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 | 421,302,456 | 1,394 | blinkov | Does anyone ended up using ClickHouse on top of fuse in long term? | 2018-09-14T09:59:18 | 2018-09-14T09:59:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,136,220 | 2,770 | alexey-milovidov | > here are results after optimization
That's much better!
> I am unable to implement %U and %W modifiers (week numbers) :(
Just leave it unimplemented (throw exception) and we will have a chance to implement later.
BTW, support for ISO week (%V) was added yesterday (I see you already use it - cool!).
> It ... | 2018-09-17T19:15:16 | 2018-09-17T19:16:09 | {} | MEMBER |
ClickHouse | ClickHouse | 421,457,578 | 3,123 | alexey-milovidov | ```
/usr/bin/ccache /usr/bin/clang++-7 -DBOOST_SYSTEM_NO_DEPRECATED -DLZ4_DISABLE_DEPRECATE_WARNINGS=1 -DPOCO_STATIC -DPOCO_UNBUNDLED_ZLIB -DUNALIGNED_OK -DWITH_GZFILEOP -DX86_64 -DZLIB_COMPAT -isystem contrib/libsparsehash -isystem contrib/libdivide -Idbms/src -isystem contrib/double-conversion -isystem contrib/libp... | 2018-09-14T19:14:26 | 2018-09-14T19:14:26 | {} | MEMBER |
ClickHouse | ClickHouse | 421,493,996 | 3,136 | alexey-milovidov | Привет. Оказывается, правильный вариант - `information`. | 2018-09-14T21:52:13 | 2018-09-14T21:52:13 | {} | MEMBER |
ClickHouse | ClickHouse | 421,625,835 | 3,142 | e11it | Проблема появляется при указании опции `--format_csv_delimiter `, без нее - работает.
Стандартный разделитель
```bash
echo -e "1.2,2\n3,4" | clickhouse-local -q "CREATE TABLE table (a Float64, b Int64) ENGINE = File(CSV, stdin); SELECT a, b FROM table; DROP TABLE table"
1.2 2
3 4
```
Указали раз... | 2018-09-15T20:00:22 | 2018-09-15T20:00:22 | {} | NONE |
ClickHouse | ClickHouse | 421,965,544 | 3,148 | blinkov | Здесь разница не только в наличии веса, но и в использовании разных алгоритмов: просто quantile означает приближённый, а не точный. См. https://clickhouse.yandex/docs/ru/query_language/agg_functions/reference/#quantilelevelx | 2018-09-17T10:49:05 | 2018-09-17T10:49:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,077,655 | 1,742 | amosbird | > but every modern allocator has similar behaviour.
OK... I was still under the impression that we need explicitly specify
```
mallopt(M_MMAP_MAX,0);
mallopt(M_TRIM_THRESHOLD,-1);
```
| 2018-09-17T16:14:27 | 2018-09-17T16:14:27 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 422,299,823 | 3,157 | BraisedFish | > What linux/unix distribution do you use?
> Can you show your timezone settings? Those commands can be helpful:
>
> ```shell
> $ ls -la /etc/localtime
> $ strings /etc/localtime
> $ echo $TZ
> $ echo $TZDIR
> $ date
> $ timedatectl
> ```
I defined TZ and TZDIR, and it' s ok now. thanks! | 2018-09-18T08:16:48 | 2018-09-18T08:16:48 | {} | NONE |
ClickHouse | ClickHouse | 421,783,247 | 3,144 | veloman-yunkan | This time the build itself timed out. | 2018-09-16T15:31:33 | 2018-09-16T15:31:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,889,904 | 699 | stiv-yakovenko | Please add jdbc connection property, because this problem provokes writing guly code... | 2018-09-17T04:50:20 | 2018-09-17T04:50:20 | {} | NONE |
ClickHouse | ClickHouse | 421,941,553 | 1,784 | blinkov | ReplacingMergeTree is mostly intended to save disk storage in long term, you still need to write queries assuming there will be duplicates in raw data. | 2018-09-17T09:24:52 | 2018-09-17T09:24:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,964,041 | 1,742 | amosbird | Yeah, it's a stale issue. I wasn't aware the tcmalloc behavior at that moment. | 2018-09-17T10:44:35 | 2018-09-17T10:44:35 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 422,122,820 | 1,197 | fleapapa | No. Sorry, i haven't had chance to revisit the issue since. Feel free to close this issue if it has been fixed or never existed. Thanks! | 2018-09-17T18:33:24 | 2018-09-17T18:33:24 | {} | NONE |
ClickHouse | ClickHouse | 422,257,841 | 3,144 | veloman-yunkan | @alexey-milovidov Thank you for accepting my first contribution!
I like your idea from the use model point of view, but it seems to me that implementing it would require more code (though, I can definitely be wrong). | 2018-09-18T05:15:04 | 2018-09-18T05:15:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,301,351 | 3,157 | filimonov | > Time zone: n/a (CST, +0800)
Hm... It look's like your linux distribution itself is not sure which timezone is in use :)
So clickhouse also can't determine proper timezone - may be it's caused by using /etc/localtime from another version of tzdata database.
1) What a distro do you use?
2) Could you also show... | 2018-09-18T08:21:04 | 2018-09-18T08:21:44 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,764,954 | 3,158 | sdlcwangsong | thank you very much | 2018-09-19T11:26:37 | 2018-09-19T11:26:37 | {} | NONE |
ClickHouse | ClickHouse | 421,464,509 | 3,139 | alexey-milovidov | API usage looks correct.
As I understand, this aggregate function takes leftmost and rightmost point (x, y) and returns a "slope" between them. This is all right. The only question is how useful it would be and whether "rate" is an appropriate name for it.
Examples:
When user wants to estimate "slope", "rate o... | 2018-09-14T19:42:49 | 2018-09-14T19:42:49 | {} | MEMBER |
ClickHouse | ClickHouse | 421,881,853 | 3,145 | den-crane | I did not mean that something wrong with the last version.
I mean it works and always worked incompatible with other SQL databases.
psql
create table t(a int, b int); insert into t values (1,1); insert into t values (2,2);
create table s(a int, b int);
insert into s values (1,1);
```
select * from t left out... | 2018-09-17T03:10:27 | 2018-09-17T03:10:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,910,910 | 2,862 | blinkov | @HamoonDBA I just wanted to let you know that we've configured the official ClickHouse website to show the results of your work: https://clickhouse.yandex/docs/fa/
At the moment it's kind of hard to reach from main page, but we're going to fix that too at some point.
Please let us know if you need any help. | 2018-09-17T07:13:41 | 2018-09-17T07:13:41 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,079,856 | 3,134 | zhang2014 | Can you give me some information? My development environment is working fine.
 | 2018-09-17T16:21:22 | 2018-09-17T16:21:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,158,447 | 3,144 | alexey-milovidov | Looks (almost) Ok. | 2018-09-17T20:26:05 | 2018-09-17T20:26:05 | {} | MEMBER |
ClickHouse | ClickHouse | 421,121,589 | 3,107 | alexey-milovidov | :+1: | 2018-09-13T19:19:16 | 2018-09-13T19:19:16 | {} | MEMBER |
ClickHouse | ClickHouse | 421,256,912 | 1,522 | thanhthai3010 | Yep, I'm also got this issue. Is there anybody can fix that or have information about that? | 2018-09-14T07:26:05 | 2018-09-14T07:26:05 | {} | NONE |
ClickHouse | ClickHouse | 421,442,610 | 3,126 | alexey-milovidov | Ok. But test is missing. | 2018-09-14T18:17:55 | 2018-09-14T18:17:55 | {} | MEMBER |
ClickHouse | ClickHouse | 422,017,063 | 3,134 | alexey-milovidov | Integration `test_merge_table_over_distributed` is broken. | 2018-09-17T13:31:33 | 2018-09-17T13:31:33 | {} | MEMBER |
ClickHouse | ClickHouse | 422,102,772 | 2,770 | alex-krash | @alexey-milovidov , here are results after optimization:
Reference:
```
SELECT count()
FROM
(
SELECT replaceRegexpOne(toString(t), '^(\\d\\d\\d\\d)-(\\d\\d)-(\\d\\d)( \\d\\d:\\d\\d:\\d\\d)$', '\\3/\\2/\\1\\4')
FROM
(
SELECT toDateTime('2018-01-01 00:00:00') + number AS t
FROM num... | 2018-09-17T17:30:49 | 2018-09-17T17:30:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,627,532 | 3,158 | sdlcwangsong | @alexey-milovidov , after update version, I got this error:
is my sql wrong?
thanks! | 2018-09-19T02:33:04 | 2018-09-19T02:33:04 | {} | NONE |
ClickHouse | ClickHouse | 421,266,943 | 3,112 | dsfishchen | Yes,i can connect to host 192.168.1.24 port 8123 from my application.
but sometimes it show "connect timed out" ,and i can connect after some minutes. | 2018-09-14T08:07:53 | 2018-09-14T08:07:53 | {} | NONE |
ClickHouse | ClickHouse | 421,318,670 | 3,115 | sundarv85 | Awesome that works | 2018-09-14T10:47:54 | 2018-09-14T10:47:54 | {} | NONE |
ClickHouse | ClickHouse | 421,775,843 | 3,144 | veloman-yunkan | A note on the failed Travis CI build - a couple of changed header files forced almost full recompilation and the Travis CI job was interrupted due to exceeded maximum time limit. At the time of the job termination the build was completed and most of the tests have been run. Strangely enough the only failed test is the ... | 2018-09-16T14:41:05 | 2018-09-16T14:41:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,936,513 | 1,808 | blinkov | @dqminh can you provide an instruction on how to reproduce this? | 2018-09-17T09:06:33 | 2018-09-17T09:06:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,093,383 | 2,935 | amosbird | Can we have this merged? | 2018-09-17T17:02:23 | 2018-09-17T17:02:23 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 422,144,989 | 3,134 | ztlpn | @zhang2014 Are you sure you are testing with the correct version? | 2018-09-17T19:44:16 | 2018-09-17T19:44:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,401,175 | 3,160 | alexey-milovidov | Thank you for the fix! | 2018-09-18T13:50:52 | 2018-09-18T13:50:52 | {} | MEMBER |
ClickHouse | ClickHouse | 422,401,712 | 3,159 | alexey-milovidov | Two tests have failed. | 2018-09-18T13:52:26 | 2018-09-18T13:52:26 | {} | MEMBER |
ClickHouse | ClickHouse | 422,466,974 | 2,770 | alexey-milovidov | I have started to merge. | 2018-09-18T16:48:54 | 2018-09-18T16:48:54 | {} | MEMBER |
ClickHouse | ClickHouse | 422,018,234 | 3,123 | alexey-milovidov | Integration test is broken.
`FAIL test_dictionaries/test.py::test_select_all[clickhouse_range_hashed]` | 2018-09-17T13:34:35 | 2018-09-17T13:34:35 | {} | MEMBER |
ClickHouse | ClickHouse | 422,085,159 | 1,742 | amosbird | Thanks! | 2018-09-17T16:37:57 | 2018-09-17T16:37:57 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 422,128,304 | 3,150 | alexey-milovidov | :+1: | 2018-09-17T18:50:03 | 2018-09-17T18:50:03 | {} | MEMBER |
ClickHouse | ClickHouse | 421,203,928 | 3,110 | zzl-wy | I used a old version 1.1.54198 and I get same output as alex-zaitsev on a newer version 1.1.54380 | 2018-09-14T01:41:58 | 2018-09-14T01:41:58 | {} | NONE |
ClickHouse | ClickHouse | 421,969,819 | 1,606 | blinkov | @Slind14 NVMe and SSD are not as fundamentally different as HDD and SSD. If you decided that HDD is not enough for your use case, you should start looking for specific disk drive model that provides best tradeoffs, considering not only raw performance and capacity, but also costs normalized by expected lifetime.
We... | 2018-09-17T11:04:03 | 2018-09-17T11:05:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,163,900 | 3,144 | alexey-milovidov | Thank you!
FYI. I also was thinking about another possible way for implementation.
We have named tuples. Nested column may be represented as a single column of type `Array(Tuple(a UInt8, b String))`. Then we can serialize/deserialize named tuples in natural way in JSON, like `{"a":1,"b":"Hello, world"}`. And `Arr... | 2018-09-17T20:42:52 | 2018-09-17T20:43:23 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 422,218,444 | 3,134 | zhang2014 | Next I'll try to recreate the problem in the same environment as the integration test | 2018-09-18T00:58:52 | 2018-09-18T00:58:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,575,709 | 3,167 | derekperkins | The second, it'd have to do a full table scan without `a` and `b` in the where clause. Even if you have all three, it may still do a full scan if it thinks that would be faster or your index is too sparse, but without all three, it has no way to optimize. The same is true for any non-indexed queries in any dbms. | 2018-09-18T22:15:18 | 2018-09-18T22:15:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,218,600 | 3,129 | alexey-milovidov | ```
../dbms/src/AggregateFunctions/AggregateFunctionStatisticsSimple.h:347:13: error: enumeration value 'covarPop' not handled in switch [-Werror=switch]
switch (StatFunc::kind)
``` | 2018-09-14T03:18:11 | 2018-09-14T03:18:11 | {} | MEMBER |
ClickHouse | ClickHouse | 421,878,975 | 3,145 | zhang2014 | Everything except `select t.*, s.* from t all left join s on (s.a = t.a and s.b = t.b) ` is expected.
You can refer to Backward incompatible changes of 18.12.13 [CHANGELOG](https://github.com/yandex/ClickHouse/blob/master/CHANGELOG.md) | 2018-09-17T02:43:36 | 2018-09-17T02:43:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,883,568 | 3,145 | zhang2014 | OK，Two questions, and I'll try to fix that : )
- Unexpected : `SELECT t.*, s.* FROM t ALL LEFT JOIN s ON (s.a = t.a and s.b = t.b)` .
- Remove duplicate join key when use `USING`. | 2018-09-17T03:31:30 | 2018-09-17T03:31:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,209,837 | 699 | den-crane | > Please add jdbc connection property, because this problem provokes writing guly code...
jdbc has these properties
https://github.com/yandex/clickhouse-jdbc/issues/202 | 2018-09-18T00:03:34 | 2018-09-18T00:04:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,579,187 | 2,818 | alexey-milovidov | > Are timezones with non-standard shifts are supported, e.g. 15, 30, 45 minutes?
Yes.
```
SELECT formatDateTime(parseDateTimeBestEffort('2000-01-01T00:00:00Z'), '%m/%d/%Y %H:%M:%S', 'Asia/Calcutta') AS res
┌─res─────────────────┐
│ 01/01/2000 05:30:00 │
└─────────────────────┘
1 rows in set. Elapsed: 0.0... | 2018-09-18T22:28:58 | 2018-09-18T22:28:58 | {} | MEMBER |
ClickHouse | ClickHouse | 422,758,392 | 3,158 | ztlpn | Thanks! Turns out I forgot to enable ON CLUSTER for ALTER UPDATE queries - fixed in badefe14bcc6c55c0e4ed9140379fb506f164c7b
ALTER DELETE should work though. | 2018-09-19T11:03:17 | 2018-09-19T11:03:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,328,390 | 873 | xiaods | need this feature , please asap. +1 | 2018-09-14T11:18:42 | 2018-09-14T11:18:42 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 422,010,585 | 1,475 | blinkov | I've tried to reproduce this on recent ClickHouse release and the server works fine. If you know how to do it, please let us know.
> BTW, can you provide a FAQ list in official documents?
We've just recently created it and it's practically empty at the moment https://clickhouse.yandex/docs/en/faq/general/ , but i... | 2018-09-17T13:14:17 | 2018-09-17T13:14:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,288,398 | 3,157 | filimonov | What linux/unix distribution do you use?
Can you show your timezone settings? Those commands can be helpful:
```bash
$ ls -la /etc/localtime
$ strings /etc/localtime
$ echo $TZ
$ echo $TZDIR
$ date
$ timedatectl
``` | 2018-09-18T07:40:02 | 2018-09-18T07:40:02 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,341,408 | 3,158 | ztlpn | Because a Distributed table is just a view over local tables that contain the data. You can use [distributed DDLs](https://clickhouse.yandex/docs/en/query_language/create/#distributed-ddl-queries-on-cluster-clause) to update/delete data on the whole cluster. | 2018-09-18T10:30:35 | 2018-09-18T10:30:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,385,556 | 1,736 | blinkov | @shangshujie365 Sounds perfectly normal: at point 2 the data from table A is in page cache, while point 3 forced it to become evicted. Have you tried repeating point 5 multiple times? Does performance become better again? | 2018-09-14T14:57:47 | 2018-09-14T14:57:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,011,006 | 3,044 | filimonov | Please check https://github.com/yandex/ClickHouse/pull/3104
`allow_ddl` added in v18.12.14
May be it's worth to mention that setting in documentation. | 2018-09-17T13:15:28 | 2018-09-17T13:18:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,072,872 | 1,742 | alexey-milovidov | We had switched to jemalloc, but every modern allocator has similar behaviour. | 2018-09-17T15:59:18 | 2018-09-17T15:59:18 | {} | MEMBER |
ClickHouse | ClickHouse | 422,143,279 | 2,935 | alexey-milovidov | No. The issue exists when `/usr/include/zconf.h` (some specific version of that file) from the OS is found and used instead of the same file from zlib-ng. I tried to workaround this issue with no success. | 2018-09-17T19:38:46 | 2018-09-17T19:38:46 | {} | MEMBER |
ClickHouse | ClickHouse | 422,188,347 | 3,135 | 4ertus2 | Исправление смержено в мастер. | 2018-09-17T22:09:50 | 2018-09-17T22:09:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,229,016 | 2,865 | qinzl1 | the new version can not install in centos6? or how can i avoid this conflicts Error | 2018-09-18T02:05:32 | 2018-09-18T02:05:32 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 422,290,099 | 3,139 | filimonov | Also you can use combinations of max / min / argMax / argMin, like that:
```sql
SELECT
sensor_id,
max(time) - min(time) as time_difference_in_sec,
argMax(battery_level,time) - argMin(battery_level,time) as battery_level_change,
battery_level_change / time_difference_in_sec as rate
FROM table
WHERE .... | 2018-09-18T07:45:26 | 2018-09-18T07:45:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,297,162 | 1,614 | blinkov | @zhanglistar basically HDFS is no different than any other place you could import data from, there are several interfaces provided by ClickHouse with different tradeoffs that you can choose from, for example:
* JDBC: https://clickhouse.yandex/docs/en/interfaces/jdbc/
* CLI: https://clickhouse.yandex/docs/en/interfa... | 2018-09-18T08:08:26 | 2018-09-18T08:08:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,406,339 | 208 | filimonov | Now when there is an ability to collect logs from server, that idea of "monitor"ing query from outside / with another query sounds even better.
I.e. plain HTTP doesn't support any multiplexing, so instead of creating special headers for everything, just create ability to see the details of that's going on on anothe... | 2018-09-18T14:04:40 | 2018-09-18T14:04:40 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,578,106 | 2,818 | alexey-milovidov | > Is merged implementation faster than using makeTM + strftime? By how much?
Yes.
Query 1:
`select count() from (select formatDateTime(t, '%m/%d/%Y %H:%M:%S') from (select toDateTime('2018-01-01 00:00:00')+number as t from numbers(100000000)));`
Your branch: 5.72 million rows/s.
Master: 15.33 million rows/s.
... | 2018-09-18T22:24:26 | 2018-09-18T22:25:24 | {} | MEMBER |
ClickHouse | ClickHouse | 421,290,915 | 3,107 | zhang2014 | Can we push down predicate to `WITH ... SELECT` ? Because subqueries is normalized. | 2018-09-14T09:25:34 | 2018-09-14T09:26:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,471,127 | 3,139 | tsl-karlp | Thank you, @alexey-milovidov . I agree, the name isn't right. Does it have to be an aggregate function to use with a `GROUP BY sensor_id`, for example?. We have IoT use case where we have temperature readings, for example:
| sensor_id | time | battery_level |
| ------------- | ------------- | ------------- |... | 2018-09-14T20:09:42 | 2018-09-14T20:09:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,936,734 | 1,197 | blinkov | @fleapapa do you have any further questions? | 2018-09-17T09:07:23 | 2018-09-17T09:07:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,012,625 | 3,044 | blinkov | @BayoNet could you add allow_ddl from #3104 to your todo list? | 2018-09-17T13:19:39 | 2018-09-17T13:19:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,389,005 | 3,134 | zhang2014 | @ztlpn @alexey-milovidov fixed in #3159 | 2018-09-18T13:16:38 | 2018-09-18T13:16:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,561,170 | 2,818 | bocharov | > Implemented in #2770 by @alex-krash.
>
> Now I'm going to import your tests and docs...
> I will try to preserve your authorship in commits.
Thanks @alexey-milovidov
I have the following questions:
1. Is merged implementation faster than using `makeTM` + `strftime`? By how much?
2. Are timezones with non-s... | 2018-09-18T21:25:09 | 2018-09-18T21:25:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,778,152 | 3,167 | simPod | I did some testing with following table
```sql
CREATE TABLE default.a_table (
fetched_at DateTime,
id1 UUID,
id2 UUID,
fetched_at_date Date MATERIALIZED toDate(fetched_at)
)
ENGINE = MergeTree
PARTITION BY fetched_at_date ORDER BY (fetc... | 2018-09-19T12:10:05 | 2018-09-19T12:10:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,384,776 | 3,112 | den-crane | blast from the past https://github.com/yandex/clickhouse-jdbc/issues/47
it was fixed two years ago
CH 1.1.54023 / jdbc(version 0.1.13) -- Nov 5, 2016
you need to change jdbc timeout to 10sec. or rather update to the latest jdbc driver ( release_0.1.41 ) https://mvnrepository.com/artifact/ru.yandex.clickh... | 2018-09-14T14:55:29 | 2018-09-14T14:57:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,967,680 | 3,148 | chipitsine | забавный момент. буду иметь в виду
| 2018-09-17T10:56:34 | 2018-09-17T10:56:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,080,225 | 1,742 | alexey-milovidov | AFAIK, these parameters have no sense for tcmalloc and jemalloc.
And default Linux allocator has terrible performance (don't use it even if you will get acceptable results in simple synthetic tests). | 2018-09-17T16:22:32 | 2018-09-17T16:22:32 | {} | MEMBER |
ClickHouse | ClickHouse | 422,096,313 | 3,135 | 4ertus2 | #3153 | 2018-09-17T17:11:54 | 2018-09-17T17:11:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,127,530 | 3,154 | alexey-milovidov | `00646_url_engine: [ FAIL ] - return code 1` - flappy test. | 2018-09-17T18:47:43 | 2018-09-17T18:47:43 | {} | MEMBER |
ClickHouse | ClickHouse | 422,144,875 | 3,134 | ztlpn | Current master:
```=========================================================================================== test session starts ============================================================================================
platform linux2 -- Python 2.7.12, pytest-3.7.4, py-1.6.0, pluggy-0.7.1 -- /usr/bin/python
cac... | 2018-09-17T19:43:56 | 2018-09-17T19:43:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,260,757 | 3,134 | zhang2014 | @ztlpn Thanks:)
I successfully repeated the problem in the development environment. I will fix the problem as soon as possible. | 2018-09-18T05:33:07 | 2018-09-18T05:33:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,218,843 | 3,110 | zhang2014 | The reason is that there are problems with the `query stage` coordination between the different engines. I'm trying to fix this. | 2018-09-14T03:19:40 | 2018-09-14T03:19:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,948,470 | 1,761 | blinkov | It has been supported this summer in #2779 | 2018-09-17T09:49:07 | 2018-09-17T09:49:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,960,109 | 1,742 | blinkov | @amosbird why do you think that it's a memory leak? I tried to reproduce this and it looks like the memory is correctly reused by allocator if you create and drop tables like this multiple times. It's just not immediately returned to OS. | 2018-09-17T10:31:17 | 2018-09-17T10:31:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,067,632 | 3,145 | den-crane | One more test.
There is a very useful use-case find all rows in T which are "not-exists" in S using semi-join.
mysql
```
create table t(a int, b int, c varchar(100)); insert into t values (1,1,'a'); insert into t values (2,2,'b');
create table s(a int, b int, c varchar(100)); insert into s values (1,1,'a');
... | 2018-09-17T15:44:00 | 2018-09-17T15:44:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,289,799 | 3,157 | BraisedFish | > What linux/unix distribution do you use?
> Can you show your timezone settings? Those commands can be helpful:
>
> ```shell
> $ ls -la /etc/localtime
> $ strings /etc/localtime
> $ echo $TZ
> $ echo $TZDIR
> $ date
> $ timedatectl
> ```
root@vm24177 packege]# ls -la /etc/localtime
-rw-r--r--. 1 root ... | 2018-09-18T07:44:27 | 2018-09-18T07:44:27 | {} | NONE |
ClickHouse | ClickHouse | 422,402,455 | 3,158 | alexey-milovidov | The message `ALTER queries are not yet supported for replicated tables` indicate that you are using old ClickHouse version (at least on one of your cluster nodes). | 2018-09-18T13:54:38 | 2018-09-18T13:54:38 | {} | MEMBER |
ClickHouse | ClickHouse | 422,633,303 | 2,818 | bocharov | That's great, thank you. | 2018-09-19T03:02:46 | 2018-09-19T03:02:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,555,651 | 3,141 | pavel-odintsov | I did rollback to 18.10.3 and it works fine.
Also, I noticed that you keep only very short list of previous version in apt repo and it makes rollback quite complicated:
```
apt-cache madison clickhouse-server
clickhouse-server | 18.12.14 | http://repo.yandex.ru/clickhouse/deb/stable/ main/ Packages
clickhouse-... | 2018-09-15T11:38:59 | 2018-09-15T11:38:59 | {} | NONE |
ClickHouse | ClickHouse | 421,830,115 | 3,145 | den-crane | ```
create table t(a Int64, b Int64) engine = TinyLog;
insert into t values(1,1);
insert into t values(2,2);
create table z(c Int64, d Int64, e Int64) engine = TinyLog;
insert into z values(1,1,1);
```
Wrong result
`select * from t all left join z on (z.c = t.a and z.d = t.b)`
a | b | c | d | e
-- | -- | -... | 2018-09-16T20:43:58 | 2018-09-16T20:43:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,944,267 | 1,774 | blinkov | > My Question is the attach partition act can only be performed on one node if both nodes run 'attach partition xxx', the xxx data will be duplicated.
Hopefully you've already found the relevant of documentation by now:
> ALTER TABLE [db.]table ATTACH PARTITION|PART 'name'
>
> Adds data to the table from the 'de... | 2018-09-17T09:34:33 | 2018-09-17T09:34:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 421,952,106 | 3,112 | dsfishchen | Thanks all.
I change the file 'ClickHouseConnectionSettings' as follow,and it seems working fine.
CONNECTION_TIMEOUT("connection_timeout", 10000, ""),
TIME_TO_LIVE_MILLIS("timeToLiveMillis", 60000, ""), | 2018-09-17T10:01:46 | 2018-09-17T10:01:46 | {} | NONE |
ClickHouse | ClickHouse | 421,967,527 | 3,148 | chipitsine | поменял на quantilesExact, стало вот так
```
[0,0,0,0,0.001,0.002,0.004,0.021,0.14,0.246,267.865]
[0,0,0,0,0.001,0.002,0.004,0.021,0.14,0.246,267.865]
``` | 2018-09-17T10:55:56 | 2018-09-17T10:55:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,217,096 | 3,134 | zhang2014 | Yes. commit id : 13779bc75d85a3f52299c4df34a741208ebf4cf7
| 2018-09-18T00:50:18 | 2018-09-18T00:54:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,238,603 | 3,139 | sundy-li | I do think that [runningDifference](https://clickhouse.yandex/docs/en/query_language/functions/other_functions/#runningdifferencex) could meet your needs better than creating a new AggregateFunction rate.
Eg.
```
SELECT
number AS time,
number * number AS level,
runningDifference(level) / runnin... | 2018-09-18T03:04:37 | 2018-09-18T03:04:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,351,885 | 3,158 | sdlcwangsong | @ztlpn thank you ,but my table engine is ReplicatedMergeTree, and I saw " ALTER queries are not yet supported for replicated tables" | 2018-09-18T11:08:35 | 2018-09-18T11:08:35 | {} | NONE |
ClickHouse | ClickHouse | 422,557,696 | 2,818 | alexey-milovidov | Implemented in #2770 by @alex-krash.
Now I'm going to import your tests and docs...
I will try to preserve your authorship in commits. | 2018-09-18T21:14:29 | 2018-09-18T21:15:18 | {} | MEMBER |
ClickHouse | ClickHouse | 421,495,544 | 3,136 | e11it | На всякий, обращаю внимание, что он не только давал выставить не поддерживаемый уровень логирования, но и не давал поменять на другой, т.к. уже установленный - не валидный.
Единственный выход был, отключить и заново подключиться. | 2018-09-14T22:00:13 | 2018-09-14T22:00:13 | {} | NONE |
ClickHouse | ClickHouse | 421,889,227 | 3,144 | alexey-milovidov | > This time the build itself timed out.
Don't worry, Travis build barely works. | 2018-09-17T04:42:28 | 2018-09-17T04:42:28 | {} | MEMBER |
ClickHouse | ClickHouse | 421,973,575 | 877 | blinkov | I'm not sure where link by @vavrusa pointed at that time, but max_replica_delay_for_distributed_queries is documented nowadays. | 2018-09-17T11:17:37 | 2018-09-17T11:17:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,116,861 | 3,123 | alexey-milovidov | #3154 | 2018-09-17T18:14:12 | 2018-09-17T18:14:12 | {} | MEMBER |
ClickHouse | ClickHouse | 422,773,838 | 2,827 | wake-up-neo | @alexey-milovidov Please take a look
```
drop table if exists test.trepl;
create table test.trepl
(
d Date,
a Int32,
b Int32
) engine = ReplacingMergeTree(d, (a,b), 8192);
insert into test.trepl values ('2018-09-19', 1, 1);
select b from test.trepl FINAL prewhere a < 1000;
drop table test.trepl;
```
... | 2018-09-19T11:56:16 | 2018-09-19T11:56:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 422,780,359 | 2,827 | wake-up-neo | The same for CollapsingMergeTree, and only with FINAL.
> 2018.09.19 15:15:56.804624 [ 54 ] {f446cb2b-b4af-4fad-a082-5b2bac65e6cf} <Error> executeQuery: Code: 10, e.displayText() = DB::Exception: Not found column a in block. There are only columns: b, sign, e.what() = DB::Exception (from 127.0.0.1:52210) (in query: s... | 2018-09-19T12:17:30 | 2018-09-19T12:20:56 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.