:43:23 | 2019-04-04T19:43:23 | {} | MEMBER |
ClickHouse | ClickHouse | 480,331,940 | 4,913 | alexey-milovidov | Your own test has failed. | 2019-04-05T16:06:26 | 2019-04-05T16:06:26 | {} | MEMBER |
ClickHouse | ClickHouse | 480,489,989 | 4,924 | Gladdy | I'm a bit puzzled why the latter two builds fail some of the performance tests. A quick inspection shows that moving the atomic has some performance impact even when the test doesn't even perform an asof join, which indicates that the failures have nothing to do with my changes and are just flukes? | 2019-04-06T09:35:21 | 2019-04-06T09:35:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,187,950 | 4,194 | pikhovkin | Здравствуйте!
Почти такая же проблема с памятью. Только память уходит значительно быстрее после определенного кол-ва вставленных строк и по разному на разных табличных движках.
CH 19.3.3 revision 54415.
ОЗУ 32Гб.
Надо импортировать 50млн записей. Каждая запись 90 колонок, заполнено в среднем колонок 20 числами,... | 2019-04-09T10:04:32 | 2019-04-09T10:04:32 | {} | NONE |
ClickHouse | ClickHouse | 481,696,384 | 4,959 | alesapin | Maybe you can provide some data to reproduce the error? | 2019-04-10T13:43:21 | 2019-04-10T13:43:21 | {} | MEMBER |
ClickHouse | ClickHouse | 479,842,853 | 4,902 | blinkov | ...or maybe https://clickhouse.yandex/docs/zh/query_language/create/#fen-bu-shi-ddlcha-xun-on-cluster-zi-ju given the punctuation you use | 2019-04-04T10:31:11 | 2019-04-04T10:31:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,051,138 | 4,893 | alexey-milovidov | The minimum version is `v19.5.1.246` that is available in testing. | 2019-04-04T20:27:58 | 2019-04-04T20:27:58 | {} | MEMBER |
ClickHouse | ClickHouse | 481,195,082 | 3,700 | alesapin | На какой версии ClickHouse это воспроизводится? Сделайте, пожалуйста, полный пример воспроизведения проблемы.
> Использую версию v1.0.0.20181017
Это версия чего? | 2019-04-09T10:26:20 | 2019-04-09T10:26:28 | {} | MEMBER |
ClickHouse | ClickHouse | 481,537,101 | 4,388 | fandyushin | @alexey-milovidov, finaly im done. Im sorry it took so long. | 2019-04-10T05:17:40 | 2019-04-10T05:17:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,676,366 | 4,953 | yedy99 | > Looks like you show CMakeOutput.log and CMakeError.log , but what cmake prints to screen ?
-- The C compiler identification is GNU 8.3.0
-- The CXX compiler identification is GNU 8.3.0
-- Checking whether C compiler has -isysroot
-- Checking whether C compiler has -isysroot - yes
-- Checking whether C compiler... | 2019-04-10T12:48:12 | 2019-04-10T12:48:12 | {} | NONE |
ClickHouse | ClickHouse | 481,677,591 | 4,947 | alesapin | Merged, will be fixed in next releases (19.4, 19.5, 19.6) | 2019-04-10T12:51:34 | 2019-04-10T12:51:34 | {} | MEMBER |
ClickHouse | ClickHouse | 481,691,919 | 4,959 | alesapin | Do you use setting `min_bytes_to_use_direct_io`? | 2019-04-10T13:31:30 | 2019-04-10T13:31:30 | {} | MEMBER |
ClickHouse | ClickHouse | 480,051,512 | 4,888 | alexey-milovidov | The fix is not released yet. | 2019-04-04T20:29:04 | 2019-04-04T20:29:04 | {
"eyes": 2
} | MEMBER |
ClickHouse | ClickHouse | 480,502,307 | 4,903 | kindred77 | thanks | 2019-04-06T13:00:59 | 2019-04-06T13:00:59 | {} | NONE |
ClickHouse | ClickHouse | 480,777,581 | 4,924 | Gladdy | Thanks, sorry, I missed that. Might also be wise to just move it with the other performance tests which I didn't get around to yet. | 2019-04-08T10:34:10 | 2019-04-08T10:34:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,545,420 | 4,939 | blinkov | тесты не прошли, а так вроде норм | 2019-04-10T06:01:00 | 2019-04-10T06:01:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,676,254 | 4,954 | 4ertus2 | Decimal conversion for external storages and dictionaries is not supported yet.
As workaround you may try to create Decimal ALIAS columns for Strings ones with toDecimal() expression. | 2019-04-10T12:47:53 | 2019-04-10T12:48:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 479,845,754 | 4,904 | blinkov | @vkingnew how exactly have you've tried to increase the memory limit? | 2019-04-04T10:41:08 | 2019-04-04T10:41:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,272,643 | 1,370 | 4ertus2 | Already fixed
```
CREATE TABLE payments
(
`net` UInt32,
`payments_amount` UInt32
)
ENGINE = Memory
Ok.
0 rows in set. Elapsed: 0.019 sec.
SELECT avg(net) AS payments_amount
FROM payments AS t
WHERE t.payments_amount > 0
┌─payments_amount─┐
│ nan │
└─────────────────┘
1 r... | 2019-04-05T13:20:33 | 2019-04-05T13:20:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,331,329 | 4,915 | alexey-milovidov | Missing `backQuoteIfNeed` but still Ok. | 2019-04-05T16:04:41 | 2019-04-05T16:04:41 | {} | MEMBER |
ClickHouse | ClickHouse | 480,511,167 | 4,927 | alexey-milovidov | More details: https://clickhouse-test-reports.s3.yandex.net/0/f862bb2ed291382e820e5b8656520fa50667b55c/stress_test_(thread).html | 2019-04-06T15:04:14 | 2019-04-06T15:04:14 | {} | MEMBER |
ClickHouse | ClickHouse | 481,360,482 | 4,242 | 4ertus2 | Fix unknown identifier #4952
It also fix optimisation for first condition
```
SELECT count() FROM constCondOptimization WHERE if(0, 1, n = 1000);
1 rows in set. Elapsed: 0.038 sec.
```
Where's one unresolved issue here
```
SELECT count() FROM constCondOptimization WHERE n = 1000 AND time between now() - ... | 2019-04-09T17:47:16 | 2019-04-10T12:21:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,687,910 | 4,879 | 4ertus2 | As workaround you may change the tables order in query:
```
select t2.name from test.other_table as t2 global right join remote('127.0.0.2', 'test.local_table') as t1 on t1.oth_id = t2.id;
Two Hundred
One Hundred
``` | 2019-04-10T13:21:12 | 2019-04-10T13:21:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,357,944 | 4,904 | vkingnew | The default memory about 10G，I tried 20G ,30G,40G appears the same error.The Mache total Memory is 48G.And I how to avoid this error message?I need to change some settings? | 2019-04-05T17:30:15 | 2019-04-05T17:30:15 | {} | NONE |
ClickHouse | ClickHouse | 480,794,668 | 4,902 | yunhuizhu | @blinkov @kindred77 thanks,we had used old 2017 clickhouse-server,now upgrade the newest clickhouse-server,it is usefull | 2019-04-08T11:33:30 | 2019-04-08T11:33:30 | {} | NONE |
ClickHouse | ClickHouse | 481,693,672 | 4,879 | dfulachier | Thank you Artem for studying this bug, and for all your workaround proposals.
None can be systematically applied. By instance, I am not sure the latest is transparent in term of performances (should we not always keep the large table on the left of join, and the small one on the right...?).
This regression is quite a... | 2019-04-10T13:36:06 | 2019-04-10T13:36:06 | {} | NONE |
ClickHouse | ClickHouse | 481,680,161 | 4,955 | gusevartemasd | > I did make a workaround for simple use cases. Add time and param id to primary key. Then run a query to group number of rows for a param is by buckets (say 1h bucket). Find last buckets that contain more than limit (500 in your case) records.
>
> Re-run the query with constraint on time that is derived from time b... | 2019-04-10T12:59:13 | 2019-04-10T13:47:56 | {} | NONE |
ClickHouse | ClickHouse | 480,347,386 | 4,893 | TH-HA | Thanks a lot for your answer.
A small comment on this part , it may be usefull to avoid such kind of question to put on the actual documentation the clickhouse version on which the documentation apply. I was naively and wrongly thinking that it was always on the latest stable version.
Regards | 2019-04-05T16:55:13 | 2019-04-05T16:55:13 | {} | NONE |
ClickHouse | ClickHouse | 480,525,065 | 4,388 | alexey-milovidov | Ok, and what's wrong? | 2019-04-06T18:05:58 | 2019-04-06T18:05:58 | {} | MEMBER |
ClickHouse | ClickHouse | 481,192,969 | 4,194 | blinkov | @Rakovskiy28 память по задумке и не должна освобождаться в операционную систему, но должна переиспользоваться при дальнейшей работе clickhouse-server.
@pikhovkin не стоит делать дедупликацию и замену при вставке через ALTER TABLE. Для этого лучше сохранять предыдущие значения в любом внешним хранилище, оптимизирован... | 2019-04-09T10:20:00 | 2019-04-09T10:20:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,647,963 | 4,957 | alvin85 | Good idea! thanks. | 2019-04-10T11:14:42 | 2019-04-10T11:14:42 | {} | NONE |
ClickHouse | ClickHouse | 481,674,649 | 4,947 | Felixoid | Hey @2ur1st. You could check out the branch from #4958 via
```
git checkout entrypoint_skip_empty
git pull https://github.com/Felixoid/ClickHouse.git entrypoint_skip_empty
```
and build image with fix as
```
cd docker/server
docker build --build-arg=version=19.4.3.11 -t yandex/clickhouse-server:19.4.3.11 .
``... | 2019-04-10T12:43:00 | 2019-04-10T12:43:44 | {} | MEMBER |
ClickHouse | ClickHouse | 479,842,322 | 4,905 | blinkov | This is on roadmap #4786, but for now you should use RAID. | 2019-04-04T10:29:21 | 2019-04-04T10:29:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,043,930 | 4,906 | alexey-milovidov | Summary: no bug exists, but we rely on weak assumption. | 2019-04-04T20:05:55 | 2019-04-04T20:05:55 | {} | MEMBER |
ClickHouse | ClickHouse | 477,250,658 | 4,242 | 4ertus2 | The last result is an error itself. #4858
```
USE test;
DROP TABLE IF EXISTS t1;
DROP TABLE IF EXISTS t2;
CREATE TABLE t1 (n Int32) ENGINE = Memory;
CREATE TABLE t2 (a Int32, n Int32) ENGINE = Memory;
SELECT count() FROM t1 WHERE if(1, 1, n = 0); -- OK
SELECT count() FROM t2 WHERE if(1, 1, n = 0); -- FAIL... | 2019-03-27T16:52:08 | 2019-04-05T14:49:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,397,713 | 4,923 | KochetovNicolai | It would be helpful if you can show some rows that can't match.
Can you also check that it's a problem with LowCardinality? Just cast LowCardinality to String:
```sql
SELECT TS,Region,Metric, QuantAvg, TrendAvg FROM
(
SELECT
TS, cast(Region as String) as Region, cast(Metric as String) as Metr... | 2019-04-05T19:37:22 | 2019-04-05T19:37:22 | {} | MEMBER |
ClickHouse | ClickHouse | 480,410,194 | 4,913 | alesapin | Have no idea why. Seems like query was absent in `system.query_log`, but server log contains:
```
2019.04.05 16:36:30.836471 [ 29 ] {a718ce00-620d-4975-898f-906bbcbecca0} <Debug> executeQuery: (from [::1]:56788) SELECT ProfileEvents.Values[indexOf(ProfileEvents.Names, 'Seek')], ProfileEvents.Values[indexOf(ProfileEve... | 2019-04-05T20:22:30 | 2019-04-05T20:23:46 | {} | MEMBER |
ClickHouse | ClickHouse | 480,820,978 | 4,923 | agaurav | The above query didn't work as well but its seems removing `LowCardinality` field as partitioning key and recreating the tables worked.
old partitioning key : `PARTITION BY (toYYYYMM(LastAggr), Region, Phase)`
new partitioning key : `PARTITION BY (toYYYYMM(LastAggr), Phase)`
where the LastAggr,Region and ... | 2019-04-08T12:58:52 | 2019-04-08T13:04:18 | {} | NONE |
ClickHouse | ClickHouse | 481,520,361 | 4,953 | yedy99 | here is the CMakeError.log
Compiling the CXX compiler identification source file "CMakeCXXCompilerId.cpp" failed.
Compiler: /usr/local/bin/g++-8
Build flags: -march=core2;-mtune=haswell;-mssse3;-ftree-vectorize;-fPIC;-fPIE;-fstack-protector-strong;-O2;-pipe;-stdlib=libc++;-fvisibility-inlines-hidden;-std=c++14;-f... | 2019-04-10T03:36:08 | 2019-04-10T03:36:08 | {} | NONE |
ClickHouse | ClickHouse | 481,559,813 | 4,939 | BayoNet | Пофиксил ссылки. Локально собралось вроде. | 2019-04-10T06:57:57 | 2019-04-10T06:57:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,183,641 | 4,944 | alesapin | Cannot reproduce on official (ubuntu) 19.4.2.7 package:
```
:) CREATE TABLE default.testTable (SomeDate DateTime, SomeId UInt32, SomeCount UInt64) ENGINE = SummingMergeTree([SomeCount]) PARTITION BY toYYYYMM(SomeDate) ORDER BY (SomeDate);
CREATE TABLE default.testTable
(
SomeDate DateTime,
SomeId UInt3... | 2019-04-09T09:51:36 | 2019-04-09T09:53:19 | {} | MEMBER |
ClickHouse | ClickHouse | 481,324,422 | 4,886 | bzon | @alex-zaitsev Perfect! We currently have a single Clickhouse cluster to serve 1 billion rows per day. | 2019-04-09T16:27:21 | 2019-04-09T16:27:21 | {} | NONE |
ClickHouse | ClickHouse | 481,698,021 | 4,959 | alesapin | You can compress partitions stored on disk or use `SELECT xxx from yyy INTO OUTFILE 'xxx.tmp' Format TSV`. Can we communicate in Telegram? | 2019-04-10T13:47:36 | 2019-04-10T13:47:36 | {} | MEMBER |
ClickHouse | ClickHouse | 479,845,328 | 4,903 | blinkov | This obviously needs to be done at some point, but no ETA yet. | 2019-04-04T10:39:46 | 2019-04-04T10:39:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,272,785 | 4,518 | jangorecki | @a-rodin Thanks, that resolves my problem. I would say it is not a bug then, but a lack of documentation. | 2019-04-05T13:21:01 | 2019-04-05T13:21:01 | {} | NONE |
ClickHouse | ClickHouse | 480,354,377 | 4,893 | alexey-milovidov | Yes, this is a frequent complaint. Today, the docs are not in sync with any ClickHouse release.
(For most features, the docs lag behind, but there are some features where docs are prepared in advance.)
We're only going to prepare a better process of docs publishing. I cannot guarantee when it will happen. | 2019-04-05T17:18:32 | 2019-04-05T17:18:32 | {} | MEMBER |
ClickHouse | ClickHouse | 480,814,230 | 4,929 | 4ertus2 | #4938 | 2019-04-08T12:38:36 | 2019-04-08T12:38:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,829,270 | 4,695 | SaltTan | Here is an excerpt from CH logs when it hangs initializing dictionaries.
dictionaries_lazy_load was set to true.
[initializing_dictionaries.txt](https://github.com/yandex/ClickHouse/files/3054435/initializing_dictionaries.txt)
| 2019-04-08T13:22:25 | 2019-04-08T13:22:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,135,243 | 4,946 | blinkov | Sign column should have 1 or -1 values, you need to fix the program that have written 0 there. | 2019-04-09T07:27:01 | 2019-04-09T07:27:01 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,644,874 | 4,957 | blinkov | At the moment only via CROSS JOIN + WHERE | 2019-04-10T11:03:45 | 2019-04-10T11:03:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,648,597 | 4,953 | proller | Looks like you show CMakeOutput.log and CMakeError.log , but what cmake prints to screen ? | 2019-04-10T11:17:04 | 2019-04-10T11:17:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,666,243 | 4,955 | nvartolomei | I did make a workaround for simple use cases. Add time and param id to primary key. Then run a query to group number of rows for a param is by buckets (say 1h bucket). Find last buckets that contain more than limit (500 in your case) records.
Re-run the query with constraint on time that is derived from time bucket... | 2019-04-10T12:17:03 | 2019-04-10T12:17:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 479,847,644 | 4,865 | abyss7 | Yes. That's the fix. | 2019-04-04T10:47:36 | 2019-04-04T10:47:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,902,670 | 4,848 | BayoNet | Вроде поправил. | 2019-04-08T16:24:55 | 2019-04-08T16:24:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,999,420 | 4,388 | fandyushin | Its segfaults here, in WriteBufferFromOStream distructor, because ostream pointer is null. Only in unbundled build and only if using any compression. Using internal brotli lib doesn't help.
```
#0 0x00007ffff645a6fc in std::ostream::sentry::sentry(std::ostream&) ()
from /usr/lib/x86_64-linux-gnu/libstdc++.so.6
... | 2019-04-08T20:41:12 | 2019-04-08T20:41:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,133,800 | 4,945 | blinkov | Are there any writes to this table? If not, could you produce the full script to reproduce this? | 2019-04-09T07:22:05 | 2019-04-09T07:22:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,200,908 | 4,194 | pikhovkin | @blinkov движки *MergeTree не гарантируют уникальность записей. А при выборке с агрегацией нет возможности получить строковое поле или min\max. Из-за адски быстрой работы выборки в CH меня устраивает медленная запись при дедупликации (~16rps). Но куда расходуется память? | 2019-04-09T10:46:59 | 2019-04-09T10:46:59 | {} | NONE |
ClickHouse | ClickHouse | 481,485,778 | 4,944 | serge-r | It's a packages from Altinity_clickhouse repo:
```
[root@noc ~]# cat /etc/yum.repos.d/Altinity_clickhouse.repo
[Altinity_clickhouse]
name=Altinity_clickhouse
baseurl=https://packagecloud.io/Altinity/clickhouse/el/7/$basearch
repo_gpgcheck=1
gpgcheck=0
enabled=1
gpgkey=https://packagecloud.io/Altinity/click... | 2019-04-10T00:13:30 | 2019-04-10T00:13:30 | {} | NONE |
ClickHouse | ClickHouse | 481,693,885 | 4,959 | alesapin | Ok, which version of clickhouse do you use? | 2019-04-10T13:36:40 | 2019-04-10T13:36:40 | {} | MEMBER |
ClickHouse | ClickHouse | 479,563,564 | 4,875 | Gladdy | Ok. After running some bigger tests under perf which I'll add soon, I'll try to also have a look at the locking structure (obviously the inserts have to be guarded), but I hope to make it slightly more multicore friendly than having a single std::mutex protecting all of the AsofRowRefs (needed per object for the sortin... | 2019-04-03T16:30:00 | 2019-04-03T16:30:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,526,558 | 4,904 | 4ertus2 | There's a related issue #4731
The reason why you get it is disabled enable_optimize_predicate_expression. That is still experimental feature. You may try to enable it or use workaround and rewrite your query with subselects duplicating where sections into them manually:
```
from (selct * from SaleOrders where AND ... | 2019-04-06T18:26:16 | 2019-04-06T18:30:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,527,439 | 4,924 | 4ertus2 | > are just flukes?
Probably they are. Auto perf tests are new and not stable enough. | 2019-04-06T18:37:37 | 2019-04-06T18:37:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,652,010 | 4,905 | RunningHu | > This is on roadmap #4786, but for now you should use RAID.
ok, I got it, thank you.
| 2019-04-08T01:19:01 | 2019-04-08T01:19:01 | {} | NONE |
ClickHouse | ClickHouse | 481,302,667 | 4,882 | filimonov | vs https://github.com/yandex/ClickHouse/pull/4771 ? | 2019-04-09T15:31:49 | 2019-04-09T15:31:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,694,767 | 4,959 | free6k | @alesapin ClickHouse server version 19.4.2.7. | 2019-04-10T13:39:04 | 2019-04-10T13:39:04 | {} | NONE |
ClickHouse | ClickHouse | 479,842,598 | 4,902 | blinkov | @yunhuizhu see https://clickhouse.yandex/docs/en/query_language/create/#distributed-ddl-queries-on-cluster-clause | 2019-04-04T10:30:23 | 2019-04-04T10:30:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 480,341,773 | 4,921 | aadant | Incidently, the clickhouse will not restart after creating such a table ...
Workaround : edit the file to add the missing backticks. I suggest to escape all column and table name with backticks in mysql / mysqldump fashion ...
/var/lib/mysql/clickhouse/metadata/test//reserved_word_table.sql
| 2019-04-05T16:36:54 | 2019-04-05T16:38:28 | {} | NONE |
ClickHouse | ClickHouse | 480,345,948 | 4,921 | alexey-milovidov | The fix is only available in v19.5.1.246-testing.
It's currently in release candidate stage.
You can install this version from testing repository.
```
echo "deb http://repo.yandex.ru/clickhouse/deb/testing/ main/" | sudo tee /etc/apt/sources.list.d/clickhouse.list
``` | 2019-04-05T16:50:47 | 2019-04-05T16:50:47 | {} | MEMBER |
ClickHouse | ClickHouse | 480,895,269 | 4,848 | blinkov | не дает смерджить из-за конфликтов | 2019-04-08T16:04:55 | 2019-04-08T16:04:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 481,682,033 | 4,959 | free6k | if execute query for this part
```
select * from orgs_stats_views where toDate(date) = toDate('2019-01-18 01:01:01') ORDER BY date desc;
```
got error
```
Code: 49, e.displayText() = DB::Exception: Can't adjust last granule because it has 8192rows, but try to subtract 40960 rows.
``` | 2019-04-10T13:04:52 | 2019-04-10T13:04:52 | {} | NONE |
ClickHouse | ClickHouse | 481,699,401 | 4,879 | 4ertus2 | > should we not always keep the large table on the left of join, and the small one on the right...?
You are right. It's not possible to save performance with the last workaround. | 2019-04-10T13:51:00 | 2019-04-10T13:51:07 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.