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 | 554,992,201 | 7,750 | akuzm | https://github.com/ClickHouse/ClickHouse/commit/7084d090e805f76c660d14e1a7af162f6951fbd5
Style check was broken, fixed it. | 2019-11-18T12:20:12 | 2019-11-18T12:20:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,029,811 | 7,811 | korjavin | Not sure if I dumped correctly to the file.
But i've gotCode: 62. DB::Exception: Syntax error: failed at position 1: ...... Unrecognized token
| 2019-11-18T14:07:37 | 2019-11-18T14:07:37 | {} | NONE |
ClickHouse | ClickHouse | 555,031,530 | 7,819 | den-crane | It's an outdated documentation.
Read it as:
INSERT sorts the input data by primary key and splits them into partitions by **partition key expression**. If you insert data for mixed **partition expression (evaluated to several values)**, it can significantly reduce the performance of the INSERT query. | 2019-11-18T14:12:01 | 2019-11-18T14:12:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,271,491 | 6,749 | den-crane | I am also wondering.
I think it is not atomical.
https://github.com/ClickHouse/ClickHouse/blob/master/dbms/src/Storages/StorageMergeTree.cpp#L1160
Anything could happen and partitions will stay not removed. | 2019-11-19T00:18:43 | 2019-11-19T00:21:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,320,644 | 7,830 | den-crane | Software developers who like strong typing. | 2019-11-19T03:50:05 | 2019-11-19T03:50:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,388,480 | 7,537 | chijiajian | @filimonov The Version of mysql is 5.7.2. I find out the problem.
If using `set global session_track_state_change = 1;` setting on mysql, It throw this exception every time.
And If using `set global session_track_state_change = 0;` mysql engine is working well.
thx.
| 2019-11-19T08:23:30 | 2019-11-19T08:23:30 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 555,594,055 | 7,781 | m1khal3v | Have the same problem on 19.16. 19.4 worked properly
edit: I want to note that the error occurs on a single request with DISTINCT, the rest of the requests with JOIN work fine
edit2: without DISTINCT have the same error :( | 2019-11-19T16:37:18 | 2019-11-19T17:02:27 | {} | NONE |
ClickHouse | ClickHouse | 555,666,335 | 7,850 | blinkov | @yigitkonur fyi: some of these changes are just switching between American and British styles of English :) | 2019-11-19T19:18:44 | 2019-11-19T19:18:44 | {
"+1": 1,
"laugh": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,846,097 | 7,848 | nauu | @sundy-li thats so useful , thx. | 2019-11-20T05:34:37 | 2019-11-20T06:29:00 | {} | MEMBER |
ClickHouse | ClickHouse | 555,861,868 | 7,856 | aa8808021 | I had the same problem | 2019-11-20T06:37:38 | 2019-11-20T06:37:38 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 554,677,611 | 7,790 | filimonov | Is it possible to create a test case? | 2019-11-16T21:50:18 | 2019-11-16T21:50:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,783,844 | 7,790 | alexey-milovidov | @filimonov Yes, it is possible and is required. | 2019-11-17T20:35:35 | 2019-11-17T20:35:52 | {} | MEMBER |
ClickHouse | ClickHouse | 554,965,007 | 7,653 | vitlibar | > > It seems the client will be able to calculate that `can_ignore` flag without extra settings.
>
> Does it mean that the decision allowing / forbidding ignorance will be hardcoded?
> What if i prefer not to ignore that?
The names and meanings of the settings are hardcoded, so it seems the ignorance flag should... | 2019-11-18T10:58:44 | 2019-11-18T10:58:44 | {} | MEMBER |
ClickHouse | ClickHouse | 554,998,160 | 7,780 | akuzm | There is `ReadBuffer::readBig` that may be suitable, but it's not implemented for plain files, only for compressed ones. Should be easy to do. | 2019-11-18T12:38:51 | 2019-11-18T12:38:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,524,484 | 5,682 | den-crane | still broken | 2019-11-19T14:11:43 | 2019-11-19T14:11:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,810,099 | 7,855 | den-crane | There is no type boolean in CH. UINT8 is just a substitute for this case. TRUE is any value UINT8 except zero.
WHERE 1 === true (UINT8)
WHERE 'a'=='a' === true (UINT8) | 2019-11-20T02:42:43 | 2019-11-20T02:42:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,915,120 | 7,261 | simpl1g | Do you have any updates on this issue? | 2019-11-18T08:50:11 | 2019-11-18T08:50:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,013,892 | 6,847 | CurtizJ | Failed tests with memory sanitizer are ok, because they were inroduced only in 19.15. | 2019-11-18T13:25:32 | 2019-11-18T13:25:32 | {} | MEMBER |
ClickHouse | ClickHouse | 555,064,692 | 7,817 | den-crane | Not related to HTTP.
cat f.values |clickhouse-client -q "INSERT INTO strange_bug_val FORMAT Values" | 2019-11-18T15:26:04 | 2019-11-18T15:26:04 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,097,208 | 7,820 | den-crane | 19.17.2.4
```
CREATE DICTIONARY city_map
(
`city_map_id` UInt32,
`city_name` String
)
PRIMARY KEY city_map_id
SOURCE(CLICKHOUSE(HOST 'localhost' PORT 9011 USER 'default' PASSWORD '' DB 'default' TABLE 'city_map'))
LIFETIME(MIN 1 MAX 10)
LAYOUT(HASHED())
```
Seems `IF NOT EXISTS` is not supported | 2019-11-18T16:33:29 | 2019-11-18T16:34:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,184,795 | 7,780 | akuzm | > There is `ReadBuffer::readBig` that may be suitable, but it's not implemented for plain files, only for compressed ones. Should be easy to do.
Turned out to be not easy at all -- being basically an unbuffered read (`ReadBuffer::readUnbuffered`..?) , it has unclear interplay with `nextImpl()`. This probably require... | 2019-11-18T20:00:54 | 2019-11-18T20:02:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,318,360 | 7,830 | den-crane | We are not humans. | 2019-11-19T03:38:48 | 2019-11-19T03:38:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,323,386 | 7,807 | den-crane | No difference. Except early versions of CH did not support `drop partition` for MV. I assumed that you created this issue because you are experiencing some issues with `drop partition`. | 2019-11-19T04:05:29 | 2019-11-19T04:09:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,517,122 | 7,646 | abyss7 | According to your logs the issue is about too heavy message load - CH can't keep up with incoming messages between poll intervals, and consumers get dropped. Also they require time to reconnect again later. You can try to tune the `stream_poll_timeout_ms` setting to a greater value. | 2019-11-19T13:53:17 | 2019-11-19T13:53:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,582,180 | 7,840 | nikitamikhaylov | You have only 5/15 successful builds. Tap on *details* button and read the build logs. | 2019-11-19T16:16:37 | 2019-11-19T16:16:37 | {} | MEMBER |
ClickHouse | ClickHouse | 555,321,118 | 7,813 | Tasselmi | it seems that you can't create a view upon a MATERIALIZED view.
I try to create MATERIALIZED view upon a MATERIALIZED view, but the upper one never updates. | 2019-11-19T03:52:44 | 2019-11-19T03:52:44 | {} | NONE |
ClickHouse | ClickHouse | 555,473,420 | 7,816 | akuzm | > Strange to see that `multAsync` failed, even if `zkutil` is untouched and up to date with `master`.
Probably not related to your changes, it's a known flappy test. | 2019-11-19T11:50:07 | 2019-11-19T11:50:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,521,368 | 7,799 | SaltTan | I think the issue is more complicated.
This query fails as well:
```
select count()
FROM system.query_log
ANY LEFT JOIN system.query_log USING port
PREWHERE type in (1)
WHERE event_date = today()
```
And this query:
```
select count()
FROM system.query_log
ANY LEFT JOIN system.query_thread_log USING qu... | 2019-11-19T14:03:46 | 2019-11-19T14:03:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,653,889 | 7,804 | den-crane | But `toDateTime(intDiv(toUInt32(ts),7200)*7200) as ts ...... ts BETWEEN `
does not give same as result `WHERE flows3.ts BETWEEN `
| 2019-11-16T16:48:57 | 2019-11-16T16:49:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,980,154 | 7,780 | nikitamikhaylov | But we can disable this feature by default or mark it as experimental. | 2019-11-18T11:42:02 | 2019-11-18T11:42:02 | {} | MEMBER |
ClickHouse | ClickHouse | 555,024,304 | 7,811 | filimonov | Please dump one message to a file (kafkacat or kafka-console-consumer).
Try to parse that file/insert into clickhouse with simple query. We need to know if format itself is correct / can be parsed.
```
cat protobuf.file | clickhouse-client 'INSERT INTO sometable FORMAT Protobuf SETTINGS format_schema = 'md:mddata... | 2019-11-18T13:53:36 | 2019-11-18T13:56:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,070,963 | 7,817 | den-crane | ```
CREATE TABLE strange_bug_val
( a Float32,
b Float32)
ENGINE = Memory;
clickhouse-client -q "select '(-160.32605134916085,37.70584056842162),' from numbers(1000000)" >test.values
cat test.values |clickhouse-client -q "INSERT INTO strange_bug_val FORMAT Values"
SELECT DISTINCT *
FROM strange_bug_va... | 2019-11-18T15:39:13 | 2019-11-18T15:48:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,100,320 | 7,820 | blitzarx1 | @den-crane yeah, I got the same result. Looks like it doesn't support IF NOT EXISTS clause | 2019-11-18T16:40:17 | 2019-11-18T16:40:50 | {} | NONE |
ClickHouse | ClickHouse | 555,269,929 | 7,811 | kap6 | I deleted the topic and started over. The issue is now resolved.
Thanks. | 2019-11-19T00:12:17 | 2019-11-19T00:12:17 | {} | NONE |
ClickHouse | ClickHouse | 555,402,142 | 7,832 | filimonov | 1. yes, but w/o zookeeper you have no reliable failover.
2. yes. You should create tables on each server separately, best practice is to create exactly the same set of tables on every server. You can have multiple cluster configurations, so you can create table A (using sharded cluster config) and table B (using clus... | 2019-11-19T09:00:43 | 2019-11-19T09:01:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,533,226 | 7,290 | abyss7 | Duplicated by #7417 | 2019-11-19T14:30:44 | 2019-11-19T14:30:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,677,859 | 7,757 | filimonov | @Tasselmi just run clickhouse-client with `-m` switch, and it start executing only after you finish the line with semicolon. | 2019-11-16T21:54:08 | 2019-11-16T21:54:08 | {
"+1": 13
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,757,139 | 7,804 | dotNox | I get the exact results ( same data, same number of output records ). The only thing that changes is how much data is read from storage.
```
# time clickhouse-client -n -d nfa -q "SELECT toDateTime(intDiv(toUInt32(ts),7200)*7200) as ts, sum(octets) FROM flows3 WHERE ts BETWEEN '2019-11-01 00:00:00' AND '2019-11-18 00... | 2019-11-17T15:32:38 | 2019-11-17T15:33:30 | {} | NONE |
ClickHouse | ClickHouse | 554,794,326 | 7,727 | alexey-milovidov | Ok but please add a functional test with https://clickhouse.yandex/docs/en/query_language/create/#create-dictionary-query | 2019-11-17T22:24:17 | 2019-11-17T22:24:17 | {} | MEMBER |
ClickHouse | ClickHouse | 555,102,472 | 7,817 | den-crane | Reproduced with 19.18.1.1667
No issue with 19.15.5.18
| 2019-11-18T16:45:00 | 2019-11-18T16:56:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,314,193 | 7,830 | Tasselmi | > Because CH has a strong typing.
>
> select * from (select toInt8(1) a) t join (select toUInt8(1) b) t1 on a=b;
> DB::Exception: Type mismatch of columns to JOIN by: a Int8 at left, b UInt8 at right.
>
> select * from (select toInt8(1) a) t join (select toUInt8(1) b) t1 on **toUInt8**(a)=b
> ┌─a─┬─b─┐
> │ 1 │... | 2019-11-19T03:19:04 | 2019-11-19T03:19:04 | {} | NONE |
ClickHouse | ClickHouse | 555,443,449 | 7,511 | nikitamikhaylov | Thank you for this PR! But there are many unrelated changes such as reordering in includes and fixing codestyle mistakes. Please, create **another** PR with only insignificant changes (refactoring). It will be easier for me to read and for you to debug. Also there are many problems with tests. If you have any mistakes,... | 2019-11-19T10:34:19 | 2019-11-19T10:34:19 | {} | MEMBER |
ClickHouse | ClickHouse | 555,539,524 | 7,780 | akuzm | ```
┌────chunk─┬─median time─┐
│ 1000 │ 294.10 │
│ 10000 │ 54.37 │
│ 100000 │ 14.14 │
│ 1000000 │ 9.47 │
│ 10000000 │ 8.61 │
└──────────┴─────────────┘
```
10MB chunk is even faster than 1MB
(hits_v1.tsv, 6 threads total)
Another run:
```
┌─────chunk─┬─median ti... | 2019-11-19T14:45:03 | 2019-11-19T16:29:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,771,974 | 2,415 | viktortnk | In case it will be useful for someone, there is `arrayFlatten` function, so for above query it would be:
```sql
SELECT
[['query1'], ['query2']] AS queries,
[['hit1', 'hit2'], ['hit1']] AS hits,
arrayFlatten(queries) AS queries_flat,
arrayFlatten(hits) AS hits_flat,
arrayMap(x -> length(x), hi... | 2019-11-17T18:20:09 | 2019-11-17T18:20:09 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,799,476 | 7,811 | filimonov | Please try version 19.15 or above. | 2019-11-17T23:19:37 | 2019-11-17T23:19:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,961,996 | 7,565 | ghost | I forcely committed changes to resolve confliction with latest master branch and cleaned up commits by rebasing.
* Rebased commits:
* Implemented basic LRU-Query-Cache.
* iImplemented 'DROP QUERY_CACHE' to delete all query_caches. | 2019-11-18T10:49:57 | 2019-11-18T10:50:26 | {} | NONE |
ClickHouse | ClickHouse | 555,035,605 | 7,814 | doom369 | @tavplubix thanks. Yes, I read the documentation. Still I think it is not consistent. From my point of view all json formats should return the additional meta or all json format should not return the additional data. Right now it is a mix. So I think it should be fixed. It is not big deal. Just want to point this out.... | 2019-11-18T14:21:29 | 2019-11-18T14:23:47 | {} | NONE |
ClickHouse | ClickHouse | 555,143,030 | 7,817 | den-crane | BTW, about CSV, the cause is FLOAT32 vs FLOAT64
```
┌─plus(row_id, 1)─┬─longitude─┬──latitude─┬─computed_geohash─┬─geohash─┐
│ 360307 │ -171.5625 │ -81.95313 │ 01q │ 01m │
└─────────────────┴───────────┴───────────┴──────────────────┴─────────┘
select geohashEncode(cast(-171.562505373... | 2019-11-18T18:20:03 | 2019-11-18T18:20:23 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,642,152 | 7,781 | m1khal3v | The problem is solved by adding aliases for all duplicate column names, e.g.:
```
SELECT
a.col1 AS col1
FROM
`default`.a
LEFT JOIN (SELECT col1 AS "col1_from_b" FROM `default`.b) ON a.col1 = b.col1_from_b
LEFT JOIN (SELECT col1 AS "col1_from_c" FROM `default`.c) ON a.col1 = `c.col1_from_c
``` | 2019-11-19T18:23:38 | 2019-11-19T18:23:38 | {
"heart": 1
} | NONE |
ClickHouse | ClickHouse | 555,717,388 | 7,838 | filimonov | gcc-9 failure - looks the same as https://github.com/ClickHouse/ClickHouse/issues/7114
It seems like the reason is pthread is not linked into protoc directly, and `protoc` works only when one of the libraries it links (for example libunwind) requires pthread.
That dirty hack fixes the gcc-9 build:
```diff
dif... | 2019-11-19T21:13:21 | 2019-11-19T21:37:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,972,579 | 7,733 | maskshell | I have the same problem, but I can not confirm the conditions is totally same.
Not sure, but check this issue may have some help:
#3953
| 2019-11-18T11:20:14 | 2019-11-18T11:20:14 | {} | NONE |
ClickHouse | ClickHouse | 555,210,331 | 7,811 | kap6 | Are you referring to logs in /var/log/clickhouse-server ?
Yes. I did. I don't see any thing there.
Here is my command line to inject data into kafka queue. I am not sure if i am doing this the right way:
bin/kafka-console-producer --broker-list localhost:9092 --topic binary5 < ~/orig/person.bin
i tested fur... | 2019-11-18T21:10:24 | 2019-11-18T21:10:24 | {} | NONE |
ClickHouse | ClickHouse | 555,347,004 | 7,831 | zhang2014 | LGTM | 2019-11-19T05:59:43 | 2019-11-19T05:59:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,418,415 | 7,816 | dingxiangfei2009 | Strange to see that `multAsync` failed, even if `zkutil` is untouched and up to date with `master`. | 2019-11-19T09:39:44 | 2019-11-19T09:39:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,597,785 | 7,114 | filimonov | ```
cd /home/mfilimonov/workspace/ClickHouse/build-vscode/Debug/contrib/libhdfs3-cmake && gdb --args /home/mfilimonov/workspace/ClickHouse/build-vscode/Debug/contrib/protobuf/cmake/protoc --cpp_out /home/mfilimonov/workspace/ClickHouse/build-vscode/Debug/contrib/libhdfs3-cmake -I /home/mfilimonov/workspace/ClickHouse/... | 2019-11-19T16:45:42 | 2019-11-19T16:45:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,801,007 | 7,811 | den-crane | Did you try to load your data with insert (using clickhouse-client) ?
> Please note that Protobuf stream is length-delimited: every message is prefixed by its length in varint..
Do you supply message length? | 2019-11-17T23:36:52 | 2019-11-17T23:36:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,313,785 | 1,952 | honmaple | This issue still happen on `ClickHouse server version 19.15.2.2 (official build).`
How can I fix it?
```
2019.11.19 11:08:30.515068 [ 20 ] {} <Warning> logtest.test_local: Part 318_26055_30571_5 is covered by 318_26055_30571_6 but should be merged into 318_26055_32627_46. This shouldn't happen often.
2019.11.19 11... | 2019-11-19T03:17:16 | 2019-11-19T03:17:16 | {} | NONE |
ClickHouse | ClickHouse | 555,315,062 | 7,816 | dingxiangfei2009 | Sorry, I will let the functional tests finish before patching the style problems. | 2019-11-19T03:23:15 | 2019-11-19T03:23:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,405,960 | 7,835 | filimonov | Not sure if needed. Maybe cmake cache cleanup was enough.
See also:
https://github.com/ClickHouse/ClickHouse/blob/15cb620f422e8cac76a790b3fb75cb2fa93c65c8/cmake/find/poco.cmake#L81-L84
https://github.com/ClickHouse/ClickHouse/blob/6e28c22876578ac2b375d999d695bb1382b76e7e/dbms/src/Dictionaries/CMakeLists.txt#L39-... | 2019-11-19T09:10:21 | 2019-11-19T09:15:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,635,518 | 7,849 | den-crane | >I understand why this error can occur in small inserts by hand
No. It's not about this.
It's about partitioning key in your MV. One insert into kafka (into MV) tries to create more than 100 partitions.
For example
```
create table A( D Date) Engine=MergeTree partition by D order by D;
insert into A select... | 2019-11-19T18:10:48 | 2019-11-19T18:11:25 | {
"rocket": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,219,040 | 7,804 | den-crane | Correct. Very useful feature.
CH:
```
select 1 as x from (select 2 as x) t where x=2;
0 rows in set
```
PG:
```
select 1 as x from (select 2 as x) t where x=2;
x
---
1
```
Though, some warning / error about ambiguity would be handy. | 2019-11-18T21:33:07 | 2019-11-18T21:37:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,534,928 | 7,260 | abyss7 | According to stacktrace it's a known internal rdkafka problem, but it should be very rare - so more details about repro scenario are required. | 2019-11-19T14:34:37 | 2019-11-19T14:34:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,541,450 | 7,843 | amosbird | It looks like a revert to https://github.com/ClickHouse/ClickHouse/pull/6675 | 2019-11-19T14:49:12 | 2019-11-19T14:49:12 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 555,011,495 | 7,733 | mdonkers | Thx for the hint. Have been thinking about this, and have my doubts its the same issue.
The 'deletes' are not depending on another, and the script makes sure only one `DELETE` query runs per node. The only thing that might happen because we manually killed 'stuck' mutations, is that the script would trigger a `DELETE`... | 2019-11-18T13:19:27 | 2019-11-18T13:19:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,110,350 | 7,817 | yamrzou | @den-crane Thanks for the concise example.
It seems that with less decimals in the input, wrong values such as `37.000008` disappear, but `inf` and values without decimals persist. | 2019-11-18T17:02:09 | 2019-11-18T17:02:09 | {} | NONE |
ClickHouse | ClickHouse | 555,216,431 | 7,804 | dotNox | Now I understand what happens. Renames are also taken in consideration in WHERE clause, unlike in systems like MySQL or similar, where it is not possible ( as per my tests ) to use renamed fields in WHERE clause.
In this matter CH documentation is a little confusing.
Thank you for you help. | 2019-11-18T21:25:59 | 2019-11-18T21:25:59 | {} | NONE |
ClickHouse | ClickHouse | 555,491,793 | 7,767 | anheicang | > May be you should decrease `query_profiler_real_time_period_ns` / `query_profiler_cpu_time_period_ns` ?
>
> By default both are set to 1000000000, i.e. 1 second, so if your queries finishes faster than 1 sec, you will not get anything in `trace_log`. Try longer queries or smaller values for `query_profiler_real_ti... | 2019-11-19T12:45:41 | 2019-11-19T12:45:41 | {} | NONE |
ClickHouse | ClickHouse | 555,504,913 | 7,780 | nikitamikhaylov | 
Speedup scales almost linearly up to 6 threads. | 2019-11-19T13:22:17 | 2019-11-19T13:22:17 | {} | MEMBER |
ClickHouse | ClickHouse | 555,544,689 | 7,843 | abyss7 | > It looks like a revert to #6675
Hmm, will discuss with Alexey what was his reasoning. | 2019-11-19T14:56:21 | 2019-11-19T14:56:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,842,111 | 7,832 | sundy-li | > 1. can I query those sharded tables from my distribution engine without zookeeper?
It's fine without zookeeper when your tables are all none replicated tables.
But I suggest you better have zookeeper to use 'cluster ddls', it'll be easier to create tables on clusters.
> 2. can I create duplicated tables in m... | 2019-11-20T05:17:08 | 2019-11-20T05:17:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,763,283 | 7,804 | den-crane | I see 2 different issue here.
1. Not CH performance issue. But logical problem that `toDateTime(intDiv(toUInt32(t),7200)*7200) BETWEEN` and `ts between` are not equivalent and provide different results.
```
CREATE TABLE x (t DateTime) ENGINE = MergeTree PARTITION BY toDate(t) ORDER BY (t);
insert into x values('2... | 2019-11-17T16:36:59 | 2019-11-17T16:51:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,246,390 | 7,814 | den-crane | output_format_write_statistics
cl --output_format_write_statistics=0 -q "select 1 format JSON" | 2019-11-18T22:48:18 | 2019-11-18T22:48:18 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,644,721 | 7,849 | LizardWizzard | Thank you so much for explanation! Got it :) Changed partition expression and error is gone | 2019-11-19T18:30:26 | 2019-11-19T18:30:26 | {} | NONE |
ClickHouse | ClickHouse | 554,698,581 | 7,057 | zhang2014 | Does we get some news about this PR ? | 2019-11-17T03:10:07 | 2019-11-17T03:10:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,744,684 | 4,470 | rtbsolutions | same here
launched via docker image as a temporary solution | 2019-11-17T13:16:49 | 2019-11-17T13:16:49 | {} | NONE |
ClickHouse | ClickHouse | 554,884,617 | 7,807 | byx313 | > MaterializedView stores data in a usual table. This storage table can created explicitly (before MV) or implicitly (by MV). In the second case the name of the table will be '.inner.{MV_NAME}'
>
> ```
> create table testx (A Int64, B Date) Engine=MergeTree partition by toYYYYMM(B) order by A;
> create materialize... | 2019-11-18T07:17:32 | 2019-11-18T07:17:32 | {} | NONE |
ClickHouse | ClickHouse | 555,000,882 | 7,790 | bgranvea | so far I haven't been able to find a simple test case. There must be some case where AggregatingSortedBlockInputStream.merge is called with empty blocks to merge. In that case, we insert an extra value in columns with SimpleAggregateFunction types. The fix is easy.
I have seen this bug on my test platform with Clic... | 2019-11-18T12:47:45 | 2019-11-18T12:47:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,006,187 | 7,780 | akuzm | > But we can disable this feature by default or mark it as experimental.
I think I'll start with merging the segmentation for TSV that I already rewrote, and then we can add other formats one by one.
For now, I'll have to rewrite TSV segmentation once again so that it doesn't do so much copying :see_no_evil: | 2019-11-18T13:04:21 | 2019-11-18T13:04:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,015,057 | 7,811 | korjavin | >Please try version 19.15 or above.
I upgraded to ClickHouse client version 19.17.2.4 (official build).
and still see the same
> Do you supply message length?
as far as I understand there is some Le... | 2019-11-18T13:28:29 | 2019-11-18T13:28:29 | {} | NONE |
ClickHouse | ClickHouse | 555,530,617 | 7,417 | abyss7 | Should be fully fixed with https://github.com/ClickHouse/ClickHouse/pull/7821 | 2019-11-19T14:25:23 | 2019-11-19T14:25:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,533,793 | 7,261 | abyss7 | Will try to solve it in #7837 | 2019-11-19T14:32:00 | 2019-11-19T14:32:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,752,421 | 7,807 | den-crane | MaterializedView stores data in a usual table. This storage table can created explicitly (before MV) or implicitly (by MV). In the second case the name of the table will be '.inner.{MV_NAME}'
```
create table testx (A Int64, B Date) Engine=MergeTree partition by toYYYYMM(B) order by A;
create materialized view tes... | 2019-11-17T14:46:07 | 2019-11-17T14:54:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,041,188 | 7,811 | den-crane | > as far as I understand there is some Lenght that I can see in kafkatool
It's a length of kafka message. CH expects that one message can contain several protobufs. And to parse the message need to know the lengths of protobufs.
`_VarintBytes(len(serialized_msg)) + serialized_msg` https://github.com/ClickHouse/Cl... | 2019-11-18T14:35:07 | 2019-11-18T14:35:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,129,973 | 7,511 | apbodrov | Added feature to read and write gzip files based on engine parameters. Supported options are
- 'none' enforce no compression even ".gz" exists,
- 'auto' choose compression by file extension
- 'gzip' enforce gzip compression
Supported engines are S3/File/URL/HDFS.
| 2019-11-18T17:48:31 | 2019-11-18T17:48:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,198,663 | 7,811 | kap6 | I thought i would chime in, as I am running into an issue that relates to protobuf and kafka engine. I am taking the binary data from 00825_protobuf_format.insh testcase file. I am sending the message to a kafka queue and am trying to read the same into a CH table (kafka engine, protobuf format). There are no errors, b... | 2019-11-18T20:38:29 | 2019-11-18T20:41:06 | {} | NONE |
ClickHouse | ClickHouse | 555,206,753 | 7,811 | den-crane | @kap6 sounds like completely another issue. Did u check CH logs ? | 2019-11-18T21:00:33 | 2019-11-18T21:00:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,311,848 | 7,830 | den-crane | Because CH has a strong typing.
select * from (select toInt8(1) a) t join (select toUInt8(1) b) t1 on a=b;
DB::Exception: Type mismatch of columns to JOIN by: a Int8 at left, b UInt8 at right.
select * from (select toInt8(1) a) t join (select toUInt8(1) b) t1 on **toUInt8**(a)=b
┌─a─┬─b─┐
│ 1 │ 1 │
└───┴───┘... | 2019-11-19T03:07:56 | 2019-11-19T03:07:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,332,764 | 7,799 | maskshell | Data type mismatch. U can use toUInt16 to convert.
```SQL
SELECT
count()
FROM
system.query_log ANY
LEFT JOIN (
SELECT toUInt16(1) AS port)
USING port PREWHERE TYPE IN (1)
WHERE
event_date = today();
```
BUT, In version 19.11.x, system will report the exception message:
```
DB::Exception: Type mi... | 2019-11-19T04:54:27 | 2019-11-19T04:54:27 | {} | NONE |
ClickHouse | ClickHouse | 555,407,400 | 7,834 | SergeyMirvoda | Same error text also in #7463 | 2019-11-19T09:14:10 | 2019-11-19T09:14:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,499,389 | 7,780 | akuzm | ```
┌─threads─┬─median time─┬─GB/s─┬─per thread─┐
│ 1 │ 27.29 │ 0.28 │ 0.28 │
│ 2 │ 14.58 │ 0.54 │ 0.27 │
│ 3 │ 10.17 │ 0.77 │ 0.25 │
│ 4 │ 8.30 │ 0.94 │ 0.23 │
│ 5 │ 7.09 │ 1.11 │ 0.22 │
│ 6 │ 6.77 │ 1.16 │ ... | 2019-11-19T13:07:26 | 2019-11-19T13:07:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,532,780 | 7,302 | abyss7 | Will test and then close this issue. | 2019-11-19T14:29:53 | 2019-11-19T14:29:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,803,525 | 7,807 | byx313 | > No difference. Except early versions of CH did not support `drop partition` for MV. I assumed that you created this issue because you are experiencing some issues with `drop partition`.
yep! thank you so much! | 2019-11-20T02:14:10 | 2019-11-20T02:14:10 | {} | NONE |
ClickHouse | ClickHouse | 554,741,813 | 7,798 | hashbash | I'm faced with the same problem after update for other dictGet* functions. | 2019-11-17T12:42:22 | 2019-11-17T12:42:22 | {} | NONE |
ClickHouse | ClickHouse | 554,911,236 | 4,833 | yuzhichang | I noticed similar problem on system.query_log table.
My ClickHouse is 19.15.3.6-stable.
```
2019.11.18 13:09:31.397505 [ 13 ] {} <Debug> system.query_log (MergerMutator): Selected 5 parts from 20191118_34225_34295_18 to 20191118_34299_34299_0_34299
2019.11.18 13:09:31.397624 [ 13 ] {} <Debug> DiskSpaceMonitor: Re... | 2019-11-18T08:39:44 | 2019-11-18T08:43:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 555,030,056 | 7,811 | korjavin | But I am able to read from kafka with another app using the same proto file. | 2019-11-18T14:08:10 | 2019-11-18T14:08:10 | {} | NONE |
ClickHouse | ClickHouse | 555,532,312 | 7,348 | abyss7 | It's really is just tracing message, and INVALID - is some internal rdkafka value. | 2019-11-19T14:28:53 | 2019-11-19T14:28:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 554,830,297 | 6,267 | zhuboshuai | It's a severe problem to significantly reduce the data volumn clickhouse cluster reserved | 2019-11-18T02:59:31 | 2019-11-18T03:01:54 | {} | NONE |
ClickHouse | ClickHouse | 555,026,408 | 7,814 | tavplubix | It's documented feature: https://clickhouse.yandex/docs/en/interfaces/formats/#json
`JSON` and `JSONCompact` formats output some additional information, which may be useful. `JSONEachRow` outputs each row as a separate JSON object without any additional information. You can use `JSONEachRow` if you don't need those ... | 2019-11-18T13:59:03 | 2019-11-18T13:59:03 | {} | MEMBER |
ClickHouse | ClickHouse | 555,320,342 | 7,830 | Tasselmi | > We are not humans.
who are you ? ...............orz. | 2019-11-19T03:48:31 | 2019-11-19T03:48:31 | {} | NONE |
ClickHouse | ClickHouse | 549,183,576 | 7,421 | alexey-milovidov | BTW, different replicas can have different volumes and different placement of data parts (but the set of data parts as a whole will be synchronized). | 2019-11-03T22:01:44 | 2019-11-03T22:01:44 | {
"+1": 1
} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.