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 | 449,481,051 | 3,901 | den-crane | >In this case, is it still true that Clickhouse cannot find the max(timestamp) by index?
It does not matter how was created a table. The only way for CH to find max -- to scan everything. There is no such optimization -- to use index for max -- yet.
create table mx(a Int64) engine=MergeTree order by a;
insert in... | 2018-12-21T19:43:01 | 2018-12-21T19:43:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,581,241 | 3,905 | alexey-milovidov | Ok. And what about performance test? | 2018-12-22T16:21:26 | 2018-12-22T16:21:26 | {} | MEMBER |
ClickHouse | ClickHouse | 449,598,949 | 3,579 | alexey-milovidov | If lines in Kafka messages doesn't contain trailing newline character, you can specify implicit newline character when creating Kafka table using `kafka_row_delimiter` setting.
`CREATE TABLE ... SETTINGS kafka_row_delimiter = '\n'`
This feature was added in #2701 and subsequently refined in #2781 | 2018-12-22T21:36:17 | 2018-12-22T21:36:17 | {} | MEMBER |
ClickHouse | ClickHouse | 449,818,973 | 3,917 | alesapin | Need example and full error text. | 2018-12-25T07:49:55 | 2018-12-25T07:49:55 | {} | MEMBER |
ClickHouse | ClickHouse | 449,929,323 | 3,876 | qietingfengling | OK,thanks | 2018-12-26T08:32:35 | 2018-12-26T08:32:35 | {} | NONE |
ClickHouse | ClickHouse | 449,971,304 | 3,927 | wenxingxing | @den-crane @blinkov I've tried the `mysql` before `MySQL`, getting the same answer.
I'll tried it tomorrow, thank's for ur reply, happy new year! | 2018-12-26T14:06:51 | 2018-12-26T14:06:51 | {} | NONE |
ClickHouse | ClickHouse | 449,999,453 | 3,891 | ztlpn | The `No message received` error is a separate problem. @alesapin is looking into it. | 2018-12-26T17:42:07 | 2018-12-26T17:42:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,428,315 | 3,780 | blinkov | @PascalWithopf do you have any further questions? | 2018-12-21T16:07:26 | 2018-12-21T16:07:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,439,033 | 3,782 | ztlpn | @blinkov don't close it yet :)
I'll close when an easy method to cancel a mutation will be implemented. | 2018-12-21T16:48:28 | 2018-12-21T16:48:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,537,564 | 3,893 | sliontc | I'm using DBeaver Enterprise version, the website is [https://dbeaver.com/](https://dbeaver.com/) | 2018-12-22T02:07:05 | 2018-12-22T02:07:05 | {} | NONE |
ClickHouse | ClickHouse | 449,640,861 | 3,910 | amosbird | > Why we can capture references in LazyBlockInputStream?
I assume you mean why we can't?
Because those references will be invalid upon the first read attempt of the `LazyBlockInputStream` . | 2018-12-23T14:44:19 | 2018-12-23T14:44:19 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 449,719,952 | 3,912 | alesapin | We don't maintain rpm packages, only debian. Maybe [altinity](https://www.altinity.com/) will help you. | 2018-12-24T11:05:46 | 2018-12-24T11:05:53 | {
"-1": 2
} | MEMBER |
ClickHouse | ClickHouse | 449,733,701 | 3,889 | alexey-milovidov | It should be named `bitEquals` to also have simple implementation for `nan`s.
(nans may be either bit equal or not) | 2018-12-24T13:21:13 | 2018-12-24T13:21:13 | {} | MEMBER |
ClickHouse | ClickHouse | 449,418,282 | 3,899 | alesapin | ~~Failed statefull test. There is dangerous bug.~~ | 2018-12-21T15:30:34 | 2018-12-24T15:13:33 | {} | MEMBER |
ClickHouse | ClickHouse | 450,093,275 | 3,931 | blinkov | @alexey-milovidov I think it makes sense only for compatibility reasons, @simPod have you seen this syntax with single quotes in some other system? | 2018-12-27T07:24:52 | 2018-12-27T07:24:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,164,323 | 3,939 | den-crane | Can you show the output from cl-client with processed stats for one shard
select idc,sum(cpuload18),count() from mdb.met430_base group by idc limit 1
If you have huge number of idc, huge results need to be transferred (so it maybe a network is a bottleneck) to a node query-initiator and then the initiator needs t... | 2018-12-27T14:51:14 | 2018-12-27T14:51:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,167,275 | 3,943 | alesapin | Please, show clickhouse-server logs. Even when server cannot start it write some information to log files. | 2018-12-27T15:06:04 | 2018-12-27T15:06:04 | {} | MEMBER |
ClickHouse | ClickHouse | 450,203,816 | 3,569 | alexey-milovidov | Fixed in master.
Please note that this bug actually had no effect. | 2018-12-27T18:17:00 | 2018-12-27T18:17:00 | {} | MEMBER |
ClickHouse | ClickHouse | 446,837,620 | 2,586 | zhang2014 | @alesapin Well, actually I just solved the conflict : (
That is my todo list hoping it can help you : )
- [x] Support Delta(UInt64) Codec
- [x] Support ALTER TABLE xxx MODIFY COLUMN CODEC Delta(UInt64)
- [ ] Select Codec using the compression_setting & refactor compression_setting config
- [ ] Support for engine... | 2018-12-13T04:19:02 | 2018-12-22T01:46:35 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,731,202 | 2,731 | code-of-kpp | I am facing similar problem with ODBC dictionary and complex keys: Clickhouse generates huge SQL with `WHERE (a = ... AND b = ...) OR (a = ... AND b = ...)...` filter. I am looking for a way to at least modify WHERE clause to leverage `IN` operator. | 2018-12-24T12:57:33 | 2018-12-24T12:57:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,852,363 | 3,899 | alesapin | 00653_verification_monotonic_data_load and 00700_decimal_round broken in master. 00806_alter_update in unbundled mode seems strange. | 2018-12-25T13:56:56 | 2018-12-25T13:56:56 | {} | MEMBER |
ClickHouse | ClickHouse | 449,869,595 | 3,924 | alexey-milovidov | Upstream PR: https://github.com/MariaDB/mariadb-connector-c/pull/91 | 2018-12-25T19:30:20 | 2018-12-25T19:30:20 | {} | MEMBER |
ClickHouse | ClickHouse | 449,642,103 | 3,911 | amosbird | This is obviously wrong. | 2018-12-23T15:02:33 | 2018-12-23T15:02:33 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 449,720,475 | 3,916 | alexey-milovidov | Ok.
PS. Missing test. I've made WHERE clause mandatory to avoid this issue. | 2018-12-24T11:10:54 | 2018-12-24T11:10:54 | {} | MEMBER |
ClickHouse | ClickHouse | 449,854,558 | 1,344 | alikrubin | Yes, that will be a great feature, @alexey-milovidov, could you share when it will be available.
Meanwhile I came up with the following workaround:
```select * from rc where created_utc > (select max(created_utc) from rc) - 60*60*24 and subreddit='programming' order by created_utc desc limit 10;
10 rows in set... | 2018-12-25T14:31:44 | 2018-12-25T14:33:34 | {} | NONE |
ClickHouse | ClickHouse | 450,178,325 | 3,945 | alesapin | Also, fixed compose version to 2.2 because in later version ipv6 network configuration is not available:
https://github.com/docker/compose/issues/4958#issuecomment-311466816 | 2018-12-27T15:59:09 | 2018-12-27T15:59:09 | {} | MEMBER |
ClickHouse | ClickHouse | 449,748,398 | 3,870 | alexey-milovidov | You are reading uninitialized data. | 2018-12-24T15:47:19 | 2018-12-24T15:47:19 | {} | MEMBER |
ClickHouse | ClickHouse | 449,814,626 | 3,917 | dmitriyshashkin | But the error said that the directory for the OLD table wasn't empty. Not for the new one. | 2018-12-25T07:08:42 | 2018-12-25T07:08:42 | {} | NONE |
ClickHouse | ClickHouse | 449,880,228 | 3,569 | alexey-milovidov | UBSan have a suggestion for us:
```
test.cpp:3:9: runtime error: negation of -2147483648 cannot be represented in type 'int'; cast to an unsigned type to negate this value to itself
```
C++ supports negate operator for unsigned types, that do exactly what we need without UB. | 2018-12-26T00:01:26 | 2018-12-26T00:01:26 | {} | MEMBER |
ClickHouse | ClickHouse | 449,880,341 | 3,569 | alexey-milovidov | ```
milovidov@milovidov-Latitude-E7470:~/work/ClickHouse$ cat test.cpp
void __attribute__((__noinline__)) f(int & x)
{
x = -x;
}
void __attribute__((__noinline__)) g(int & x)
{
x = ~x + 1;
}
void __attribute__((__noinline__)) h(int & x)
{
x = -static_cast<unsigned>(x);
}
#include <limits... | 2018-12-26T00:03:31 | 2018-12-26T00:03:31 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 450,113,248 | 3,939 | ghost | this is part of my users.xml configure:
<max_memory_usage>128000000000</max_memory_usage>
<max_block_size>65536100</max_block_size>
<max_threads>200</max_threads>
<max_streams_to_max_threads_ratio>10</max_streams_to_max_threads_ratio>
<max_read_buffer_size... | 2018-12-27T09:26:58 | 2018-12-27T09:26:58 | {} | NONE |
ClickHouse | ClickHouse | 450,140,257 | 3,933 | code-of-kpp | I wanted to build all possible combinations
```
b c d
a a a
b c a
b a a
b a d
a a d
a c d
a a d
```
(`['b', 'c', 'd']` is small but dynamic array) | 2018-12-27T12:16:16 | 2018-12-27T12:16:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,159,074 | 2,602 | TimonKK | I got same error for MergeTree. And after that I have gotten error in each query for table. But when I execute ALTER again - it helped fix error | 2018-12-27T14:20:27 | 2018-12-27T14:20:27 | {} | NONE |
ClickHouse | ClickHouse | 450,289,044 | 3,927 | wenxingxing | solved, thanks! | 2018-12-28T04:44:56 | 2018-12-28T04:44:56 | {} | NONE |
ClickHouse | ClickHouse | 449,569,471 | 3,764 | viongpanzi | @shangshujie365 thx~ | 2018-12-22T13:16:46 | 2018-12-22T13:16:46 | {} | NONE |
ClickHouse | ClickHouse | 449,732,888 | 3,870 | alexey-milovidov | The business logic is broken.
```
SELECT toInt8(CounterID)
FROM test.hits
WHERE CounterID = 23192098
┌─toInt8(CounterID)─┐
│ 34 │
│ 34 │
│ 34 │
└───────────────────┘
SELECT CounterID
FROM test.hits
WHERE (CounterID = 23192098) AND (toInt8(CounterID) = 3... | 2018-12-24T13:13:24 | 2018-12-24T13:13:24 | {} | MEMBER |
ClickHouse | ClickHouse | 449,957,179 | 3,882 | shurshun | Hi!
I have the similar problem with Clickhouse cluster in Kubernetes.
Deployment: 3 shards with replica factor 2, and 1 separate node with "aggregator"
Version: 18.14.17
So, when one of 6 CH pod goes down, I catch error like this:
> Poco::Exception. Code: 1000, e.code() = 0, e.displayText() = Host not found: clic... | 2018-12-26T12:05:14 | 2018-12-26T12:05:14 | {} | NONE |
ClickHouse | ClickHouse | 449,979,914 | 2,369 | jvmlet | Hello
Any progress with protobuf messages support ? DWH that natively supports storing/querying protobuf messages is the most demanding feature ;-) | 2018-12-26T15:15:02 | 2018-12-26T15:15:02 | {} | NONE |
ClickHouse | ClickHouse | 449,734,915 | 3,912 | yonglezhu | yes，altinity' repo works.
another question，if i compiling from src,can I set some parameter like 'prefix' to put all files to single directory? | 2018-12-24T13:33:03 | 2018-12-24T13:33:03 | {} | NONE |
ClickHouse | ClickHouse | 449,744,628 | 3,716 | tishkinis | 18.16.0, #3716 fixed | 2018-12-24T15:09:58 | 2018-12-24T15:09:58 | {} | NONE |
ClickHouse | ClickHouse | 449,751,674 | 3,912 | alesapin | Yes, just follow the doc https://clickhouse.yandex/docs/en/development/build/#build-clickhouse. | 2018-12-24T16:21:33 | 2018-12-24T16:21:42 | {} | MEMBER |
ClickHouse | ClickHouse | 449,835,643 | 3,899 | alesapin | It's not 100% ready, I still can do some refactoring and optimizations. However I'd be glad to see some review notices. | 2018-12-25T10:10:20 | 2018-12-25T10:10:20 | {} | MEMBER |
ClickHouse | ClickHouse | 449,867,692 | 3,870 | alexey-milovidov | #3921 | 2018-12-25T18:44:42 | 2018-12-25T18:44:42 | {} | MEMBER |
ClickHouse | ClickHouse | 450,020,255 | 3,889 | den-crane | I mentioned <=> , because it's almost standard SQL operator.
https://dev.mysql.com/doc/refman/8.0/en/comparison-operators.html#operator_equal-to
https://www.vertica.com/docs/9.2.x/HTML/Content/Authoring/SQLReferenceManual/LanguageElements/Operators/ComparisonOperators.htm | 2018-12-26T20:20:41 | 2018-12-26T20:21:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,095,941 | 3,931 | simPod | @blinkov yes, in [Postgres](https://wiki.postgresql.org/wiki/Working_with_Dates_and_Times_in_PostgreSQL#3._You_may_add_or_subtract_two_INTERVALS). Also [Oracle](https://oracle-base.com/articles/misc/oracle-dates-timestamps-and-intervals#interval) but that is slightly different. | 2018-12-27T07:42:26 | 2018-12-27T07:42:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,731,910 | 3,910 | alexey-milovidov | Do you have a test case that fails under ASan or UBSan? | 2018-12-24T13:04:32 | 2018-12-24T13:04:32 | {} | MEMBER |
ClickHouse | ClickHouse | 449,970,693 | 3,927 | den-crane | actually table functions are case sensitive
SELECT *
FROM MySQL('x', 'x', 'x', 'x', 'x')
Received exception from server (version 18.14.18):
Code: 46. DB::Exception: Received from localhost:9000, ::1. DB::Exception: Unknown table function MySQL.
SELECT *
FROM mysql('x', 'x', 'x', 'x', 'x')
Received excepti... | 2018-12-26T14:01:54 | 2018-12-26T14:01:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,980,526 | 3,931 | alesapin | Yes, this way is also supported:
```
:) select now() + INTERVAL 1 day
SELECT now() + toIntervalDay(1)
┌─plus(now(), toIntervalDay(1))─┐
│ 2018-12-27 18:18:16 │
└───────────────────────────────┘
1 rows in set. Elapsed: 0.006 sec.
```
| 2018-12-26T15:19:35 | 2018-12-26T15:19:35 | {} | MEMBER |
ClickHouse | ClickHouse | 449,980,643 | 3,931 | alexey-milovidov | Note that `now() + 60 * 60 * 24` does not respect clock adjustments (example: daylight saving time).
And the `INTERVAL` syntax or `addDays` function is always preferred.
| 2018-12-26T15:20:32 | 2018-12-26T15:20:32 | {} | MEMBER |
ClickHouse | ClickHouse | 450,128,016 | 3,929 | zhang2014 | > I have doc and some tests to not complete. Because I'm not sure if 3 and 4 are popular, They do a lot of implicit work
I've already done that and can I block some tests in unbundled functional test?
Because it doesn't seem to compile MySQL engine.
| 2018-12-27T10:50:12 | 2018-12-27T10:55:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,270,238 | 3,927 | wenxingxing | I'll rebuild it
| 2018-12-28T01:46:01 | 2018-12-28T01:46:01 | {} | NONE |
ClickHouse | ClickHouse | 449,433,353 | 3,782 | simPod | @blinkov no, just have read it all from top and it seems crystal clear. Thanks | 2018-12-21T16:26:35 | 2018-12-21T16:26:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,473,093 | 3,901 | Jack012a | Hello den-crane,
> 2>CH unable to find max(timestamp) from index so it will scan all records.
`ENGINE = MergeTree PARTITION BY date ORDER BY (date, symbol, timestamp) SETTINGS index_granularity = 8192
`
The table is created by the index of (date, symbol, timestamp).
In this case, is it still true that Clickhou... | 2018-12-21T19:08:20 | 2018-12-21T19:08:20 | {} | NONE |
ClickHouse | ClickHouse | 450,102,978 | 3,931 | simPod | tru tru. My preference is Postgres, it's more intuitive | 2018-12-27T08:28:35 | 2018-12-27T08:28:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,359,308 | 2,586 | zhang2014 | @alesapin Hopefully https://github.com/yandex/ClickHouse/pull/2586#issuecomment-446837620 didn't mislead you (In fact, what I want to express is that these unfinished plans will be helpful for your subsequent plans, for example, to help you understand my thoughts) 😄
In order not to mislead you, I will continue to... | 2018-12-21T11:04:33 | 2018-12-22T02:13:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,610,188 | 2,675 | alexey-milovidov | The settings input_format_* are for user profiles. For Kafka tables, context with settings is initialized at table creation (at server startup). So, these settings might be applied only if they are set in default profile. Can you please check if they work if you set them in default profile and restart the server?
To... | 2018-12-23T02:35:29 | 2018-12-23T02:35:29 | {} | MEMBER |
ClickHouse | ClickHouse | 449,731,874 | 3,910 | alexey-milovidov | Ok. | 2018-12-24T13:04:11 | 2018-12-24T13:04:11 | {} | MEMBER |
ClickHouse | ClickHouse | 449,949,890 | 3,929 | zhang2014 | I have doc and some tests to not complete. Because I'm not sure if 3 and 4 are popular, They do a lot of implicit work | 2018-12-26T10:59:43 | 2018-12-26T10:59:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,966,987 | 3,198 | turbotankist | But still there is problem with netlink file descriptors open
```
> ps aux | grep clickhouse
clickho+ 23434 12.9 11.8 7978952 967100 ? Ssl Dec25 184:24 /usr/bin/clickhouse-server --config=/etc/clickhouse-server/config.xml
> sudo lsof -p 23434 | wc -l
184505
```
The number is constantly growing!
And upo... | 2018-12-26T13:30:32 | 2018-12-26T13:30:32 | {} | NONE |
ClickHouse | ClickHouse | 449,971,964 | 3,927 | den-crane | then check mysql is supported by your CH server
select * from system.build_options where name like 'USE_MYSQL'
```
┌─name──────┬─value─┐
│ USE_MYSQL │ 1 │
└───────────┴───────┘
```
mysqllib is statically linked to CH binary. | 2018-12-26T14:12:16 | 2018-12-26T14:12:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,980,285 | 2,369 | alexey-milovidov | This task is reassigned and scheduled for January.
> DWH that natively supports storing/querying protobuf messages is the most demanding feature ;-)
It will allow to import/export/store subset of Protobuf messages. Nested messages will be limited to Arrays. | 2018-12-26T15:17:55 | 2018-12-26T15:17:55 | {} | MEMBER |
ClickHouse | ClickHouse | 449,999,031 | 1,761 | SaltTan | Unfortunately it is not fully supported.
If a dictionary is described in a separate file, and that file gets deleted, the dictionary will not be removed. | 2018-12-26T17:38:33 | 2018-12-26T17:38:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,021,781 | 3,933 | den-crane | sorry, maybe you have a confusion with arrayStringConcat vs arrayJoin?
SELECT arrayMap(x -> arrayStringConcat([x, 'a']), ['b', 'c', 'd'])
SELECT arrayStringConcat(['b', 'c', 'd'], 'a')
SELECT toString(arrayJoin(arrayMap(x -> arrayConcat([x], ['a']), ['b', 'c', 'd']))) | 2018-12-26T20:31:55 | 2018-12-26T20:32:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,093,197 | 3,931 | simPod | Would be useful as `1 day 3 seconds` etc | 2018-12-27T07:24:17 | 2018-12-27T07:24:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,160,960 | 3,908 | proller | d9bc934b7d808878f06e37ed80c17d6efeca86f3 | 2018-12-27T14:32:52 | 2018-12-27T14:32:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,488,103 | 3,901 | Jack012a | I will make a change based on your suggestion.
Thank you. | 2018-12-21T20:15:55 | 2018-12-21T20:15:55 | {} | NONE |
ClickHouse | ClickHouse | 449,719,726 | 3,917 | alesapin | This message mean, that table `new_name` already exists or some files for this table persist on disk.
- In first case you have to remove table `new_name` with query `DROP TABLE db_name.new_name`. `db_name` may be omitted if table stored in `default` database.
- In other case you have to remove files for this tabl... | 2018-12-24T11:03:25 | 2018-12-24T11:03:25 | {} | MEMBER |
ClickHouse | ClickHouse | 449,732,091 | 3,870 | alexey-milovidov | The test fails constantly in debug build. | 2018-12-24T13:06:18 | 2018-12-24T13:06:18 | {} | MEMBER |
ClickHouse | ClickHouse | 449,980,982 | 3,931 | simPod | Ah nice! But that isn't documented yet, right? Or did I just overlook? https://clickhouse.yandex/docs/en/query_language/functions/date_time_functions/ | 2018-12-26T15:23:13 | 2018-12-26T15:23:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,434,359 | 3,838 | ztlpn | fixed in v18.16.1 | 2018-12-21T16:30:23 | 2018-12-21T16:30:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,487,351 | 3,901 | den-crane | And BTW, you can use argMax & tuple if you need only 1 row
```
SELECT argMax(('MyTable', date, date_time, timestamp, symbol), timestamp)
FROM default.MyTable
WHERE date =
(
SELECT max(max_date) FROM system.parts
WHERE (table = 'MyTable') AND (database = 'default')
)
``` | 2018-12-21T20:11:47 | 2018-12-21T20:11:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,743,558 | 3,918 | alexey-milovidov | Testing with set of strings with constant length is non-representative, because it makes branch predictor too happy. Better to test on real dataset with different distribution of length of strings.
Numbers in MB/sec looks like MB of source data per second, where source data is something like system.numbers table. It... | 2018-12-24T14:59:58 | 2018-12-24T15:50:02 | {} | MEMBER |
ClickHouse | ClickHouse | 449,782,623 | 3,870 | zhang2014 | > The business logic is broken.
Might the be added to the test case?
> You are reading uninitialized data.
Into a Null value when convert RNP? the logic here is not very clear to me. | 2018-12-25T01:09:07 | 2018-12-25T01:09:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,098,755 | 3,931 | blinkov | @simPod thanks. Actually, they look significantly different and incompatible with each other. | 2018-12-27T08:01:20 | 2018-12-27T08:01:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,154,718 | 3,939 | TH-HA | Hello
We had a similar issue where queries were slow without resource usage (CPU and IO ). In our context the issue was related to a bad partitioning schema leading to too many files on nodes. In this case CH was spending all the time opening/closing files ( and i 'm just guessing but the file system was saturated). ... | 2018-12-27T13:54:42 | 2018-12-27T13:54:42 | {} | NONE |
ClickHouse | ClickHouse | 449,427,871 | 3,782 | blinkov | @simPod do you have any further questions? | 2018-12-21T16:05:49 | 2018-12-21T16:05:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,867,595 | 3,870 | alexey-milovidov | > Into a Null value when convert RNP? the logic here is not very clear to me.
It's about `field.get<Int64>()` when field can be NULL.
Probably this error was already present.
I've rewriting it.
> Might the be added to the test case?
This was found by a test case in internal ("stateful") part of tests.
| 2018-12-25T18:42:08 | 2018-12-25T18:42:08 | {} | MEMBER |
ClickHouse | ClickHouse | 449,869,381 | 2,898 | KochetovNicolai | Moved to #3925 | 2018-12-25T19:25:16 | 2018-12-25T19:25:16 | {} | MEMBER |
ClickHouse | ClickHouse | 449,952,980 | 3,928 | alesapin | Related code https://github.com/yandex/ClickHouse/pull/2609/files#diff-6d7b8a71e9cf645fa1e86f3e40f7ff25 | 2018-12-26T11:26:54 | 2018-12-26T11:26:54 | {} | MEMBER |
ClickHouse | ClickHouse | 450,027,600 | 3,933 | alexey-milovidov | Confirmed.
You cannot use arrayJoin inside higher order functions.
`SELECT arrayMap(x -> arrayJoin([x, 1]), [1, 2])` | 2018-12-26T21:17:43 | 2018-12-26T21:17:43 | {} | MEMBER |
ClickHouse | ClickHouse | 450,233,675 | 3,948 | alexey-milovidov | Now all functional tests have passed with UBSan. | 2018-12-27T21:23:41 | 2018-12-27T21:23:41 | {} | MEMBER |
ClickHouse | ClickHouse | 449,412,640 | 3,893 | blinkov | @sliontc What is this client you are trying to use? Looks like Eclipse? If yes, it's probably using JDBC driver so you should check it's settings. | 2018-12-21T15:09:55 | 2018-12-21T15:09:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,425,662 | 3,902 | blinkov | Probably there's some requirement on `wkhtmltopdf` version or it's dependencies, but I haven't researched this in detail. The option it complains about is important to get the PDF version right. At the moment PDF version can be disabled only together with normal single-page version, but it can be improved. | 2018-12-21T15:57:27 | 2018-12-21T15:57:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,442,241 | 3,901 | den-crane | 1. It's completely wrong task for CH. CH is not an OLTP database.
Just save somehow the last inserted time in additional table from client.
2. CH unable to find max(timestamp) from index so it will scan all records.
3. maybe you missed date = ... predicate in the first query.
```
SELECT tx,...,
FROM MyTabl... | 2018-12-21T17:01:23 | 2018-12-21T17:07:46 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,695,200 | 3,780 | ghost | @blinkov No, thanks for answering my questions. | 2018-12-24T07:07:43 | 2018-12-24T07:07:43 | {} | NONE |
ClickHouse | ClickHouse | 449,720,855 | 3,889 | alesapin | `<=>` may be semantically confused with ["spaceship"](https://en.wikipedia.org/wiki/Three-way_comparison) operator from C++20.
`equalsOrNulls` may be good idea. | 2018-12-24T11:14:22 | 2018-12-24T11:14:22 | {} | MEMBER |
ClickHouse | ClickHouse | 449,726,137 | 3,918 | blinkov | This table could have been published in docs, but at the moment it lacks the description at least about used hardware. | 2018-12-24T12:06:26 | 2018-12-24T12:06:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,755,978 | 3,604 | abyss7 | There are several problems with RIGHT JOINs right now. The better solution will be available after internal ticket CLICKHOUSE-4139 gets closed. | 2018-12-24T17:10:14 | 2018-12-24T17:10:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,643,892 | 3,818 | lamberken | @blinkov , thanks very match. I'll your suggestion | 2018-12-23T15:32:08 | 2018-12-23T15:32:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,721,041 | 3,882 | alesapin | This is not the first report on this issue. I will try to look at this problem. | 2018-12-24T11:16:06 | 2018-12-24T11:16:06 | {} | MEMBER |
ClickHouse | ClickHouse | 449,740,243 | 3,918 | filimonov | > This table could have been published in docs, but at the moment it lacks the description at least about used hardware.
I was checking it on quite old workstation. May be it would be better to check it on better cpu, and also with some extra metrics.
Also results are quite confusing, cause i was expecting to get... | 2018-12-24T14:26:35 | 2018-12-24T14:45:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,929,426 | 3,877 | qietingfengling | I think I understand it,thanks | 2018-12-26T08:33:19 | 2018-12-26T08:33:19 | {
"hooray": 2
} | NONE |
ClickHouse | ClickHouse | 449,981,727 | 2,369 | jvmlet | Thanks for the quick response
> Nested messages will be limited to Arrays.
Does it mean that only `repeated` fields are going to be supported ? Or nested field will be exposed as array of size `1` ? | 2018-12-26T15:28:43 | 2018-12-26T15:29:17 | {} | NONE |
ClickHouse | ClickHouse | 450,148,082 | 3,892 | proller | /var/lib/clickhouse/preprocessed_configs now contains all preprocessed config including from /dictionaries/ dir.
All '/' in path relative to etc/clickhouse-server/ changed to '_' (to avoid too deep tree in preprocessed dir)
Now you can delete all *dictionary-preprocessed.xml in /etc/clickhouse-server/dictionarie... | 2018-12-27T13:11:27 | 2018-12-27T13:11:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 449,639,196 | 3,910 | alexey-milovidov | What is the mechanics of this change?
Why we can capture references in LazyBlockInputStream? | 2018-12-23T14:17:00 | 2018-12-23T14:17:00 | {} | MEMBER |
ClickHouse | ClickHouse | 449,883,513 | 2,401 | alexey-milovidov | Это полностью корректно.
Ведь формат TabSeparated требует экранирования, а JSON тоже требует экранирование внутри строковых литералов. Получается, что для корректной вставки в формате TabSeparated строки, содержащей JSON, содержащий строковый литерал, содержащий двойную кавычку, вам потребуется два бэкслеша. | 2018-12-26T01:11:16 | 2018-12-26T01:11:16 | {} | MEMBER |
ClickHouse | ClickHouse | 449,932,051 | 3,927 | blinkov | <strike>Try `mysql` in lower case.</strike> does it work without create table? | 2018-12-26T08:51:00 | 2018-12-26T08:59:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,040,321 | 3,931 | alexey-milovidov | @blinkov Should we also support interval description in a string literal like '1 day'? | 2018-12-26T23:22:57 | 2018-12-26T23:22:57 | {} | MEMBER |
ClickHouse | ClickHouse | 450,267,345 | 3,927 | wenxingxing | looks like mysqllib is not built with clickhouse
```
SELECT *
FROM system.build_options
WHERE name LIKE 'USE_MYSQL'
┌─name──────┬─value─┐
│ USE_MYSQL │ │
└───────────┴───────┘
``` | 2018-12-28T01:20:28 | 2018-12-28T01:20:28 | {} | NONE |
ClickHouse | ClickHouse | 450,267,683 | 3,927 | wenxingxing | ```
SELECT *
FROM system.build_options
┌─name──────────────────────┬─value───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┐
│ VERSION_FULL │ ClickHouse 18.14.18 ... | 2018-12-28T01:23:11 | 2018-12-28T01:23:11 | {} | NONE |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.