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 | 631,027,678 | 10,994 | den-crane | https://github.com/ClickHouse/ClickHouse/issues/10397#issuecomment-631027204 | 2020-05-19T19:15:37 | 2020-05-19T19:15:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,523,958 | 11,071 | r-zenine | Thanks @blinkov ,
I'll update the documentation following your remarks.
Have a good day. | 2020-05-20T14:51:30 | 2020-05-20T14:51:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,469,080 | 11,023 | azat | Please note that I've also disabled `system_tables_lazy_load` (description updated accordingly)
And this requires docker image update (although tests will pass even without updating the container, since none of existing tests in upstream/master depends on this). | 2020-05-18T22:29:40 | 2020-05-18T22:29:40 | {} | MEMBER |
ClickHouse | ClickHouse | 630,650,422 | 11,030 | dependabot-preview[bot] | OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting `@dependabot ignore this major version` or `@dependabot ignore this minor version`.
If you change your mind, just re-... | 2020-05-19T07:52:50 | 2020-05-19T07:52:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,145,493 | 10,917 | den-crane | Then I believe it's expected behavior.
Jemalloc does not free allocated memory immediately. To save CPU resources on deallocation/allocation cycles. | 2020-05-19T23:42:58 | 2020-05-19T23:42:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,481,375 | 11,056 | den-crane | Yes, Cassandra, InfluxDB much better for such queries. | 2020-05-20T13:42:13 | 2020-05-20T13:42:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,533,886 | 10,917 | den-crane | I don't really understand what is the problem here. CH able to use all server's memory. What is the difference is it used all the time or only part of time?
>Is it possible for users to configure ClickHouse to use an alternative memory allocator, so that memory is freed earlier?
Old versions of CH used tcmalloc a... | 2020-05-20T15:06:05 | 2020-05-20T15:11:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,554,447 | 10,605 | den-crane | not related to #10093.
Still relevant 20.3.9.70
? CREATE DICTIONARY RANGE_HASHED does not support arbitrary types in RANGE (only date). ? | 2020-05-20T15:38:25 | 2020-05-20T15:40:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,641,497 | 11,029 | azat | >node3_1 | Indirect leak of 5904 byte(s) in 82 object(s) allocated from:
>node1_1 | 10 0x1fc47388 (/usr/bin/clickhouse+0x1fc47388)
Is there any plans/work in progress to configure symbolizer?
Anyway looks like leak is here (if I not messed up anywhere)
```
addr2line -e ../clickhouse-common-stati... | 2020-05-19T07:33:33 | 2020-05-19T08:12:09 | {} | MEMBER |
ClickHouse | ClickHouse | 630,696,161 | 10,777 | alesapin | All tests passed (Performance flap). Need to fix only synchronization check | 2020-05-19T09:16:05 | 2020-05-19T09:16:05 | {} | MEMBER |
ClickHouse | ClickHouse | 630,714,872 | 8,492 | zhicwu | Close the issue as I'm happy with the custom jdbc-bridge implementation :) | 2020-05-19T09:51:45 | 2020-05-19T09:51:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,022,574 | 11,029 | azat | About test failures:
> test_replicated_merge_tree_s3/test.py::test_insert_select_replicated | ERROR
```
2020.05.19 19:28:25.278207 [ 1 ] {} <Error> Application: Caught exception while loading metadata: Code: 499, e.displayText() = DB::Exception: The specified bucket does not exist, Stack trace (when copying this... | 2020-05-19T19:05:30 | 2020-05-19T19:05:30 | {} | MEMBER |
ClickHouse | ClickHouse | 631,508,348 | 11,068 | den-crane | STR:
```
create table vt(datetime_value DateTime, value Float64) Engine=Memory;
create view computed_datum_hours as
SELECT
toStartOfHour(b.datetime_value) AS datetime_desc,
sum(b.value) AS value
FROM vt AS b
GROUP BY toStartOfHour(b.datetime_value)
Received exception from server (version... | 2020-05-20T14:27:12 | 2020-05-20T14:27:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,920,506 | 11,002 | qoega | Let's change PR name? | 2020-05-19T16:06:58 | 2020-05-19T16:06:58 | {} | MEMBER |
ClickHouse | ClickHouse | 630,960,925 | 11,029 | azat | >Build/Testing/Packaging Improvement
Did not find the right tag, feel free to update it to the right one. | 2020-05-19T17:18:30 | 2020-05-19T17:21:00 | {} | MEMBER |
ClickHouse | ClickHouse | 631,051,322 | 11,005 | filimonov | Maybe we should check version requirements in generated spec for rpm creation. | 2020-05-19T20:02:52 | 2020-05-19T20:02:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,140,424 | 11,003 | azat | >looks like we are unable to set the UID - but you checked this in the entrypoint - is there a other solution available to setup the docker images without root ?
Looks like you have docker with userns enabled and you tried `CLICKHOUSE_DO_NOT_CHOWN=1`, i.e. something like this:
```
docker run -e CLICKHOUSE_DO_NOT... | 2020-05-19T23:28:09 | 2020-05-19T23:28:09 | {} | MEMBER |
ClickHouse | ClickHouse | 631,527,909 | 11,059 | den-crane | Also we need a warp drive and laser guns | 2020-05-20T14:57:38 | 2020-05-20T14:57:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,543,987 | 11,063 | den-crane | >why we can not have primary key ( DateTime, other_field_1 , other_field_2) ?
Because high cardinality column `DateTime` makes `other_field_1` values far from each other and reduces effectiveness of descending to `where other_field_1=?`
>Honestly I do not completely understand the concept of index in Clickhouse... | 2020-05-20T15:21:55 | 2020-05-20T15:26:14 | {
"+1": 5
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,480,431 | 11,023 | azat | >Is there any reasons why we shouldn't disable system_tables_lazy_load by default?
That was my [initial suggestion](https://github.com/ClickHouse/ClickHouse/pull/9642#discussion_r401421174) :) | 2020-05-18T23:06:37 | 2020-05-18T23:06:37 | {} | MEMBER |
ClickHouse | ClickHouse | 630,611,511 | 5,681 | andyyzh | nice! | 2020-05-19T06:28:05 | 2020-05-19T06:28:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,636,757 | 11,023 | azat | >Functional stateless tests (release)
was ~28 min, now 32 min (may be time related, but don't think so) | 2020-05-19T07:24:29 | 2020-05-19T07:24:40 | {} | MEMBER |
ClickHouse | ClickHouse | 630,776,447 | 10,994 | nikitamikhaylov | What will be if you disable *dictionaries_lazy_load* ? | 2020-05-19T12:08:46 | 2020-05-19T12:08:46 | {} | MEMBER |
ClickHouse | ClickHouse | 630,854,237 | 11,041 | den-crane | It's not solved in 19.17.4.11 (and no plans to solve it, branch 19.17 is closed for good), but solved in 19.16.19.85.
So you can downgrade 19.17.4.11 to 19.16.19.85. Or install 20.4.2.9 on all replica. | 2020-05-19T14:25:06 | 2020-05-19T14:25:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,881,566 | 9,545 | akuzm | > There is a special exception in Linux license that permits to use API of system calls, but I'm not sure that these names are part of API.
I doubt it's possible to have any kind of license for something like 'dTLB-store-misses', these are just common words. And we won't use any of their code, we just need to use th... | 2020-05-19T15:04:58 | 2020-05-19T15:04:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,207,575 | 10,917 | huang--yijian | Is there a gurantteed deadline to free memory finally in jemalloc, so that the memory footprint is contained at certain level?
Jemalloc can't keep memory until almost all memory is consumed, as in our case.
Is it possible for users to configure ClickHouse to use an alternative memory allocator, so that memory is ... | 2020-05-20T03:06:43 | 2020-05-20T03:09:06 | {} | NONE |
ClickHouse | ClickHouse | 631,438,174 | 10,884 | brefre | I have increased kafka_max_block_size for all large topics, and tried to adjust various other settings, but the errors are still coming (3-4 per day).
I will now try to set offsets.commit.timeout.ms to 10k for my brokers, and set retries to 4, to see if that helps. | 2020-05-20T12:19:07 | 2020-05-20T12:19:07 | {} | NONE |
ClickHouse | ClickHouse | 630,475,339 | 11,023 | alexey-milovidov | Is there any reasons why we shouldn't disable `system_tables_lazy_load` by default? | 2020-05-18T22:49:45 | 2020-05-18T22:49:45 | {} | MEMBER |
ClickHouse | ClickHouse | 630,668,329 | 10,897 | filimonov | The issue caused by PR: https://github.com/ClickHouse/ClickHouse/pull/6617 (the feature was introduced in 19.14). | 2020-05-19T08:26:55 | 2020-05-19T08:26:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,716,704 | 8,723 | zhicwu | Close this as it's exact the same issue as #9609. | 2020-05-19T09:55:41 | 2020-05-19T09:55:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,753,362 | 11,029 | Jokser | Hi @azat . Thank you for reporting this. Could you please create an issue about this leak, I'll try to fix it in the nearest time. | 2020-05-19T11:16:16 | 2020-05-19T11:16:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,789,661 | 9,545 | And42 | So, for now the remaining issues are:
1. Too many file descriptors overall
2. Multiplexing + scaling in case there are more hardware counters enabled than the number of actual hardware counters units (not an issue by itself, but no ability to avoid it is an issue)
There is no way around to avoid creating a file de... | 2020-05-19T12:37:51 | 2020-05-19T12:38:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 629,808,947 | 9,545 | And42 | @akuzm seems like the best way is to just copy-paste event names code from https://github.com/torvalds/linux/blob/master/tools/perf/util/parse-events.c (i.e. `perf stat` names you've suggested). The only problem is perf has GPL-2.0 license. Don't really know how to better integrate the part of its code here without nee... | 2020-05-17T14:40:46 | 2020-05-19T21:00:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,192,174 | 11,056 | den-crane | There is no good solution, this task is not typical for OLAP.
```
select groupArray(date), facility, groupArray(value) from
(select date, facility, value from
(select '2017-09-09' date, 10002 facility, 10 value union all
select '2017-09-10' , 10001, 12 union all
select '2017-09... | 2020-05-20T02:15:43 | 2020-05-20T02:24:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,219,862 | 10,200 | vumdao | `Non-default settings: <joined_subquery_requires_alias>0</joined_subquery_requires_alias>
`
Should we enable or disable this for 20.3.9.70? I plan to use this version to fix the segment fault issue from my system
| 2020-05-20T03:51:20 | 2020-05-20T03:51:20 | {} | NONE |
ClickHouse | ClickHouse | 631,330,058 | 11,062 | ghost | @alexey-milovidov
Thank you very much.! | 2020-05-20T08:39:35 | 2020-05-20T08:39:50 | {} | NONE |
ClickHouse | ClickHouse | 631,411,964 | 11,070 | yfinkelstein | May be virtual columns can be used for this? I presume a virtual column can be passed to sharding key function as a parameter? | 2020-05-20T11:18:46 | 2020-05-20T11:18:46 | {} | NONE |
ClickHouse | ClickHouse | 630,613,850 | 10,917 | huang--yijian | As below, the memory footprint of ClickHouse server increases every 6 hours, in a zig-zag fashion.
We identified an applicaiton which launches memory-consuming ClickHouse queries, right every 6 hours.
We ... | 2020-05-19T06:33:26 | 2020-05-19T06:39:54 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 630,731,665 | 10,777 | alesapin | All tests passed (on previous commit), ya.make fixed. Merge! | 2020-05-19T10:27:07 | 2020-05-19T10:27:07 | {} | MEMBER |
ClickHouse | ClickHouse | 630,735,090 | 10,962 | ObjatieGroba | @hczhcz
I do research about easiest way of implementation UDF. It is possible to implement isolated C++ UDF without strong performance degradation (#10750).
I will implement agreed by @alexey-milovidov UDF soon. | 2020-05-19T10:34:29 | 2020-05-19T10:34:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,853,063 | 10,128 | filimonov | #11042 | 2020-05-19T14:23:23 | 2020-05-19T14:23:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,612,223 | 11,074 | azat | @Jokser Maybe test with unavailable S3 worth adding? (that triggers the initial issue) | 2020-05-20T17:19:59 | 2020-05-20T17:19:59 | {} | MEMBER |
ClickHouse | ClickHouse | 630,403,988 | 11,022 | azat | > force-pushed the azat:memory-tracking branch from c6715e0 to b587120 now
Drop the patch with removing the test `leak_when_memory_limit_exceeded.sql` | 2020-05-18T19:58:10 | 2020-05-18T19:58:10 | {} | MEMBER |
ClickHouse | ClickHouse | 630,523,305 | 6,462 | weiyongyuan | The same problem!I modified the ttl os the system.trace_log,but the auto ttl doesn't work.
>Here is the table sql:CREATE TABLE system.trace_log (`event_date` Date, `event_time` DateTime, `revision` UInt32, `timer_type` Enum8('Real' = 0, 'CPU' = 1), `thread_number` UInt32, `query_id` String, `trace` Array(UInt64)) ENGI... | 2020-05-19T01:35:55 | 2020-05-19T02:11:37 | {} | NONE |
ClickHouse | ClickHouse | 630,789,344 | 11,025 | qoega | Actually it may by too risky to add them as synonyms without difference in behavior.
https://stackoverflow.com/a/176565
It can be hard to make it more close to ANSI if it has different behavior. Have you checked the difference? | 2020-05-19T12:37:11 | 2020-05-19T12:37:11 | {} | MEMBER |
ClickHouse | ClickHouse | 631,045,758 | 10,775 | azat | I was looking into `test_replicated_merge_tree_s3` [failures](https://clickhouse-test-reports.s3.yandex.net/11029/f57ac3d3f025de15dc1255671a976bfeac641650/integration_tests_(asan).html), there was some issues with bucket availability (which also is an issue), and in this case an exception while loading metadata will be... | 2020-05-19T19:52:16 | 2020-05-19T19:52:16 | {} | MEMBER |
ClickHouse | ClickHouse | 631,365,386 | 11,051 | 4ertus2 | 20.3- are not affected | 2020-05-20T09:41:49 | 2020-05-20T09:41:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,522,991 | 11,071 | blinkov | Hi @r-zenine, it's just a marker so we're less likely to miss pull-requests that definitely need to be documented.
Your docs draft looks ok in general. Here are some comments:
* The return data type is not mentioned explicitly (sounds like UInt8).
* We could be more specific on what "processed as a value" means.... | 2020-05-20T14:49:58 | 2020-05-20T14:49:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,529,422 | 11,071 | blinkov | @r-zenine thanks, you too :) | 2020-05-20T15:00:04 | 2020-05-20T15:00:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,476,585 | 9,032 | alexey-milovidov | @yurriy is not working on this issue right now, we have to reassign. | 2020-05-18T22:53:53 | 2020-05-18T22:53:53 | {} | MEMBER |
ClickHouse | ClickHouse | 630,809,528 | 10,914 | 4ertus2 | ```
create table t (x UInt32, id UInt32) ENGINE = MergeTree() order by (id);
insert into t values (0, 0);
select 1 from t as l join t as r on l.id = r.id prewhere l.x = 1; -- OK
select 1 from t as l join t as r on l.id = r.id where r.x = 1; -- OK
select 1 from t as l join t as r on l.id = r.id prewhere l.x = 1 w... | 2020-05-19T13:15:01 | 2020-05-19T13:15:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,413,277 | 8,329 | vitlibar | The changes have been merged in https://github.com/ClickHouse/ClickHouse/pull/11050. | 2020-05-20T11:21:44 | 2020-05-20T11:21:44 | {} | MEMBER |
ClickHouse | ClickHouse | 631,427,380 | 11,063 | nvartolomei | https://clickhouse.tech/docs/en/engines/table-engines/mergetree-family/mergetree/#selecting-the-primary-key
High cardinality is good as just by looking at the index (which is in memory) a large set of rows will be discarded.
| 2020-05-20T11:55:16 | 2020-05-20T11:55:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,201,631 | 10,994 | maison2710 | @nikitamikhaylov I don't think disable _dictionaries_lazy_load_ should resolve the root cause. | 2020-05-20T02:45:21 | 2020-05-20T02:45:21 | {} | NONE |
ClickHouse | ClickHouse | 631,486,543 | 11,071 | r-zenine | HI @blinkov ,
I see a doc-alert flag on the PR. I guess I might have missed something where preparing it. Sorry About that, Can you tell me what's missing ? I will gladly update my work :)
Ps : The documentation is already updated in the document docs/en/sql-reference/functions/array-functions.md
Thanks | 2020-05-20T13:51:09 | 2020-05-20T13:51:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,506,719 | 10,605 | georgantasp | https://github.com/ClickHouse/ClickHouse/issues/10093 | 2020-05-20T14:24:37 | 2020-05-20T14:24:37 | {} | NONE |
ClickHouse | ClickHouse | 630,546,653 | 10,976 | liuxun666 | I have the same problem | 2020-05-19T02:48:47 | 2020-05-19T02:48:47 | {} | NONE |
ClickHouse | ClickHouse | 630,638,180 | 11,021 | tavplubix | We don't have such test, but this scenario is used in performance tests report builder. `clickhouse-local` doesn't drop database by itself on exit | 2020-05-19T07:27:18 | 2020-05-19T07:27:18 | {} | MEMBER |
ClickHouse | ClickHouse | 630,644,088 | 5,615 | amit-gshe | Same issue at ClickHouse server version 20.4.2.9 (official build). | 2020-05-19T07:39:18 | 2020-05-19T07:39:18 | {
"+1": 4
} | NONE |
ClickHouse | ClickHouse | 630,653,697 | 10,962 | hczhcz | @jason-godden
We have this kind of UDF function in C++ (#6991).
It is not viable in typical scenarios, considering the user code can crash the whole server (no matter if intentionally or unintentionally). We use it to do some quick jobs in our private server. | 2020-05-19T07:59:12 | 2020-05-19T07:59:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,720,896 | 9,609 | zhicwu | Thank you @sergey-buturlakin. I can confirm your solution works very well - tested using 50 concurrent users and no more timeout :) Would be great if someone can create a pull request to fix this in trunk. | 2020-05-19T10:04:08 | 2020-05-19T10:04:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,735,281 | 11,011 | nikitamikhaylov | Try to test official build for ARM, which you can download from [here](https://clickhouse-builds.s3.yandex.net/0/3d5e8e0f2c070a17bdfd4cf47f8ce2184dce433c/clang-10-aarch64_relwithdebuginfo_none_bundled_unsplitted_disable_binary/clickhouse).
Note, that it is master with newer version of ClickHouse. If it is reproduced ... | 2020-05-19T10:34:56 | 2020-05-19T10:34:56 | {} | MEMBER |
ClickHouse | ClickHouse | 630,995,924 | 10,858 | azat | >Merge remote-tracking branch 'upstream/master' into min_mv_insert_blo
FWIW I can rebase it against upstream/master to avoid history branching | 2020-05-19T18:20:39 | 2020-05-19T18:20:39 | {} | MEMBER |
ClickHouse | ClickHouse | 631,324,416 | 11,002 | alesapin | test failure is not related to changes. Performance is also not affected. | 2020-05-20T08:30:05 | 2020-05-20T08:30:05 | {} | MEMBER |
ClickHouse | ClickHouse | 631,372,471 | 11,056 | Shanginre | thanks for the answer. I'm used to working with relational databases, this is a common case there. Please tell me in which DBMS such a query will be executed quickly for 100 million rows? Cassandra, InfluxDB, ....? I am now choosing a DBMS for IOT project. | 2020-05-20T09:55:19 | 2020-05-20T09:55:19 | {} | NONE |
ClickHouse | ClickHouse | 631,528,750 | 11,028 | vitlibar | There are some code style-related errors, could you fix them?
https://clickhouse-test-reports.s3.yandex.net/11028/0b51e25a6991196f4fcedda0475a7f1428eaf6bf/style_check.html#fail1 | 2020-05-20T14:58:55 | 2020-05-20T14:59:30 | {} | MEMBER |
ClickHouse | ClickHouse | 631,570,987 | 11,028 | vitlibar | A new [test](https://github.com/ClickHouse/ClickHouse/tree/master/tests/queries/0_stateless) for this new syntax would be useful. Can you write it please? | 2020-05-20T16:05:26 | 2020-05-20T16:05:26 | {} | MEMBER |
ClickHouse | ClickHouse | 631,616,763 | 11,028 | Potya | OMW
| 2020-05-20T17:28:22 | 2020-05-20T17:28:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,407,307 | 10,983 | azat | >Thread names are a bit 'out of convention' (other threads use CamelCasing with shortcuts) but ok.
By initial thoughts was:
- keep `rdk` for "compatibility"
- broker/main/background abbrevs
- table name
Don't see where CamelCase can be used here, but suggestions are welcome! | 2020-05-18T20:05:28 | 2020-05-18T20:05:28 | {} | MEMBER |
ClickHouse | ClickHouse | 630,475,817 | 9,545 | alexey-milovidov | There is a special exception in Linux license that permits to use API of system calls, but I'm not sure that these names are part of API. | 2020-05-18T22:51:23 | 2020-05-18T22:51:23 | {} | MEMBER |
ClickHouse | ClickHouse | 630,489,104 | 11,023 | azat | >Let's do it...
Ok, will do separately (and I removed patch with disable it for tests separately) | 2020-05-18T23:35:06 | 2020-05-18T23:35:06 | {} | MEMBER |
ClickHouse | ClickHouse | 630,531,994 | 7,900 | zhang2014 | > So yep add more custom queries possible for the same htttp route / path (/metrics) and add the possibilities to have more custom output.
@ut0mt8 Maybe HTTP handler config + template format will help you ?
reference:
https://github.com/ClickHouse/ClickHouse/pull/7900#issuecomment-558993423
https://clickhouse... | 2020-05-19T02:02:56 | 2020-05-19T02:02:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,457,240 | 11,063 | hoseiney | thanks. so having DateTime column in primary key is OK ?
based on the doc(you also shared):
> Long data ranges (several times longer than the index_granularity) with identical values for (a, b) are common
for one value of (a,b) we should have thousands of records, right ?
so if we have : primary key (Da... | 2020-05-20T12:58:20 | 2020-05-20T13:01:10 | {} | NONE |
ClickHouse | ClickHouse | 631,529,927 | 11,028 | Potya | OK, i will fix them soon | 2020-05-20T15:00:50 | 2020-05-20T15:00:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,598,990 | 3,830 | 4ertus2 | The pattern could be combined via `AND` with normal JOIN keys that should not join NULLs in special way. We need special join per key not per JOIN.
```
... JOIN t3 ON
(
( (t0.NAME = t3.NAME) OR ((t0.NAME IS NULL) AND (t3.NAME IS NULL)) )
AND (t0.MONTH_ID = t3.MONTH_ID)
)
``` | 2020-05-20T16:55:22 | 2020-05-20T16:55:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,608,361 | 11,005 | winter7 | @den-crane Thanks!
But why hasn't the other version encountered this situation, what is `clickhouse-common-static`? | 2020-05-19T06:20:54 | 2020-05-19T06:20:54 | {} | NONE |
ClickHouse | ClickHouse | 631,287,494 | 11,062 | alexey-milovidov | @achimbab We're alright.
> 20.3.5.21
We have already released four bugfixes after this version, could you please check 20.3.9? | 2020-05-20T07:21:11 | 2020-05-20T07:21:11 | {} | MEMBER |
ClickHouse | ClickHouse | 631,290,592 | 11,062 | alexey-milovidov | I have checked master version and the issue does not reproduce. | 2020-05-20T07:27:51 | 2020-05-20T07:27:51 | {} | MEMBER |
ClickHouse | ClickHouse | 631,475,129 | 8,976 | sundarv85 | Thanks @blinkov We will try upgrading and check. | 2020-05-20T13:30:53 | 2020-05-20T13:30:53 | {} | NONE |
ClickHouse | ClickHouse | 631,517,273 | 11,063 | den-crane | if you need queries like this ` where other_field_1=? and DateTime between ? and ?`
` primary key ( other_field_1 , other_field_2, DateTime)`
Or
` primary key ( toStartOfHour(DateTime), other_field_1 , other_field_2, )` | 2020-05-20T14:40:57 | 2020-05-20T14:40:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,529,585 | 11,063 | hoseiney | yes this is exactly our use case.
why we can not have primary key ( DateTime, other_field_1 , other_field_2) ?
Honestly I do not completely understand the concept of index in Clickhouse, I read the doc but I am not sure how to design it.
How can I read more about the concept (in addition to Clickhouse MergeTree ... | 2020-05-20T15:00:20 | 2020-05-20T15:00:20 | {} | NONE |
ClickHouse | ClickHouse | 625,226,195 | 10,726 | hhell | > `dateDiff()` function
Other SQL-level hacks to do this:
* `cast(val as Int64)` (as in the example above) – returns unix timestamp (UTC)
* `(cast(val as DateTime('UTC'), toTypeName(val))` – UTC isoformat + original timezone (also sufficient to reconstruct the value).
* Construct the ISO8601 format with t... | 2020-05-07T12:31:48 | 2020-05-20T15:29:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,467,643 | 11,021 | alexey-milovidov | I remember that we discussed some use case for `clickhouse-local` - the user creates tables and fills them with data, then `clickhouse-local` exits and data remains available for further manipulations.
Do we have a test for this scenario? | 2020-05-18T22:25:09 | 2020-05-18T22:25:40 | {} | MEMBER |
ClickHouse | ClickHouse | 630,483,524 | 11,023 | alexey-milovidov | Let's do it... | 2020-05-18T23:16:47 | 2020-05-18T23:16:47 | {} | MEMBER |
ClickHouse | ClickHouse | 630,547,147 | 10,976 | liuxun666 | I have a table that is partitioned by date, and I want to drop the specified partition with the passed date parameter | 2020-05-19T02:50:26 | 2020-05-19T02:50:26 | {} | NONE |
ClickHouse | ClickHouse | 630,859,204 | 11,005 | den-crane | yum version resolver installed the latest clickhouse-common-static because clickhouse-server-20.3.8.53 depends on it. But the latest clickhouse-common-static-20.4.2.9 is require clickhouse-server-20.4.2.9
clickhouse-common-static -- is the main CH package. It is the real CH binary.
clickhouse-server & clickhouse-... | 2020-05-19T14:32:26 | 2020-05-19T14:32:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,882,327 | 9,545 | akuzm | > IMHO, the best way for now is to add two settings (`enable_perf_events`: `defaults to false` and `perf_events_list`: `defaults to all events`) and a threshold.
Sounds good to me. | 2020-05-19T15:06:08 | 2020-05-19T15:06:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,976,970 | 3,950 | haoch | Hi @alexey-milovidov @alexelex, any update on this issue?
Looks like clickhouse `insert_quorum` works with some kind of `exclusive lock` with zookeep path `storage.zookeeper_path + "/quorum/status" ` (`/clickhouse/tables/${SHARD}/${TABLE}/quorum/status`).
So here are my questions:
1) What's the design purpose?
... | 2020-05-19T17:44:08 | 2020-05-19T18:28:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,039,372 | 9,609 | filimonov | @sergey-buturlakin why don't you send a PR? | 2020-05-19T19:39:37 | 2020-05-19T19:39:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,201,265 | 10,994 | maison2710 | https://github.com/ClickHouse/ClickHouse/issues/10397#issuecomment-631200765 | 2020-05-20T02:44:02 | 2020-05-20T02:44:02 | {} | NONE |
ClickHouse | ClickHouse | 631,524,634 | 11,061 | den-crane | Your table is partitioned by date. You can just drop partitions https://clickhouse.tech/docs/en/sql-reference/statements/alter/#alter_drop-partition and attach partition from the temporary table https://clickhouse.tech/docs/en/sql-reference/statements/alter/#alter_attach-partition-from
---
Engine GraphiteMergeTre... | 2020-05-20T14:52:34 | 2020-05-20T14:52:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,481,035 | 9,624 | CurtizJ | UBSan ```../contrib/lz4/lib/lz4.c:818:44: runtime error: applying zero offset to null pointer``` - ok. See #10631
```00834_cancel_http_readonly_queries_on_client_close``` fails, beacuse Poco update, which fixes this bug, has broken all odbc tests and was reverted. We have to apply this patch properly later and relea... | 2020-05-18T23:08:43 | 2020-05-18T23:22:36 | {} | MEMBER |
ClickHouse | ClickHouse | 630,740,258 | 11,011 | wushiming540 | it is build error, we have solved.。 thank you | 2020-05-19T10:45:35 | 2020-05-19T10:45:35 | {} | NONE |
ClickHouse | ClickHouse | 630,851,140 | 11,031 | den-crane | duplicate https://github.com/ClickHouse/ClickHouse/issues/9008 | 2020-05-19T14:20:18 | 2020-05-19T14:20:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 630,960,749 | 11,022 | azat | >Functional stateless tests (ubsan) — fail: 1, passed: 1848, skipped: 6
```
2020-05-19 14:11:43 Code: 60. DB::Exception: Received from localhost:9000. DB::Exception: Table system.trace_log doesn't exist..
```
Can be fixed with some hack, but it will go away anyway after #11029 | 2020-05-19T17:18:08 | 2020-05-19T17:18:08 | {} | MEMBER |
ClickHouse | ClickHouse | 631,027,204 | 10,397 | den-crane | Since 19.6? CH unable to use dictionaries from main folder using wildcards *_dictionary.xml
Try to create a new folder `mkdir /etc/clickhouse-server/dict` and put dictionarie's xml there.
and add to config:
` <dictionaries_config>dict/*.xml</dictionaries_config> ` | 2020-05-19T19:14:40 | 2020-05-19T19:15:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 631,200,765 | 10,397 | maison2710 | @den-crane in my setup, I already did that. Actually, server still can load the dictionaries from .xml files when server starts. Then I still can create Dictionary table. After that, the issue only happens when server needs to restart. At this time, server can not ATTACH the Dictionary table, which I created before res... | 2020-05-20T02:42:30 | 2020-05-20T02:42:30 | {} | NONE |
ClickHouse | ClickHouse | 631,502,030 | 11,067 | den-crane | ```
SELECT value
FROM system.events
WHERE event LIKE 'InsertQuery'
SELECT
event_time,
ProfileEvent_InsertQuery,
runningDifference(ProfileEvent_InsertQuery) ins_per_sec
FROM system.metric_log
WHERE event_date = today()
ORDER BY event_time DESC
LIMIT 50
``` | 2020-05-20T14:17:17 | 2020-05-20T14:17:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 636,125,436 | 11,296 | alexey-milovidov | The issue https://github.com/ClickHouse/ClickHouse/issues/6417 is not relevant to this bug but has to be checked again.
No more relevant issues found on GitHub. | 2020-05-29T18:37:17 | 2020-05-29T18:37:27 | {} | MEMBER |
ClickHouse | ClickHouse | 636,370,553 | 9,154 | alexey-milovidov | Sorry, this PR introduced race condition:
https://clickhouse-test-reports.s3.yandex.net/11301/8a4c2380ddd7a1e316f4d3d1eb5d178bccabec1f/stress_test_(thread)/stderr.log
(I will fix it) | 2020-05-30T18:55:20 | 2020-05-30T18:55:20 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.