Datasets:
open-index
/
open-github-issues

Dataset card Data Studio Files
xet
 Community
1
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
549,184,693
7,413
alexey-milovidov
@reflection What is the `SHOW CREATE TABLE brain.brain_mb`?
2019-11-03T22:15:27
2019-11-03T22:15:27
{}
MEMBER
ClickHouse
ClickHouse
549,185,167
5,493
alexey-milovidov
In recent versions: `ALTER TABLE ... MODIFY SETTING merge_with_ttl_timeout = ...`
2019-11-03T22:20:39
2019-11-03T22:20:39
{}
MEMBER
ClickHouse
ClickHouse
549,186,405
7,312
alexey-milovidov
> so what's this znode means? Replica has a session with ZooKeeper and can serve INSERT queries. > byw,when I use replicated table,every insert operation one znode would be created,should I need to delete znode manually after some days like a month to prevent so many znodes in zk? No, you don't need to do anyt...
2019-11-03T22:36:02
2019-11-03T22:36:24
{}
MEMBER
ClickHouse
ClickHouse
549,187,921
7,223
alexey-milovidov
There can be multiple reasons why the same query on identical data in different tables may take different time. For example, the sequential read speed from HDD can depend on data location in sectors (close or far from the center of the disk). For ext4, it's typical that two files can be placed far from each other and y...
2019-11-03T22:54:24
2019-11-03T22:54:24
{}
MEMBER
ClickHouse
ClickHouse
549,189,263
6,985
alexey-milovidov
It's about https://github.com/ClickHouse/ClickHouse/blob/master/CHANGELOG.md#backward-incompatible-change-3 (input_format_defaults_for_omitted_fields)
2019-11-03T23:10:04
2019-11-03T23:10:31
{}
MEMBER
ClickHouse
ClickHouse
549,199,638
7,563
introspection3
> @564064202 see https://clickhouse.yandex/docs/en/operations/table_engines/replication/ but there is only shard config (with zookeeper)
2019-11-04T01:10:09
2019-11-04T01:10:09
{}
NONE
ClickHouse
ClickHouse
549,282,728
7,554
filimonov
ERROR: type should be string, got "\r\nhttps://arrow.apache.org/blog/2019/10/13/introducing-arrow-flight/\r\nhttps://arrow.apache.org/docs/format/Flight.html"
2019-11-04T09:52:30
2019-11-04T09:52:30
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,463,702
7,617
cw9
Yes, here is the example: create table ORX( A Int64, B Int64, V Int64) Engine=MergeTree order by (A,B) create table m as ORX ENGINE = Merge(default, 'ORX') insert into ORX select 111, number, number from numbers(10000000) select * from ORX where A = 111 order by A,B limit 3 settings max_threads=1; ``` 3...
2019-11-04T17:34:11
2019-11-04T17:34:11
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,596,769
7,224
alexey-milovidov
We can add two ProfileEvents: FailedSelectQuery, FailedInsertQuery. This is very easy task.
2019-11-04T23:38:24
2019-11-04T23:38:24
{}
MEMBER
ClickHouse
ClickHouse
549,735,290
7,628
Tasselmi
> place I download all the 3 rpm files -- clickhosue-common, clickhouse-client, clickhouse-server locally. and use `yum localinstall xxxx` to install them.
2019-11-05T09:19:59
2019-11-05T09:19:59
{}
NONE
ClickHouse
ClickHouse
549,852,067
6,546
meob
We upgraded to 19.15 and we didn't get the error again. I think it had been fixed. Thanks!
2019-11-05T14:41:27
2019-11-05T14:41:27
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,253,675
7,488
filimonov
https://github.com/ClickHouse/ClickHouse/blob/8c24d5e66a4982f4faaa31106496e6eecd886aaa/dbms/src/Interpreters/ReplaceQueryParameterVisitor.cpp#L57 deserializeTextEscaped ?
2019-11-04T08:07:47
2019-11-04T08:12:59
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,314,648
7,606
alexey-milovidov
Probably your use case outweight my use case :)
2019-11-04T11:30:58
2019-11-04T11:30:58
{}
MEMBER
ClickHouse
ClickHouse
549,726,387
7,546
ariesdevil
Use ```shell cmake .. -DCMAKE_CXX_COMPILER=`which g++-9` -DCMAKE_C_COMPILER=`which gcc-9` -DCMAKE_BUILD_TYPE=Debug ``` on MacOS Catalina(ver: 10.15.1 (19B88)) and got: ``` CMake Error at cmake/darwin/default_libs.cmake:4 (message): Darwin build is supported only for Clang Call Stack (most recent call first):...
2019-11-05T08:57:05
2019-11-05T08:59:18
{}
NONE
ClickHouse
ClickHouse
549,165,204
7,425
excitoon
Reproduction scenario: ``` clickhouse client -q "drop table if exists t" clickhouse client -q "create table t (number UInt64, x String) Engine=MergeTree ORDER BY number SETTINGS storage_policy='jbods_with_external'" clickhouse client -q "system stop merges t" clickhouse client -q "insert into t values (1, '1')" c...
2019-11-03T18:31:52
2019-11-03T18:31:52
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,185,366
521
alexey-milovidov
The `input` table function was implemented. https://clickhouse.github.io/clickhouse-presentations/meetup30/new_features/#21
2019-11-03T22:23:07
2019-11-03T22:23:07
{}
MEMBER
ClickHouse
ClickHouse
549,313,471
7,606
alexey-milovidov
I always kept this in mind but refused to implement because of limited benefits and compromised simplicity of code. But it's still nice to have.
2019-11-04T11:27:23
2019-11-04T11:27:23
{}
MEMBER
ClickHouse
ClickHouse
549,811,478
7,631
alesapin
Thank you.
2019-11-05T12:56:31
2019-11-05T12:56:31
{}
MEMBER
ClickHouse
ClickHouse
549,181,873
7,514
alexey-milovidov
The relevant code is here: https://github.com/ClickHouse/ClickHouse/blob/594c535bd419d153efe97bd9e93caef4b96cd692/dbms/src/Common/SymbolIndex.cpp#L284
2019-11-03T21:41:53
2019-11-03T21:41:53
{}
MEMBER
ClickHouse
ClickHouse
549,182,016
7,514
alexey-milovidov
How to check: https://sourceware.org/gdb/onlinedocs/gdb/Separate-Debug-Files.html
2019-11-03T21:43:39
2019-11-03T21:43:39
{}
MEMBER
ClickHouse
ClickHouse
549,182,278
7,488
alexey-milovidov
@filimonov I wonder why it is not solved three days ago, when we discussed the solution.
2019-11-03T21:46:38
2019-11-03T21:46:38
{}
MEMBER
ClickHouse
ClickHouse
549,187,764
7,223
alexey-milovidov
You can look at the query metics and compare them: https://clickhouse.github.io/clickhouse-presentations/dataops_2019/#29 (press "right" and look at the following slides)
2019-11-03T22:52:24
2019-11-03T22:52:24
{}
MEMBER
ClickHouse
ClickHouse
549,346,255
7,612
manojfim
Is there a out of the box solution for importing all the csv files in a s3 directory?
2019-11-04T13:10:19
2019-11-04T13:10:19
{}
NONE
ClickHouse
ClickHouse
549,463,184
7,612
manojfim
Is this will be supported in future? On Nov 4, 2019, 8:59 AM -0500, Denis Zhuravlev <notifications@github.com>, wrote: > There is no out of the box solution right now (engine URL does not support auth, engine S3 is not released). > You should use local files. > — > You are receiving this because you authored the t...
2019-11-04T17:32:49
2019-11-04T17:33:14
{}
NONE
ClickHouse
ClickHouse
549,495,714
7,612
den-crane
Not sure, probably it works in test releases like this ``` insert into mytable select * from s3('http://s3auth:s3auth@host/bucket/test.csv', 'CSV', 'column1 UInt32, column2 UInt32, column3 UInt32') ```
2019-11-04T18:53:45
2019-11-04T18:54:26
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,665,586
7,608
amosbird
btw, some perf tests use `total_time` as its main_metric, which is confusing.
2019-11-05T05:12:10
2019-11-05T05:12:10
{}
COLLABORATOR
ClickHouse
ClickHouse
549,681,006
6,526
boqu
@alexey-milovidov Could you please share some details about the fix(e.g. commit, release)? Thanks!
2019-11-05T06:27:32
2019-11-05T06:27:32
{}
NONE
ClickHouse
ClickHouse
549,704,081
7,611
RaulGuo
> https://www.altinity.com/blog/2018/5/10/circular-replication-cluster-topology-in-clickhouse > > But it's not fully supported: > > * DDL on cluster won't work. > * Double distributed joins will fail with exception "Table not found" > > Personally I don't recommend this schema because second replicas kill per...
2019-11-05T07:50:43
2019-11-05T07:50:43
{}
NONE
ClickHouse
ClickHouse
549,795,313
7,610
alesapin
I'm also interested in the root cause of this problem. @den-crane
2019-11-05T12:05:29
2019-11-05T12:05:29
{}
MEMBER
ClickHouse
ClickHouse
549,189,421
6,981
alexey-milovidov
We should never allow to do build without submodules by default, but it's possible to introduce a flag like `-D TROUBLE=1` that the user can set for its own trouble.
2019-11-03T23:12:14
2019-11-03T23:12:14
{}
MEMBER
ClickHouse
ClickHouse
549,316,465
7,606
amosbird
> I also used Null format to test network throughput: > > `clickhouse-client --host remote-server --query 'SELECT rand64() FROM system.numbers FORMAT Null'` > > Now this use case will fail :( Interesting use case. Perhaps we can revert this and implement alternative. I also tried extending the `SELECT INTO OUT...
2019-11-04T11:36:58
2019-11-04T11:36:58
{}
COLLABORATOR
ClickHouse
ClickHouse
549,319,993
7,600
alexey-milovidov
> ClickHouse build check Will be fixed after you merge with master.
2019-11-04T11:48:29
2019-11-04T11:48:29
{}
MEMBER
ClickHouse
ClickHouse
549,387,548
7,613
nvartolomei
cc @alesapin need `can be tested` label as I can't run these tests on my setup. All this should be ok since #5757.
2019-11-04T14:48:07
2019-11-04T15:02:39
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,432,847
7,611
den-crane
https://www.altinity.com/blog/2018/5/10/circular-replication-cluster-topology-in-clickhouse But it's not fully supported: * DDL on cluster won't work. * Double distributed joins will fail with exception "Table not found" Personally I don't recommend this schema because second replicas kill performance (they use...
2019-11-04T16:25:35
2019-11-04T16:28:44
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,536,953
7,601
azat
Hm, this is not the only place that can have problems, some of callers does not affected, but some are, in particular: - `MutationsInterpreter::execute` - `StorageMerge::createSourceStreams` So maybe instead the callers should care (copy it?)? i.e. `FunctionsExternalDictionaries.h` and related
2019-11-04T20:40:13
2019-11-04T20:40:13
{ "+1": 1 }
MEMBER
ClickHouse
ClickHouse
549,704,629
7,611
Tasselmi
> performance you need ask boss for more money......
2019-11-05T07:52:27
2019-11-05T07:52:27
{}
NONE
ClickHouse
ClickHouse
549,790,742
7,628
alesapin
@Tasselmi How your download URL looked like?
2019-11-05T11:50:53
2019-11-05T11:51:10
{}
MEMBER
ClickHouse
ClickHouse
549,180,882
7,599
azat
I thought that since it is internal, tests is not needed, anyway added.
2019-11-03T21:31:26
2019-11-03T21:31:26
{}
MEMBER
ClickHouse
ClickHouse
549,240,518
7,598
yuzhichang
> I would name it `bitmapTransform` as columns are very much immutable at SQL level. Done as you suggested.
2019-11-04T06:57:58
2019-11-04T06:57:58
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,313,986
7,606
alexey-milovidov
I also used Null format to test network throughput: `clickhouse-client --progress --host remote-server --query 'SELECT rand64() FROM system.numbers FORMAT Null'` Now this use case will fail :(
2019-11-04T11:29:05
2019-11-04T11:30:01
{}
MEMBER
ClickHouse
ClickHouse
549,476,624
7,619
excitoon
Opened by mistake.
2019-11-04T18:06:19
2019-11-04T18:06:19
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,563,715
7,608
alexey-milovidov
Results of performance test look strange: https://clickhouse-test-reports.s3.yandex.net/7608/5c160d5efb0ca08b6b515d66d2abb0bb847713b4/performance_test/comparison_min_time_gcc_9.html Could you please compare the most different result manually with `clickhouse-benchmark`?
2019-11-04T21:51:20
2019-11-04T21:51:20
{}
MEMBER
ClickHouse
ClickHouse
549,778,192
7,601
azat
> Can you explain why we want InterpreterSelectQuery to live longer? To expand lifetime of the `InterpreterSelectQuery::context`, since [`FunctionDictGet*::execute`](https://github.com/ClickHouse/ClickHouse/blob/6a871f579fa9638acc08ac8275f3aa1d2736a8f6/dbms/src/Functions/FunctionsExternalDictionaries.h#L833) uses i...
2019-11-05T11:12:03
2019-11-05T11:13:30
{}
MEMBER
ClickHouse
ClickHouse
549,184,795
7,410
alexey-milovidov
https://clickhouse.github.io/clickhouse-presentations/meetup30/new_features/#12
2019-11-03T22:16:31
2019-11-03T22:16:31
{}
MEMBER
ClickHouse
ClickHouse
549,186,516
7,302
alexey-milovidov
Does it run successfully on Raspberry Pi, your Android phone and Cavium?
2019-11-03T22:37:16
2019-11-03T22:37:16
{}
MEMBER
ClickHouse
ClickHouse
549,233,316
7,421
ghost
@alexey-milovidov Thank you for your explanation.
2019-11-04T06:11:24
2019-11-04T06:11:24
{}
NONE
ClickHouse
ClickHouse
549,311,008
7,604
4ertus2
It's expected. There's no join order optimizer. You should create joins with small right sides. There's Hash-Join under the hood so your right table would be entirely read and placed into hash-map, then stream from left table would be filtered by this hash-map. It's cheaper to have small map + big stream then big map +...
2019-11-04T11:19:02
2019-11-04T11:19:02
{ "+1": 2 }
CONTRIBUTOR
ClickHouse
ClickHouse
549,374,359
7,610
den-crane
Stop queries / ingestion ``` detach table transformed_ts_1minute; # mv /data/clickhouse/data/spoor/transformed_ts_1minute/20191104_15374027_15374051_3 /tmp attach table transformed_ts_1minute; ``` start queries / ingestion
2019-11-04T14:18:45
2019-11-04T14:18:57
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,445,697
7,617
den-crane
Do you mean Engine=Merge? Do you have reproducible example?
2019-11-04T16:52:48
2019-11-04T16:53:06
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,664,827
7,608
amosbird
The perf results almost don't make any sense to me. The slow down queries aren't related to any of this pr's changes
2019-11-05T05:08:22
2019-11-05T05:16:49
{}
COLLABORATOR
ClickHouse
ClickHouse
549,728,658
7,607
ghost
@den-crane I will try newer version of ClickHouse. :) Thank you~
2019-11-05T09:02:48
2019-11-05T09:02:48
{}
NONE
ClickHouse
ClickHouse
549,181,666
7,514
alexey-milovidov
It does not relate to `libunwind`.
2019-11-03T21:40:01
2019-11-03T21:40:01
{}
MEMBER
ClickHouse
ClickHouse
549,182,676
7,441
alexey-milovidov
It can be easily fixed, we just need to look at the mount point of `data` as the default.
2019-11-03T21:51:24
2019-11-03T21:51:34
{ "+1": 2 }
MEMBER
ClickHouse
ClickHouse
549,620,297
7,627
den-crane
That means that 4 parts will be read because 4 parts is suitable for WHERE conditions by partition key (_parts by date_)/ primary key index(parts by key). 1030 marks means that this range of marks need to read to solve the query, 1030 granules by '8192' rows. ``` create table ZZ (A Int64, B Int64, C String) Engin...
2019-11-05T01:18:08
2019-11-05T01:21:20
{ "+1": 1 }
CONTRIBUTOR
ClickHouse
ClickHouse
549,648,188
7,625
Joeywzr
@alexey-milovidov Thanks! I've read your code, which is much more efficient than mine. I learned a lot.
2019-11-05T03:31:56
2019-11-05T03:31:56
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,717,164
7,628
blinkov
@Tasselmi which method have you used to install ClickHouse in the first place?
2019-11-05T08:32:26
2019-11-05T08:32:34
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,183,016
7,438
alexey-milovidov
@vitlibar Is it possible to fix? Should we look at the Enum indices instead of names? If it's non trivial and too cumbersome - then just close this issue.
2019-11-03T21:54:56
2019-11-03T21:54:56
{}
MEMBER
ClickHouse
ClickHouse
549,184,634
7,413
alexey-milovidov
@CurtizJ If there is no WHERE and the overall LIMIT is small, then lower the number of query execution threads.
2019-11-03T22:14:45
2019-11-03T22:14:45
{}
MEMBER
ClickHouse
ClickHouse
549,186,172
7,438
vitlibar
The current code matches the enum types by the names of the constants. But there are other options: enum can be automatically converted to integer or string and vice versa. So the following script should work: ``` clickhouse-client --password=default --query="CREATE TABLE bug_enum (e Enum8('A'=0, 'B'=1, 'C'=2, ''=...
2019-11-03T22:32:56
2019-11-03T22:32:56
{}
MEMBER
ClickHouse
ClickHouse
549,186,922
7,262
alexey-milovidov
I also have this use case quite frequently. I'd like if it will be named `cluster_all_replicas`.
2019-11-03T22:42:51
2019-11-03T22:42:51
{}
MEMBER
ClickHouse
ClickHouse
549,188,449
7,189
alexey-milovidov
> but I am unable to start How does it look?
2019-11-03T23:00:06
2019-11-03T23:00:06
{}
MEMBER
ClickHouse
ClickHouse
549,189,642
6,967
alexey-milovidov
> Trying to read the yandex metrics data using Python, but facing encoding errors when using the utf-8. The data is neither in UTF-8 nor in ISO-8859-1 or ASCII. You should treat this data as binary (although TSV dump mostly looks like text). After you read this data, you can treat separate string field values li...
2019-11-03T23:14:51
2019-11-03T23:18:51
{}
MEMBER
ClickHouse
ClickHouse
549,224,927
7,517
zhang2014
related: #329 In ClickHouse, the data is only stored in order of the `order by clause` and primary key info is generated based on the `primary key clause`(default same to `order by clause`). When executing query, the `primary key index` always works on the granules level (filtering or retaining a granule). ...
2019-11-04T04:58:09
2019-11-04T04:58:09
{ "+1": 3 }
CONTRIBUTOR
ClickHouse
ClickHouse
549,225,497
7,517
chu1070y
Thanks for information. but, if you find the answer of below question, > Does the Clickhouse re-sort the data to make granule`s info? or uses the data already sorted by primary key? Please comment on this issue.
2019-11-04T05:03:33
2019-11-04T05:09:09
{}
NONE
ClickHouse
ClickHouse
549,298,444
7,610
huang--yijian
We see an issue https://github.com/ClickHouse/ClickHouse/pull/6748 with the same error message. However, we are not using skip indices mentioned in issue 6748, so it is not related to ours.
2019-11-04T10:39:55
2019-11-04T10:39:55
{}
NONE
ClickHouse
ClickHouse
549,584,523
7,626
alexey-milovidov
BTW, it's in question why not return FixedString instead of String. What do you think?
2019-11-04T22:53:38
2019-11-04T22:53:38
{}
MEMBER
ClickHouse
ClickHouse
549,793,771
7,628
blinkov
@Tasselmi why haven't you just followed the official instruction from https://clickhouse.yandex/#quick-start ? (which shows up after clicking on CentOS/RedHat)
2019-11-05T12:00:25
2019-11-05T12:00:33
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,835,783
7,546
ernestp
@ariesdevil As stated in error message `Darwin build is supported only for Clang` compiler. GCC is not supported for Darwin. Try without specifying compiler. This should auto detect clang compiler. `cmake ..` Make sure that XCode and command line tools is installed. If issues without specifying compiler the...
2019-11-05T14:03:14
2019-11-05T14:05:54
{ "+1": 1 }
CONTRIBUTOR
ClickHouse
ClickHouse
549,184,420
7,416
alexey-milovidov
Pointing home directory of the daemon user to `/nonexistent` is a common practice. But why you need a home directory for `clickhouse` user?
2019-11-03T22:12:01
2019-11-03T22:12:01
{}
MEMBER
ClickHouse
ClickHouse
549,185,054
6,263
alexey-milovidov
This issue will be addressed by development of "polymorphic parts" that is currently in progress by @CurtizJ
2019-11-03T22:19:23
2019-11-03T22:19:23
{}
MEMBER
ClickHouse
ClickHouse
549,188,084
7,207
alexey-milovidov
Many users successfully run ClickHouse on servers with 4 GiB RAM, although this usage is indeed limited.
2019-11-03T22:56:03
2019-11-03T22:56:03
{}
MEMBER
ClickHouse
ClickHouse
549,365,576
7,612
den-crane
There is no out of the box solution right now (engine URL does not support auth, engine S3 is not released). You should use local files.
2019-11-04T13:59:06
2019-11-04T13:59:06
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,182,485
7,468
alexey-milovidov
@filimonov Are you sure that this setting should be used to limit mutations?
2019-11-03T21:49:02
2019-11-03T21:49:02
{}
MEMBER
ClickHouse
ClickHouse
549,183,325
7,421
alexey-milovidov
If data1 is broken, you should replace the disk (create filesystem and mount at the same location) and restart clickhouse-server. ClickHouse will download missing parts from replica.
2019-11-03T21:58:38
2019-11-03T21:59:21
{ "+1": 1 }
MEMBER
ClickHouse
ClickHouse
549,184,449
7,416
alexey-milovidov
@proller please help to clarify.
2019-11-03T22:12:19
2019-11-03T22:12:19
{}
MEMBER
ClickHouse
ClickHouse
549,191,446
6,688
alexey-milovidov
@BayoNet has failed.
2019-11-03T23:37:14
2019-11-03T23:37:14
{}
MEMBER
ClickHouse
ClickHouse
549,197,139
7,599
amosbird
Thanks!
2019-11-04T00:44:14
2019-11-04T00:44:14
{}
COLLABORATOR
ClickHouse
ClickHouse
549,623,998
7,627
den-crane
create table ZZ (A Int64, B Int64, C String) Engine = MergeTree order by B; the same as create table ZZ (A Int64, B Int64, C String) Engine = MergeTree **partition by tuple()** order by B; One partition (tuple()) == all. No partition pruning / elimination.
2019-11-05T01:33:51
2019-11-05T01:34:48
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,821,220
7,629
alesapin
Fix https://github.com/ClickHouse/ClickHouse/pull/7635
2019-11-05T13:25:05
2019-11-05T13:25:05
{ "+1": 1 }
MEMBER
ClickHouse
ClickHouse
549,181,365
7,537
alexey-milovidov
> Do you use official build? @filimonov You can see it from the log message: `version 19.14.3.3 (official build)`
2019-11-03T21:36:31
2019-11-03T21:36:49
{}
MEMBER
ClickHouse
ClickHouse
549,252,068
7,488
filimonov
> Alexey Milovidov, [25.10.19 17:31] > Для параметров используется парсинг из текста (например, параметра URL), метод что-то типа deserializeWholeAsText. Надо уметь различать строку 'NULL' и NULL. Насколько я понял, реализация просто скопирована из CSV, где можно указывать строки без кавычек, но для разрешения неодноз...
2019-11-04T08:00:23
2019-11-04T08:00:23
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,317,593
7,605
alexey-milovidov
Ok. The meaning of this value was not documented neither tested.
2019-11-04T11:40:41
2019-11-04T11:40:41
{}
MEMBER
ClickHouse
ClickHouse
549,622,251
7,627
chu1070y
When I create table, I do not use partition key, Does the ClickHouse use partition key automatically?
2019-11-05T01:26:14
2019-11-05T01:26:14
{}
NONE
ClickHouse
ClickHouse
549,642,544
7,518
sundy-li
@alexey-milovidov Could u help provide some advice on this implementation? ` range(start, end, step) ` here we may get some partial const columns. It impacts performance if we convert it to full Column. It's hard to cover all just by the template codegen. And it may result to `C(3, 1) + C(3, 2) + c(3, ...
2019-11-05T03:01:52
2019-11-05T03:39:04
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,780,174
1,981
lexerom
Also had such error. In logs I had: <Error> Application: Exception: Failed to merge config with '/etc/clickhouse-server/users.d/default-password.xml': SAXParseException: Invalid token in '/etc/clickhouse-server/users.d/default-password.xml', line 1 column 38. It was easy to fix - I updated the password for default u...
2019-11-05T11:18:11
2019-11-05T11:18:11
{ "+1": 1 }
NONE
ClickHouse
ClickHouse
549,183,496
7,421
alexey-milovidov
You can also remove disk1 from volumes configuration and restart clickhouse-server. ClickHouse will download missing parts from replica and place them on disk2.
2019-11-03T22:00:51
2019-11-03T22:00:51
{ "+1": 4 }
MEMBER
ClickHouse
ClickHouse
549,190,441
6,820
alexey-milovidov
@namikhnenko @evillique @kekekekule
2019-11-03T23:25:20
2019-11-03T23:25:20
{}
MEMBER
ClickHouse
ClickHouse
549,426,957
7,616
SQkaer
@sundy-li ,please review when you are free
2019-11-04T16:13:29
2019-11-04T16:13:29
{}
NONE
ClickHouse
ClickHouse
549,587,054
7,518
alexey-milovidov
Ok. And please also commit this performance test.
2019-11-04T23:01:32
2019-11-04T23:01:32
{}
MEMBER
ClickHouse
ClickHouse
549,575,403
7,625
alexey-milovidov
@Joeywzr BTW, you have won Hacktoberfest!
2019-11-04T22:24:44
2019-11-04T23:10:49
{}
MEMBER
ClickHouse
ClickHouse
549,681,595
7,312
byx313
> > so what's this znode means? > > Replica has a session with ZooKeeper and can serve INSERT queries. > > > byw,when I use replicated table,every insert operation one znode would be created,should I need to delete znode manually after some days like a month to prevent so many znodes in zk? > > No, you don't n...
2019-11-05T06:29:58
2019-11-05T06:29:58
{}
NONE
ClickHouse
ClickHouse
549,293,696
7,610
huang--yijian
We can afford to delete/detach this bad part, so that the query does not fail. But how to delete/detach a specific part?
2019-11-04T10:25:25
2019-11-04T10:25:25
{}
NONE
ClickHouse
ClickHouse
549,419,464
7,607
den-crane
You should upgrade to 19.13.6.51 https://github.com/ClickHouse/ClickHouse/pull/6404
2019-11-04T15:57:03
2019-11-04T15:57:03
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,576,724
7,328
alexey-milovidov
See the linked PR.
2019-11-04T22:28:25
2019-11-04T22:28:25
{}
MEMBER
ClickHouse
ClickHouse
549,636,525
7,610
huang--yijian
> Stop queries / ingestion > > ``` > detach table transformed_ts_1minute; > # mv /data/clickhouse/data/spoor/transformed_ts_1minute/20191104_15374027_15374051_3 /tmp > attach table transformed_ts_1minute; > ``` > > start queries / ingestion It works. Thank you for your help! What's the root cause ? And is...
2019-11-05T02:32:25
2019-11-05T02:32:25
{}
NONE
ClickHouse
ClickHouse
549,664,615
7,608
amosbird
# Summary This pr optimizes three use cases. 1. arrayReduce: 0.9 sec -> 0.56 sec, **60%** improvement 2. -Array combinator: 0.9 sec -> 0.34 sec, **260%** improvement 3. -State combinator: 3.0 sec -> 0.5 sec, **600%** improvement The actual improvements depend on aggregate functions, group by methods, group...
2019-11-05T05:07:12
2019-11-05T05:07:12
{}
COLLABORATOR
ClickHouse
ClickHouse
549,677,950
7,626
sundy-li
> BTW, it's in question why not return FixedString instead of String. What do you think It will be better. Will be better to add new tests for Chinese words: ``` SELECT char(0xe4, 0xbd, 0xa0, 0xe5, 0xa5, 0xbd) AS hello; 你好 ```
2019-11-05T06:13:40
2019-11-05T06:30:35
{}
CONTRIBUTOR
ClickHouse
ClickHouse
549,683,701
7,611
Tasselmi
> I have built a clickhouse with 3 nodes(let's say it is ip1, ip2, and ip3), now I tries to build a cluster in the metrica.xml. My plan is to build one with three shards, and each shards has two replicas when I use distributed table using this cluster, so I defined my cluster named distribute_cluster like this: > > ...
2019-11-05T06:38:14
2019-11-05T06:38:14
{}
NONE
ClickHouse
ClickHouse
549,744,270
7,601
alesapin
Hi, @azat! Can you explain why we want `InterpreterSelectQuery` to live longer? All streams have type `BlockInputStreamPtr` and shouldn't be destroyed in `InterpreterSelectQuery's` destructor. I have executed your test locally and it works without this fix. Maybe I need to run it with sanitizer?
2019-11-05T09:42:03
2019-11-05T09:42:03
{}
MEMBER
ClickHouse
ClickHouse
564,241,950
8,128
alexey-milovidov
What happens when I connect to the server: Was: ``` $ sudo strace -f -e setsockopt -p $(pidof clickhouse-server) [sudo] password for milovidov: strace: Process 3879 attached with 64 threads [pid 3942] setsockopt(15, SOL_TCP, TCP_NODELAY, [1], 4) = 0 [pid 3935] setsockopt(15, SOL_TCP, TCP_NODELAY, [1], 4) = ...
2019-12-10T20:26:44
2019-12-10T20:26:44
{}
MEMBER