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
372,361,450
2,026
silviucpp
Of course using CAST will fix the problem but maybe clickhouse needs to be smart enough to allow this by default
2018-03-12T15:57:45
2018-03-12T15:57:45
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,426,918
2,021
alexey-milovidov
We don't have it in our "short-list" roadmap and it is not scheduled for implementation right now, but I'm considering this feature for implementation and started reviewing possible approaches. It will most likely to be implemented as 128 or 64 bit integer as a backend (depending on scale + precision). The implemen...
2018-03-12T19:03:26
2018-03-12T19:03:51
{ "+1": 3 }
MEMBER
ClickHouse
ClickHouse
372,557,288
936
dolbyzerr
@alexey-milovidov I have an error when I try to alter distributed table that points to `ReplicatedMergeTree` on cluster ``` CREATE TABLE log.test ON CLUSTER 'my-cluster' (date Date, value1 String) ENGINE = ReplicatedMergeTree ('/clickhouse/tables/{shard}/log/test','{replica}', date, (date), 8192) CREATE TABLE ...
2018-03-13T06:08:44
2018-03-13T06:09:41
{}
NONE
ClickHouse
ClickHouse
373,463,527
2,063
simplevil
Thanks a lot.
2018-03-15T17:42:59
2018-03-15T17:42:59
{}
NONE
ClickHouse
ClickHouse
370,863,671
2,001
kmatt
``` 2018.03.06 17:27:52.768526 [ 20 ] <Error> ServerErrorHandler: Code: 210, e.displayText() = DB::NetException: Connection reset by peer: while reading from socket ([::ffff:xxx.xxx.xxx.xxx]:50194), e.what() = DB::NetException, Stack trace: 0. clickhouse-server(StackTrace::StackTrace()+0x16) [0x38bcde6] 1. clickho...
2018-03-06T17:36:28
2018-03-06T17:36:40
{}
NONE
ClickHouse
ClickHouse
371,946,150
245
alexey-milovidov
> Now there are two queries with the same query_id! That's Ok, because query cancellation is done in background and query is still shows in processlist while being cancelled before actually finish.
2018-03-09T21:11:41
2018-03-09T21:11:55
{}
MEMBER
ClickHouse
ClickHouse
372,368,090
68
adokukin
@alexey-milovidov > To get unix timestamp for string with datetime in specified time zone: > toUnixTimestamp('2000-01-01 00:00:00', 'Asia/Yekaterinburg') Which version of clickhouse-sever does that? I just upgraded clickhouse-server-common to version 1.1.54362 and server responds with "DB::Exception: Number of ...
2018-03-12T16:16:17
2018-03-12T16:18:21
{}
NONE
ClickHouse
ClickHouse
372,621,252
1,633
silviucpp
@alexey-milovidov don't forget about this
2018-03-13T10:38:57
2018-03-13T10:38:57
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,895,053
2,012
alexey-milovidov
Missing performance test. Missing algorithm description inside the code.
2018-03-14T03:42:12
2018-03-14T03:42:12
{}
MEMBER
ClickHouse
ClickHouse
373,134,465
2,034
shahvishal8464
# Create the schema Replace cluster name if needed ``` DROP TABLE IF EXISTS PriceTable_TBL ; DROP TABLE IF EXISTS PriceTable ; DROP TABLE IF EXISTS SalesTable_TBL ; DROP TABLE IF EXISTS SalesTable ; DROP TABLE IF EXISTS ProductDscntTable_TBL ; DROP TABLE IF EXISTS ProductDscntTable ; CREATE TABLE P...
2018-03-14T18:48:32
2018-03-14T18:48:32
{}
NONE
ClickHouse
ClickHouse
373,443,424
2,060
peteychuk
Fixed. They've changed the url to: deb http://repo.yandex.ru/clickhouse/deb/stable/ main/ Not it works. Thanks @rpfennin!
2018-03-15T16:43:33
2018-03-15T16:43:33
{ "+1": 4 }
NONE
ClickHouse
ClickHouse
370,775,726
1,912
zhang2014
@alexey-milovidov I've added test, Please help review the changes : )
2018-03-06T13:08:38
2018-03-06T13:08:38
{}
CONTRIBUTOR
ClickHouse
ClickHouse
371,975,800
245
alexey-milovidov
Fixed in master.
2018-03-09T23:36:38
2018-03-09T23:36:38
{}
MEMBER
ClickHouse
ClickHouse
371,985,243
1,813
alexey-milovidov
Duplicate of #245. Fixed in master.
2018-03-10T00:42:35
2018-03-10T00:42:35
{}
MEMBER
ClickHouse
ClickHouse
372,083,972
2,017
alexey-milovidov
:+1:
2018-03-11T02:23:21
2018-03-11T02:23:21
{}
MEMBER
ClickHouse
ClickHouse
372,418,226
1,722
alexey-milovidov
I'm interested in this modification. For example, if you run quite loaded server under ASan, it will complain about "thread limit exceeded" after few days of run. And your modification will fix it. But it's rather hard to implement it without breaking something. And @ztlpn had a few complains about this modificat...
2018-03-12T18:36:42
2018-03-12T18:36:42
{}
MEMBER
ClickHouse
ClickHouse
372,432,790
1,722
silviucpp
ok great ! I'll wait for your feedback guys !
2018-03-12T19:23:12
2018-03-12T19:23:12
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,877,512
2,037
alexey-milovidov
:+1:
2018-03-14T01:49:32
2018-03-14T01:49:45
{}
MEMBER
ClickHouse
ClickHouse
372,908,566
2,030
alexey-milovidov
> uniqExact can't use disk, all unique records should really fit in memory, with some overhead. uniqExact can use disk - as any other aggregate function with enabled `max_bytes_before_external_group_by` setting, if you have multiple small states for different aggregation keys.
2018-03-14T05:25:10
2018-03-14T05:25:10
{ "+1": 2 }
MEMBER
ClickHouse
ClickHouse
372,987,746
2,041
millin
Since v1.1.54289 you can use `SYSTEM RELOAD DICTIONARY <dict_name>` query or `SYSTEM RELOAD DICTIONARIES` query. Last dictionary update is stored in `system.dictionaries` table.
2018-03-14T11:21:22
2018-03-14T11:21:22
{ "+1": 5, "heart": 4 }
CONTRIBUTOR
ClickHouse
ClickHouse
373,104,463
2,045
filimonov
That can also lead to big downgrade of compression ratio (implementing compression before encryption will be not so easy) and together with slowing down data reading / inserting can kill all the advantages.
2018-03-14T17:19:49
2018-03-14T17:31:36
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,275,661
2,030
filimonov
When max_bytes_before_external_group_by is enabled the temporary state of each aggregate function (including uniqExact) can be flushed to disk. But internal state of each function should fit into memory. I.e. if you group by let's say minutes, and make a select for 1 year, you will have a lot of small groups. Ag...
2018-03-15T06:37:42
2018-03-15T06:37:42
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,286,249
1,722
silviucpp
are you guys still interested into this pull request ? Just asking because now conflicts with master and as time you don't care about it I will close it
2018-03-12T12:00:20
2018-03-12T12:00:20
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,444,773
2,026
alexey-milovidov
Yes, these operations should be supported.
2018-03-12T20:05:10
2018-03-12T20:05:10
{}
MEMBER
ClickHouse
ClickHouse
372,891,282
2,012
alexey-milovidov
Few points: > IntersectionsMax Currently we name our functions in `camelCase`, not `CamelCase`. Also I think that `maxIntersections` sounds better than `intersectionsMax`. > if start_column value is NULL NULL values are skipped before being passed to aggregate functions. > if start_column value is NULL ...
2018-03-14T03:17:03
2018-03-14T03:17:44
{}
MEMBER
ClickHouse
ClickHouse
372,892,514
2,012
alexey-milovidov
Missing documentation in code (`.h` file). Missing test.
2018-03-14T03:25:13
2018-03-14T03:25:13
{}
MEMBER
ClickHouse
ClickHouse
371,945,505
245
alexey-milovidov
Steps to reproduce: ``` milovidov@milovidov-Latitude-E7470:~/work/ClickHouse/$ curl 'http://localhost:8123/?query_id=hello&replace_running_query=1' -d 'SELECT sum(ignore(*)) FROM (SELECT number % 1000 AS k, groupArray(number) FROM numbers(100000000) GROUP BY k)' & [1] 10045 milovidov@milovidov-Latitude-E7470:~/wo...
2018-03-09T21:08:58
2018-03-09T21:08:58
{}
MEMBER
ClickHouse
ClickHouse
372,617,920
102
ghuname
The problem remains in v1.1.54362
2018-03-13T10:26:42
2018-03-13T10:26:42
{ "+1": 1 }
NONE
ClickHouse
ClickHouse
372,909,673
2,012
alexey-milovidov
I've slightly changed the algorithm. Please take a look.
2018-03-14T05:32:46
2018-03-14T05:32:46
{}
MEMBER
ClickHouse
ClickHouse
373,395,498
2,061
alniks
May relate to #1550
2018-03-15T14:30:23
2018-03-15T14:30:23
{}
NONE
ClickHouse
ClickHouse
372,058,617
1,902
alexey-milovidov
Получилась идиосинкразия для part_log. Исправлю самостоятельно.
2018-03-10T19:32:08
2018-03-10T19:32:08
{}
MEMBER
ClickHouse
ClickHouse
372,062,002
1,630
alexey-milovidov
We have CI for Mac OS builds. It is using gcc-7. Also we have tried Travis, but for Mac OS, there are very long queues (it is basically unusable).
2018-03-10T20:10:05
2018-03-10T20:10:05
{}
MEMBER
ClickHouse
ClickHouse
372,763,278
2,016
kshvakov
Because it's the wrong place to fix this problem. Problem with incorrect quotes is in the clickhouse-client side, the server returns correct string `'2018-02-15'`. Client uses `serializeTextEscaped` method and show that string as `\'2018-02-15\'`
2018-03-13T18:06:36
2018-03-13T18:06:36
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,831,529
2,016
filimonov
> Problem with incorrect quotes is in the clickhouse-client side Check that: https://github.com/yandex/ClickHouse/issues/1729 > the server returns correct string '2018-02-15' Here the server sends the string with quotes at the begining and the end. Normally strings does not need to have quotes around. I don't ...
2018-03-13T21:53:26
2018-03-13T21:54:14
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,893,362
2,012
alexey-milovidov
``` void insertResultInto(ConstAggregateDataPtr place, IColumn & to) const override { auto & ret = static_cast<ColumnUInt64 &>(to).getData(); ret.push_back(data(place).max()); if (return_position) ret.push_back(data(place).max_pos()); } ``` That's absolutely in...
2018-03-14T03:30:56
2018-03-14T03:30:56
{}
MEMBER
ClickHouse
ClickHouse
373,065,097
2,016
ztlpn
Yes, if the partition key is a tuple, we need to mimic the behavior of DataTypeTuple::serializeText() or just use it directly. And it uses serializeTextQuoted() for serializing the tuple elements. Also I think it is better to use serializeTextEscaped in the exception message to avoid messing with the client terminal...
2018-03-14T15:34:31
2018-03-14T15:34:31
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,435,345
2,044
ludv1x
Using of arrays as the second argument of IN operator is prohibited: https://clickhouse.yandex/docs/en/single/#in-operators So, the query has to fail for this reason. But an arguable question about the following queries remains: ``` with ('a') as w select 'a' IN w with tuple('a') as w select 'a' IN w ``` They ...
2018-03-15T16:19:56
2018-03-15T16:21:22
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,134,751
2,019
alexey-milovidov
Thanks!
2018-03-11T17:54:45
2018-03-11T17:54:45
{}
MEMBER
ClickHouse
ClickHouse
372,327,892
1,474
silviucpp
@vavrusa did you found any workaround for this ?
2018-03-12T14:26:33
2018-03-12T14:26:33
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,598,227
2,030
xiao1dian
how to use the external disk for some big queries?
2018-03-13T09:19:24
2018-03-13T09:19:24
{}
NONE
ClickHouse
ClickHouse
371,250,573
1,901
alexey-milovidov
Sorry for the delay. We will review today.
2018-03-07T19:12:23
2018-03-07T19:12:23
{}
MEMBER
ClickHouse
ClickHouse
372,331,002
1,630
silviucpp
@alexey-milovidov I guess you are not using High Sierra because there master seems not compiling with none of clang or gcc-7 . On gcc-7 there is the problem of : https://github.com/yandex/ClickHouse/issues/1474 and on clang : Seems in last stable Xcode optional is still under experimental namespace. (original isue...
2018-03-12T14:35:41
2018-03-12T14:35:41
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,662,625
2,009
ludv1x
> Returns 0. I expect to return -10.2 You have two NULLs and two numbers, NULLs are skipped, only two numbers -10.2 and 10.2 remains. ``` median(a, b) = (a + b) / 2 median(-10.2, 10.2) = 0 ``` Am I right?
2018-03-13T13:21:08
2018-03-13T13:21:08
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,394,988
2,060
peteychuk
the same issue
2018-03-15T14:28:53
2018-03-15T14:28:53
{}
NONE
ClickHouse
ClickHouse
371,348,205
1,901
alexey-milovidov
Your help is very appreciated!
2018-03-08T01:39:18
2018-03-08T01:39:18
{}
MEMBER
ClickHouse
ClickHouse
372,799,609
2,016
ztlpn
> Because it's the wrong place to fix this problem. Well the problem exists not only on the client side. Suppose you want to query the parts according to some complicated predicate on the partition key. With proper serialization you could do `CAST(partition AS <partition-key-type>)` and use the partition column in a...
2018-03-13T20:02:52
2018-03-13T20:02:52
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,040,715
2,039
ludv1x
Confirmed
2018-03-14T14:28:35
2018-03-14T16:15:59
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,372,266
2,055
filimonov
Hmm... Thought it's documented. Generally each request logged twice - at the moment of start and at the moment of finishing / exception. Check the 'type' column. ``` QUERY_START = 1, QUERY_FINISH = 2, EXCEPTION_BEFORE_START = 3, EXCEPTION_WHILE_PROCESSING = 4, ```
2018-03-15T13:17:37
2018-03-15T13:17:37
{ "+1": 1 }
CONTRIBUTOR
ClickHouse
ClickHouse
373,437,910
2,060
peteychuk
What service provider do you use? I'm trying to use DigitalOcean and no luck on different 14.04/16.04 droplets.
2018-03-15T16:27:39
2018-03-15T16:27:39
{}
NONE
ClickHouse
ClickHouse
373,487,229
2,044
TimonKK
Thank you, the operator "has" helped
2018-03-15T18:59:21
2018-03-15T18:59:21
{}
NONE
ClickHouse
ClickHouse
372,753,265
2,016
ztlpn
Why did you abandon the PR? I'm thinking of introducing similar changes. The original reason for quoting was to enable using values of the partition column in ALTER queries as-is. The downsides are: ugly quotes; and difficulties with casting values of the partition column to the original partition key type. Without the...
2018-03-13T17:37:23
2018-03-13T17:37:23
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,866,488
2,016
alexey-milovidov
> Check that: #1729 The reason for escaping in Pretty format is to don't break tables formatting with newlines and tabs. We can implement more smart formatting, like this: ``` ┌─x──────┬─y──┐ │ hello, ╲ a, ╲ │ world │ b, ╲ │ │ c │ │ xxx │ yy │ └────────┴────┘ ``` It looks like a bit overkill...
2018-03-14T00:40:57
2018-03-14T00:40:57
{}
MEMBER
ClickHouse
ClickHouse
372,891,028
2,030
xiao1dian
Thank you very much. make sense for me
2018-03-14T03:15:12
2018-03-14T03:15:12
{}
NONE
ClickHouse
ClickHouse
372,938,853
2,016
kshvakov
With this patch there are problems with partition based on tuple(date...string): ``` Parts before OPTIMIZE: Parts before OPTIMIZE: (\'2000-01-01\', 1) 20000101-1_1_1_0 | (2000-01-01, 1) 20000101-1_1_1_0 (\'2000-01-01\', 1) 20000101-1_5_5_0 | (2000-01-01, 1) 20000101-1_5_5_0 (\'2000-01-01\', 2...
2018-03-14T08:18:24
2018-03-14T08:19:18
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,254,381
2,030
xiao1dian
@alexey-milovidov really?I just did set the max_bytes_before_external_group_by to 20G, but still got Memory limit (for query) exceeded error, so seems it didn't use disk.
2018-03-15T04:06:17
2018-03-15T04:06:17
{}
NONE
ClickHouse
ClickHouse
370,964,933
2,003
alexey-milovidov
The fix is in version 1.1.54337. Look at the changelog: https://github.com/yandex/ClickHouse/blob/master/CHANGELOG.md#clickhouse-release-1154337-2018-01-18 > Fixed parsing of tuples (values of the Tuple data type) in text formats.
2018-03-06T23:20:48
2018-03-06T23:20:48
{}
MEMBER
ClickHouse
ClickHouse
371,930,742
2,013
hatarist
Oops, pardon me. The documentation [says](https://clickhouse.yandex/docs/en/functions/json_functions/): > The JSON doesn't have space characters outside of string literals. I'm closing this issue, then; though it would be nice to have such an improvement.
2018-03-09T20:05:40
2018-03-09T20:05:40
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,137,328
2,008
alexey-milovidov
It doesn't. It would be nice if it will have.
2018-03-11T18:28:41
2018-03-11T18:28:41
{ "+1": 1 }
MEMBER
ClickHouse
ClickHouse
372,453,012
2,014
alexey-milovidov
Ok. Please be careful with naming. As we have function `sum`, the weighted should be named `sumWeighted`, and as we have `varSamp`, the weighted should be named `varSampWeighted`.
2018-03-12T20:33:58
2018-03-12T20:33:58
{}
MEMBER
ClickHouse
ClickHouse
372,892,941
2,012
alexey-milovidov
The function `IntersectionsMaxPos` should be named `maxIntersectionsPosition` according the fact that we have `position` function, not `pos`.
2018-03-14T03:28:14
2018-03-14T03:28:14
{}
MEMBER
ClickHouse
ClickHouse
373,447,401
2,054
ludv1x
You can freely ignore user-level (INFO) error in ZooKeeper logs if there no similar errors in ClickHouse logs. We have the same errors in our ZooKeeper installation and everything is Ok.
2018-03-15T16:54:42
2018-03-15T16:54:42
{}
CONTRIBUTOR
ClickHouse
ClickHouse
371,115,558
1,469
inkrement
@ztlpn is it somehow possible to use runningDifference in combination with a group by? For example to calculate some lag-value for each unit.
2018-03-07T11:54:59
2018-03-07T11:54:59
{}
NONE
ClickHouse
ClickHouse
371,264,380
1,901
alexey-milovidov
I wonder what is the main motivation of this feature (what is your usage scenario, where this is most important)?
2018-03-07T19:57:06
2018-03-07T19:57:06
{}
MEMBER
ClickHouse
ClickHouse
371,985,037
1,730
alexey-milovidov
Duplicate of #245. Fixed in master.
2018-03-10T00:41:19
2018-03-10T00:41:19
{ "+1": 1 }
MEMBER
ClickHouse
ClickHouse
372,445,776
2,014
bubaflub
Great. I'm thinking about implementing: - CountWeighted - SumWeighted - AvgWeighted - varSampleWeighted - varPopWeighted - stddevSampleWeighted - stddevPopWeighted I can add tests and documentation in English as well.
2018-03-12T20:08:37
2018-03-12T20:08:37
{}
NONE
ClickHouse
ClickHouse
372,876,528
2,035
alexey-milovidov
Nothing works for unknown reason.
2018-03-14T01:43:37
2018-03-14T01:43:37
{}
MEMBER
ClickHouse
ClickHouse
372,893,173
2,012
alexey-milovidov
The user could expect that the function will work for floats and signed numbers.
2018-03-14T03:29:26
2018-03-14T03:29:26
{}
MEMBER
ClickHouse
ClickHouse
373,462,083
2,063
alex-zaitsev
Performance is bad because of index granularity - 10000 in your example. For point queries try to reduce parameter to something small, like 128: ENGINE=CollapsingMergeTree(created_at, (txid), 128, sign)
2018-03-15T17:38:32
2018-03-15T17:38:32
{ "+1": 1 }
CONTRIBUTOR
ClickHouse
ClickHouse
371,105,722
1,697
millin
I checked this issue on latest stable v1.1.54342 - all works fine. Probably there are some troubles in previous versions (v1.1.54318, v1.1.54327) or something strange happened during building to rpm. Closed.
2018-03-07T11:11:46
2018-03-07T11:11:46
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,432,214
1,474
silviucpp
Yes the above line make the trick !
2018-03-12T19:21:15
2018-03-12T19:21:15
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,638,305
2,030
filimonov
uniqExact can't use disk, all unique records should really fit in memory, with some overhead. As a workarond you can use: 1) uniq() function which gives a good approximation by using fixed amount of memory 2) add more memory 3) do group by and later count row of resulting set (group by can use disk, with the opt...
2018-03-13T11:47:08
2018-03-13T11:48:11
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,658,915
2,020
ludv1x
It is fixed now
2018-03-13T13:08:08
2018-03-13T13:08:08
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,382,019
2,058
lamberken
it seems that the client version is incompatible with the server version. I fix it myselft, thank you :smile:
2018-03-15T13:50:33
2018-03-15T13:50:33
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,397,971
1,323
silviucpp
@alexey-milovidov any input on this issue. I opened it for a while ? should be considered a bug ?
2018-03-15T14:37:28
2018-03-15T14:37:28
{}
CONTRIBUTOR
ClickHouse
ClickHouse
371,130,046
1,697
millin
I just faced with this or related issue again. During the test for more different valid ip-addresses, using query `SELECT dictGetString('geolite2_city', 'iso_code', tuple('8.8.8.8'));` I found that sometimes this query hangs and in this case tee log is empty. I tried to kill hang query from CH - unsuccessfully...
2018-03-07T12:57:04
2018-03-07T12:57:04
{}
CONTRIBUTOR
ClickHouse
ClickHouse
371,348,033
2,007
alexey-milovidov
:+1:
2018-03-08T01:38:20
2018-03-08T01:38:20
{}
MEMBER
ClickHouse
ClickHouse
371,764,042
2,010
silviucpp
This function returns if I'm not wrong unique elements across all arrays. not unique arrays for example if I have [1,2], [3,4], [5,6] I expect 3 but returns 6 (1,2,3,4,5,6) Silviu On Thu, Mar 8, 2018 at 11:10 AM, 张健 <notifications@github.com> wrote: > Maybe try arrayUniq Function ? > > — > You are receiving this bec...
2018-03-09T09:44:19
2018-03-09T09:44:19
{}
CONTRIBUTOR
ClickHouse
ClickHouse
371,969,217
2,014
alexey-milovidov
I think, it is good idea! Looking forward for PR.
2018-03-09T22:57:24
2018-03-09T22:57:24
{}
MEMBER
ClickHouse
ClickHouse
372,038,740
208
Dimon70007
прописал ```<send_progress_in_http_headers>1</send_progress_in_http_headers>``` ```<http_headers_progress_interval_ms>500</http_headers_progress_interval_ms>``` в users.xml. обновился до версии 1.1.54343 b отправка информации о прогрессе заработала как надо (на версии 1.1.54236 не работало) может кому пригодится. Еще...
2018-03-10T15:35:34
2018-03-10T15:35:34
{}
NONE
ClickHouse
ClickHouse
372,175,394
1,997
buddiking
@alexey-milovidov: Thanks very much. the problem resolved.
2018-03-12T02:17:13
2018-03-12T02:17:13
{}
NONE
ClickHouse
ClickHouse
372,345,158
1,474
silviucpp
I found myself one here: https://discussions.apple.com/thread/8127677 ```sh cmake .. -DCMAKE_CXX_COMPILER=`which g++-7` -DCMAKE_C_COMPILER=`which gcc-7` -DCMAKE_BUILD_TYPE=Release -DCMAKE_OSX_DEPLOYMENT_TARGET=10.12 ```
2018-03-12T15:14:23
2018-03-12T15:14:23
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,417,387
1,474
vavrusa
I think I built it with clang as a workaround, better if you can build it with gcc-7
2018-03-12T18:34:07
2018-03-12T18:34:07
{}
CONTRIBUTOR
ClickHouse
ClickHouse
372,908,893
2,035
alexey-milovidov
``` + pwd / + mkdir -p /etc/clickhouse-server/config.d + cp ssl.xml /etc/clickhouse-server/config.d cp: cannot stat 'ssl.xml': No such file or directory ```
2018-03-14T05:27:28
2018-03-14T05:27:28
{}
MEMBER
ClickHouse
ClickHouse
373,181,435
2,021
innerr
Wonderful!
2018-03-14T21:30:27
2018-03-14T21:30:27
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,302,411
2,041
igor-sh8
Great. It works for me. Thanks.
2018-03-15T08:47:02
2018-03-15T08:47:02
{}
NONE
ClickHouse
ClickHouse
373,395,117
2,060
peteychuk
any luck on this?
2018-03-15T14:29:16
2018-03-15T14:29:16
{}
NONE
ClickHouse
ClickHouse
373,413,375
2,044
ludv1x
The problem is even simpler `with ['a'] as w select 'a' IN w`
2018-03-15T15:19:34
2018-03-15T15:43:05
{}
CONTRIBUTOR
ClickHouse
ClickHouse
370,865,445
2,001
kmatt
Other CentOS machines running same version can run DESCRIBE without error. Is there any other debug information that clients on the problem machines can emit?
2018-03-06T17:42:01
2018-03-06T17:42:01
{}
NONE
ClickHouse
ClickHouse
371,259,454
1,901
alexey-milovidov
Looks like it will be much better to implement this feature through `DatabaseMemory` class. (You can have all temporary tables inside `DatabaseMemory` in the context.)
2018-03-07T19:40:56
2018-03-07T19:40:56
{}
MEMBER
ClickHouse
ClickHouse
371,336,958
1,901
zhang2014
Sorry, we don't need it, I just solve a few issue during the holiday, hoping to help you : )
2018-03-08T00:39:50
2018-03-08T00:39:50
{}
CONTRIBUTOR
ClickHouse
ClickHouse
371,955,327
2,014
bubaflub
If folks think it is a good idea I can put a PR up that implements what I'm suggesting.
2018-03-09T21:51:06
2018-03-09T21:51:06
{}
NONE
ClickHouse
ClickHouse
372,865,289
2,016
alexey-milovidov
We can use `serializeText` in `system.parts`, `system.parts_columns` instead of `serializeTextQuoted`. We should never use `serializeText` to write to filesystem, zookeeper, etc. As I see, it doesn't happen. The patch looks fine!
2018-03-14T00:32:38
2018-03-14T00:32:38
{}
MEMBER
ClickHouse
ClickHouse
372,892,696
2,012
alexey-milovidov
``` String getName() const override { return "IntersectionsMax"; } ``` this is wrong according that you have two different functions.
2018-03-14T03:26:44
2018-03-14T03:26:52
{}
MEMBER
ClickHouse
ClickHouse
373,375,061
2,058
filimonov
Error description in not helpful, but may be you should provide node id (after shard name) according to recommendations in documentation? > Example: > ReplicatedMergeTree('/clickhouse/tables/{layer}-{shard}/hits', '{replica}', EventDate, intHash32(UserID), (CounterID, EventDate, intHash32(UserID), EventTime), 8192)...
2018-03-15T13:27:42
2018-03-15T13:27:42
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,381,775
2,012
furmur
thank you for such a lot of fixes. i see that algorithm was completely changed. i used rb-tree to keep array sorted and to really aggregate data instead of adding points to the plain array. your algorithm sorts accumulated array right in the `insertResultInto` and simply counts points. it's more straightforward bu...
2018-03-15T13:49:47
2018-03-15T13:49:47
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,395,747
2,060
rpfennin
Yes actually, I removed all clickhouse references from /etc/apt/sources.list and deleted the clickhouse.list file in /etc/apt/sources.list.d Then I re-added the repos, updated, and upgraded successfully
2018-03-15T14:31:06
2018-03-15T14:31:06
{ "+1": 1 }
NONE
ClickHouse
ClickHouse
372,563,928
1,697
millin
I found that CH send data to process's STDIN only when process already send STDOUT. So, I temporary solved this by adding **`echo ''`** to the top of my script. Is this an expected behavior?
2018-03-13T06:48:22
2018-03-13T06:48:22
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,082,765
2,044
filimonov
test case to reproduce that error message (not sure if it's correct) ``` with ['a','b','c'] as w select arrayFilter(x -> x IN w, c) FROM ( select ['a','b','c','d','e','f','g','h'] c ) ; Code: 60, e.displayText() = DB::Exception: Table default.w doesn't exist., e.what() = DB::Exception ``` workaround (use has ins...
2018-03-14T16:19:47
2018-03-14T16:19:59
{ "+1": 1 }
CONTRIBUTOR
ClickHouse
ClickHouse
373,096,114
2,034
ludv1x
I can't reproduce the problem: ``` DROP TABLE IF EXISTS test.a; DROP TABLE IF EXISTS test.b; create table test.a (t Int32, v Int32) ENGINE = Log; create table test.b (t Int32, v Int32) ENGINE = Log; INSERT INTO test.a VALUES (0, 0); INSERT INTO test.b VALUES (0, 0); SELECT t , CAST(0 AS Float32)...
2018-03-14T16:56:13
2018-03-14T16:57:31
{}
CONTRIBUTOR
ClickHouse
ClickHouse
373,474,498
2,054
rajadayalan
Thanks. We don's see errors in ClickHouse and replication looks fine. Will ignore the user-level (INFO) in Zookeeper logs.
2018-03-15T18:17:13
2018-03-15T18:17:54
{}
NONE