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 | 585,256,772 | 9,080 | qoega | OK for me. | 2020-02-12T15:21:16 | 2020-02-12T15:21:16 | {} | MEMBER |
ClickHouse | ClickHouse | 585,699,317 | 9,064 | grootman2007 | @CurtizJ same output with 20.1.4.14:
<code><pre>
2020.02.13 14:33:48.410856 [ 126 ] {} <Fatal> BaseDaemon: ########################################
2020.02.13 14:33:48.472786 [ 126 ] {} <Fatal> BaseDaemon: (version 20.1.4.14 (official build)) (from thread 100) (query_id: 622a74cc-6bd6-40cf-85bb-446c6eef5db9) Recei... | 2020-02-13T11:36:02 | 2020-02-13T11:36:02 | {} | NONE |
ClickHouse | ClickHouse | 585,765,719 | 9,077 | alexey-milovidov | That's reasonable.
Actual implementation should use divide + remainder of division by power of ten. | 2020-02-13T13:56:36 | 2020-02-13T13:56:36 | {} | MEMBER |
ClickHouse | ClickHouse | 585,929,111 | 9,099 | filimonov | > @tavplubix said that it's better to extract them in AST level.
Engine=Merge, distributed - too?
| 2020-02-13T19:23:06 | 2020-02-13T19:23:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,066,282 | 6,863 | cameronbraid | I'd like to know the same | 2020-02-14T02:22:39 | 2020-02-14T02:22:39 | {} | NONE |
ClickHouse | ClickHouse | 586,176,819 | 9,091 | YiuRULE | @alexey-milovidov Thanks for your review, I applied the requested fix. :)
(except the part you mentioned `perf top`, I would need to have an insight about what kind of information you need) | 2020-02-14T09:32:13 | 2020-02-14T09:32:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,282,355 | 6,128 | qoega | Is everything done in this issue? | 2020-02-14T13:13:28 | 2020-02-14T13:13:28 | {} | MEMBER |
ClickHouse | ClickHouse | 586,290,786 | 9,118 | aligin | I've reproduced this on `v19.17.7.52`.
After the `enable_optimize_predicate_expression = 0` I've got the same as you. But, is this right to disable enable_optimize_predicate_expression?
I don't understand the broken DateEnd column. How it must looks like? | 2020-02-14T13:37:40 | 2020-02-14T13:37:40 | {} | NONE |
ClickHouse | ClickHouse | 586,328,289 | 9,110 | den-crane | > What about
>
> > ## ALTER {#query_language_queries_alter}
>
> We should change it to ALTER TABLE using the same logic?
No, I don't think so. Topics (anchors) must be unique across .md . Before my fix that document had two topics with the same name and the same anchor about two different `alter move` commands... | 2020-02-14T15:09:16 | 2020-02-14T15:09:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,252,777 | 9,070 | akuzm | Is it possible to add some kind of integration test for this? | 2020-02-12T15:13:15 | 2020-02-12T15:13:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,476,735 | 9,088 | chu1070y | ```sql
CREATE TABLE default.test2
(
`EventDate` Date,
`CounterID` Nullable(String),
`UserID` Nullable(UInt32)
)
ENGINE = MergeTree()
ORDER BY EventDate;
INSERT INTO test2 VALUES ('2020-01-02', '', 2)('2020-01-03', 'aaa', 0);
INSERT INTO test2 VALUES ('2020-01-02', '', 2)('2020-01-02', '', 2)('20... | 2020-02-12T23:49:57 | 2020-02-12T23:49:57 | {} | NONE |
ClickHouse | ClickHouse | 585,686,005 | 9,064 | CurtizJ | Here you can find version 20.1.4.14
https://repo.yandex.ru/clickhouse/deb/stable/main/
| 2020-02-13T11:28:14 | 2020-02-13T11:28:14 | {} | MEMBER |
ClickHouse | ClickHouse | 585,760,203 | 1,713 | qoega | Thank you for your interest in ClickHouse optimization.
You can make an issue with more specific proposal what can be done.
Current question is answered. | 2020-02-13T13:44:22 | 2020-02-13T13:44:22 | {} | MEMBER |
ClickHouse | ClickHouse | 585,765,368 | 2,414 | qoega | If you still have questions about small query performance you can try to reproduce problem with actual version of ClickHouse and reopen this issue or create new one. | 2020-02-13T13:55:50 | 2020-02-13T13:55:50 | {} | MEMBER |
ClickHouse | ClickHouse | 585,775,551 | 9,052 | alexey-milovidov | The changelog entry looks like it is for developers but it is unclear for the users. | 2020-02-13T14:16:40 | 2020-02-13T14:16:47 | {} | MEMBER |
ClickHouse | ClickHouse | 586,288,680 | 9,118 | 4ertus2 | I cannot reproduce your result on master. It's not an optimisation for sure: it's broken DateEnd column. I've got these results (id's been cropped)
```
9be6f472 1 Problem1 2020-02-14 10:16:00 2020-02-14 10:17:00 60
cb35f07e 1 Problem2 2020-02-14 10:18:00 2020-02-... | 2020-02-14T13:31:54 | 2020-02-14T13:36:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,295,092 | 9,118 | 4ertus2 | Fixed in #8503
You should update to newer version. It has not been ported to 19.17 cause of code conflicts. | 2020-02-14T13:49:39 | 2020-02-14T13:58:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,311,200 | 5,104 | Melancholic | See here #9126 | 2020-02-14T14:27:40 | 2020-02-14T14:27:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,014,557 | 6,639 | Phantomape | Able to reproduce in our 4-node servers via Clickhouse 19.16.2.2. It seems like a race condition to me, and we bypass these errors by letting the two Poco Event variable in DDLWorker to send continue signal even though these two exceptions happen. However, that may not seem like a correct fix at all, more like a dirty ... | 2020-02-12T03:51:47 | 2020-02-12T03:51:47 | {} | NONE |
ClickHouse | ClickHouse | 585,252,385 | 9,088 | 4ertus2 | Could you attach your schema here too? | 2020-02-12T15:12:23 | 2020-02-12T15:12:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,599,638 | 8,573 | dstruck | Works again. Probably BADSIG was temporarily reported due to updates to the repository.
Unfortunately I can only replicate the TLS issues with the 20. series in the test setup. I have tested the 20.1.4.14 version and the problem persists.
```
20.1.2.4
requests.exceptions.SSLError: HTTPSConnectionPool(host='loca... | 2020-02-13T07:59:44 | 2020-02-13T07:59:44 | {} | NONE |
ClickHouse | ClickHouse | 585,633,971 | 9,080 | alesapin | We need motivation for each rename. Now I don't see any motivation at all. Not OK for me. | 2020-02-13T09:31:53 | 2020-02-13T09:31:53 | {} | MEMBER |
ClickHouse | ClickHouse | 585,766,641 | 9,100 | nvartolomei | In #9094 the crash was happening on clickhouse start, not on shutdown. | 2020-02-13T13:58:49 | 2020-02-13T13:58:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,898,389 | 9,101 | abyss7 | What is your exact Clickhouse version? | 2020-02-13T18:17:09 | 2020-02-13T18:17:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,268,371 | 6,331 | tavplubix | After #7215 `NULL` keyword can be used as `DEFAULT` if `input_format_null_as_default=1` and column type is not `Nullable` | 2020-02-14T12:29:49 | 2020-02-14T12:29:49 | {} | MEMBER |
ClickHouse | ClickHouse | 585,015,302 | 9,084 | den-crane | It could be a system tables with logs. Or crashdumps.
Check what tables sizes really are with:
```
select round(sum(bytes)/1024/1024) size_mb, count() parts_count, sum(rows),database,table
from system.parts
where active
group by database,table
order by size desc
``` | 2020-02-12T03:55:15 | 2020-02-12T03:55:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,257,100 | 9,072 | den-crane | Do not see any issue, closing. | 2020-02-12T15:21:55 | 2020-02-12T15:21:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,263,960 | 9,086 | den-crane | >CH version 19.15.4
https://github.com/ClickHouse/ClickHouse/pull/6168 I think `move partition to another table` is not even in release yet.
Though it's working in 20.2.1.234
related to https://github.com/ClickHouse/ClickHouse/issues/4729 | 2020-02-12T15:35:34 | 2020-02-12T15:36:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,267,814 | 9,084 | den-crane | @RangerWolf
My strongly suggestion is to use single partition.
create table .... Engine=ReplacingMergeTree order by account_id -- `partition by -- ommited`
And to stop using `alter update `-- use `insert into + optimize` instead.
Consider to switch to external dictionaries https://clickhouse.tech/docs/en/q... | 2020-02-12T15:43:27 | 2020-02-12T15:44:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,648,252 | 9,080 | proller | Integration with Yandex internal repository was performed many times as is without all these moves.
/src/bin - 'bin' usually means already compiled files.
/src/contrib - looks like sabotage
| 2020-02-13T10:05:48 | 2020-02-13T10:05:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,717,591 | 9,064 | CurtizJ | Well then, it's a new issue. Thanks. | 2020-02-13T11:48:52 | 2020-02-13T11:48:52 | {} | MEMBER |
ClickHouse | ClickHouse | 585,775,217 | 9,052 | alexey-milovidov | Why don't also add a test? | 2020-02-13T14:15:59 | 2020-02-13T14:15:59 | {} | MEMBER |
ClickHouse | ClickHouse | 586,211,646 | 9,070 | urykhy | @akuzm i'm not sure
we arrange test environment to trigger this kind of error, but:
* it's build from closed components (uploader, tests, test data)
* not 100% accuracy, this patch only reduce rate of error. it's still possible that data will not reach MV or reach MV twice.
* time to run at least 15 minutes. | 2020-02-14T11:01:19 | 2020-02-14T11:10:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,266,139 | 9,096 | tavplubix | Related: #7882, #7943 | 2020-02-14T12:22:50 | 2020-02-14T12:22:50 | {} | MEMBER |
ClickHouse | ClickHouse | 586,424,448 | 7,851 | tavplubix | The setting added to 19.14.11 and 19.16.14 | 2020-02-14T19:00:36 | 2020-02-14T19:00:36 | {} | MEMBER |
ClickHouse | ClickHouse | 585,253,192 | 8,922 | qoega | Currently I can propose to read about functions that work with binary representations that can be found in [`bitmap`](https://clickhouse.tech/docs/en/query_language/functions/bitmap_functions/) docs section.
ClickHouse is an OpenSource project where you are welcome to contribute and add any functions, that can be us... | 2020-02-12T15:14:05 | 2020-02-12T15:14:05 | {} | MEMBER |
ClickHouse | ClickHouse | 585,256,751 | 9,084 | den-crane | `f5f61b319bd8ff5933ad028f3a7c2ca3_681_109969_23_111268/`
@RangerWolf
f5f61b319bd8ff5933ad028f3a7c2ca3 -- what's your table `PARTITION BY` and why?
111268 -- do you use mutations and daily bases operations? Consider to ReplacingMergeTree https://clickhouse.tech/docs/en/operations/table_engines/replacingmerget... | 2020-02-12T15:21:14 | 2020-02-12T15:21:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,272,024 | 9,092 | den-crane | Sorry. I misread.
What CH version do you use? I could not reproduce
```
CREATE TABLE my_table
(
`purchase_id` UInt32,
`pcode_series_id` Nullable(UInt32),
`purchase_date` Date DEFAULT CAST('1970-01-01', 'Date'),
`purchase_datetime` DateTime DEFAULT CAST('0000000000', 'DateTime'),
`deal_o... | 2020-02-12T15:51:50 | 2020-02-12T15:51:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,570,650 | 8,334 | filimonov | Opened an FR in apache arrow issue tracker: https://issues.apache.org/jira/browse/ARROW-7845 | 2020-02-13T06:19:50 | 2020-02-13T06:19:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,435,741 | 9,096 | TimonKK | If I use node js [lib](https://www.npmjs.com/package/clickhouse) and insert like [this](https://pastebin.com/i5eFUGs3) I get error about ClientID:
`
No such column ClientID in table session_temp
`
Full stack trace:
`
2020.02.13 00:43:58.989567 [ 78 ] {a5d5127c-9366-434e-bf7f-8bffabe052d3} <Debug> executeQuery... | 2020-02-12T21:51:12 | 2020-02-13T09:02:46 | {} | NONE |
ClickHouse | ClickHouse | 585,670,481 | 8,645 | 4ertus2 | For now there's only bankers rounding for floats and both for integers
https://clickhouse.tech/docs/en/query_language/functions/rounding_functions/#rounding_functions-round
https://clickhouse.tech/docs/en/query_language/functions/rounding_functions/#roundbankers | 2020-02-13T11:01:22 | 2020-02-13T11:01:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,767,446 | 9,100 | nvartolomei | The trace indeed shows that it is called from shutdown 🤔. I can try and reproduce it with this patch and without next week. | 2020-02-13T14:00:33 | 2020-02-13T14:00:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,181,240 | 8,645 | lmeng987 | Thanks @4ertus2 for help me understanding the round of clickhouse.
If I want to use round half to up function, do you know if clickhouse has the function to do the round half to up? for I need make sure the result from clickhouse should be same from other DB with same SQL.
Thanks! | 2020-02-12T12:24:07 | 2020-02-12T12:24:07 | {} | NONE |
ClickHouse | ClickHouse | 585,259,629 | 9,084 | RangerWolf | @den-crane
Hi,
This table does not have date related columns. It contains columns like :
account_id / account_name / refresh_token / budget ...
I thought there are only around 1K records, so I use account_id as partition key and sort key.
Any suggestion? | 2020-02-12T15:27:01 | 2020-02-12T15:27:01 | {} | NONE |
ClickHouse | ClickHouse | 585,356,818 | 9,092 | VeryaskinMax | > What CH version do you use? I could not reproduce
Version: `20.1.2.4`
I checked it also with version `19.17.4.11` - and it works well.
But I wonder why it's doesn't work in version `20.1.2.4`. | 2020-02-12T18:42:41 | 2020-02-12T18:42:41 | {} | NONE |
ClickHouse | ClickHouse | 585,749,660 | 9,100 | alexey-milovidov | I have difficulties with providing a test.
Let's assume it is good enough. | 2020-02-13T13:18:08 | 2020-02-13T13:18:08 | {} | MEMBER |
ClickHouse | ClickHouse | 585,760,121 | 8,924 | alexey-milovidov | @den-crane Could you please submit these queries to performance tests? | 2020-02-13T13:44:11 | 2020-02-13T13:44:11 | {} | MEMBER |
ClickHouse | ClickHouse | 586,143,111 | 9,095 | yuzhichang | @artbeglaryan Try the following in `dbms/src/Core/Settings.h`: `min_insert_block_size_rows`, `min_insert_block_size_bytes`.
`StorageDistributedDirectoryMonitor::Batch::isEnoughSize()` reads those settings. | 2020-02-14T08:01:24 | 2020-02-14T08:03:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,178,182 | 9,110 | qoega | What about
> ## ALTER {#query_language_queries_alter}
We should change it to ALTER TABLE using the same logic? | 2020-02-14T09:35:39 | 2020-02-14T09:35:39 | {} | MEMBER |
ClickHouse | ClickHouse | 586,214,624 | 9,046 | 4ertus2 | > Artem,
>
> does "select distinct " guarantee unique results?
Yup, you're right. I've missed `distinct` in your query. | 2020-02-14T11:09:59 | 2020-02-14T11:09:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,291,489 | 9,118 | 4ertus2 | I think the problem is in `if` + `neighbor` functions. ORDER BY should make `neighbor` stable but you've got wrong neighbors. We have to disable it automatically in case of statefull functions (cause it's not OK to push down predicates in this case). And it's probably already done in newer versions. | 2020-02-14T13:39:40 | 2020-02-14T13:45:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,748,561 | 9,094 | alexey-milovidov | Should be easy to fix. | 2020-02-13T13:15:14 | 2020-02-13T13:15:14 | {} | MEMBER |
ClickHouse | ClickHouse | 585,935,837 | 8,573 | filimonov | That means the problem introduced by
https://github.com/ClickHouse/ClickHouse/pull/8128
Was successfully fixed by
https://github.com/ClickHouse/ClickHouse/pull/9026 | 2020-02-13T19:38:15 | 2020-02-13T19:38:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,117,553 | 8,085 | ghost | @TisonKun how did you fix this one? | 2020-02-12T09:39:11 | 2020-02-12T09:39:11 | {} | NONE |
ClickHouse | ClickHouse | 585,504,446 | 9,084 | RangerWolf | @den-crane
So many thanks! I will have a try.
but one thing I am not clear is that, you metioned `insert into + optimize`
do you mean trigger `optimize` just after `insert into`?
Thanks again! | 2020-02-13T01:31:28 | 2020-02-13T01:31:28 | {} | NONE |
ClickHouse | ClickHouse | 585,686,026 | 9,075 | nikitamikhaylov | You can't see this changes, because I'm currently working on this task and don't publish commits. | 2020-02-13T11:28:15 | 2020-02-13T11:28:15 | {} | MEMBER |
ClickHouse | ClickHouse | 585,756,754 | 4,727 | qoega | Resolved in #6055 and #7215 | 2020-02-13T13:35:50 | 2020-02-13T13:35:50 | {} | MEMBER |
ClickHouse | ClickHouse | 586,296,405 | 9,118 | aligin | Thanks for help. | 2020-02-14T13:53:19 | 2020-02-14T13:53:19 | {} | NONE |
ClickHouse | ClickHouse | 586,444,164 | 8,943 | akuzm | The functional test failure doesn't look like it's related to system logs, and we've seen on previous commits that performance test are not influenced by this PR, as expected. Will merge now. | 2020-02-14T19:41:33 | 2020-02-14T19:41:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 584,828,015 | 9,072 | den-crane | two hours idleness after benchmarks
| 2020-02-11T20:10:47 | 2020-02-11T20:10:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,893,632 | 9,097 | 4ertus2 | Is `dws_data.pay_hour_pgcr_stat` a Replicated table? What do you have in `system.repllication_queue` and `system.merges` on both nodes before `insert select`?
The ALTER DELETE command is not a transaction. It finishes when accordng queues would finish. | 2020-02-13T18:05:58 | 2020-02-13T18:06:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,048,760 | 9,111 | den-crane | предредыщуй вариант звучал так:
Возвращает то же значение, которое использовалось в качестве аргумента.
Используется для отладки и тестирования, позволяет "сломать" доступ по индексу, и получить результат и производительность запроса для полного сканирования. | 2020-02-14T01:08:12 | 2020-02-14T01:08:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,051,906 | 8,968 | den-crane | I've never downloaded that file, I don't speak Chinese. | 2020-02-14T01:21:19 | 2020-02-14T01:21:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,320,247 | 9,124 | LittleMaverick | @4ertus2
Could you please clarify, which words are special and should be masked by back quotes? | 2020-02-14T14:50:17 | 2020-02-14T15:17:50 | {} | NONE |
ClickHouse | ClickHouse | 585,018,271 | 9,083 | den-crane | Sequentially. Though several mutations can be combined into one. But still data will be updated as updates applied sequentially.
>How does the ClickHouse keep data integrity
Data will be consistent after all updates ended (eventually). But during mutations your selects will see partially updated data because muta... | 2020-02-12T04:08:20 | 2020-02-12T04:08:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,036,211 | 9,084 | den-crane | >do you mean trigger optimize just after insert into
Yes. Or you can use `from final`.
SELECT count(*) FROM ads_account **FINAL**
It will work fine against such small table. | 2020-02-14T00:18:01 | 2020-02-14T00:18:01 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,309,679 | 9,124 | 4ertus2 | It's zone of uncertainty should we support aliases that have names of functions or types. Specifically in this case it's an error cause we allow to have alias in one case but disallow in another. But it's probably OK either to disable or to enable it if it has the same behaviour in all cases. It's better to use back qu... | 2020-02-14T14:23:59 | 2020-02-14T14:25:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,356,395 | 9,124 | 4ertus2 | INTERVAL is SQL data type and could be parsed in special way in some context. If you make a column or alias with name of a type, e.x. `select x as UInt32` or function `select x as round` you may have troubles without back quotes. | 2020-02-14T16:12:22 | 2020-02-14T16:13:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,665,789 | 9,075 | nvartolomei | > By splitting each partition you can increase the effectiveness (you won't drop huge amount of data).
True, but I can't see any changes in this PR that would change this behaviour.
> Do you use copier? Do you disable merges before copying?
Used it a few times to split clusters. Didn't disable merges. | 2020-02-13T10:49:08 | 2020-02-13T10:49:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,752,970 | 3,718 | qoega | still relevant | 2020-02-13T13:26:32 | 2020-02-13T13:26:32 | {} | MEMBER |
ClickHouse | ClickHouse | 586,283,434 | 9,118 | 4ertus2 | > I asume that ClickHouse optomizes query, and do where condition in subquery.
There's `enable_optimize_predicate_expression` that enables this optimisation.
I'll try to reproduce your issue. | 2020-02-14T13:16:27 | 2020-02-14T13:16:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,020,350 | 9,083 | chu1070y | Thank you. | 2020-02-12T04:17:50 | 2020-02-12T04:17:50 | {} | NONE |
ClickHouse | ClickHouse | 585,035,753 | 9,085 | BayoNet | @4ertus2 look at the changes please. | 2020-02-12T05:24:21 | 2020-02-12T05:24:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,690,707 | 9,099 | alexey-milovidov | Maybe it can be trivially implemented with hook in getTable method in query context, that will store this data in the query context. Thoughts @tavplubix ? | 2020-02-13T11:31:00 | 2020-02-13T11:31:00 | {} | MEMBER |
ClickHouse | ClickHouse | 586,215,576 | 8,334 | wesm | Per comments on the Arrow JIRA, if someone can clarify what you mean by "This issue was caused by lack of list type support in the used version of arrow lib." We do support reading ListArray so long as the child type is either a list or a primitive type (but not struct). The main JIRA for this is
https://issues.apa... | 2020-02-14T11:13:00 | 2020-02-14T11:13:00 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 586,262,393 | 9,096 | tavplubix | `Cannot parse expression of type String here: 'Rostov'` indeed caused by mismatching of columns count (parser expects `,` after `'Rostov'`, not `)`) and we should make exception message clear.
Default expressions in `Values` format can be inserted this way:
```
INSERT INTO session_temp (ClientID, FirstName, Addre... | 2020-02-14T12:10:57 | 2020-02-14T12:10:57 | {} | MEMBER |
ClickHouse | ClickHouse | 584,834,808 | 9,081 | den-crane | >how does it explain?
Your luckiness https://clickhouse.tech/docs/en/operations/table_engines/replication/
Read _"Data blocks are deduplicated."_ part.
---
BTW, never use legacy syntax. It can be canceled in the future.
WRONG:
```
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/default/cl... | 2020-02-11T20:26:40 | 2020-02-11T20:26:40 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,037,155 | 9,045 | karlchan-cn | Sorry for some wrong description.
The problems is I want to delete data in server B&C on table hits_local in just one SQL execution with script like below
#ALTER TABLE hits_local on cluster cluster_test DROP PARTITION 20200114;#.
@den-crane Thanks for your kindly answering ,i will try the new version and check th... | 2020-02-12T05:30:35 | 2020-02-12T05:30:35 | {} | NONE |
ClickHouse | ClickHouse | 585,249,840 | 9,087 | 4ertus2 | You may look at system.merges and system.replicated_queue tables content. You may have too much inserts or too small inserts or too slow hardware.
The idea is you need progress in both queues: merges and replication. It's possible if you have enough hardware and optimezed inserts (big portions of data). It looks lik... | 2020-02-12T15:06:54 | 2020-02-12T15:07:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,666,370 | 9,064 | grootman2007 | @CurtizJ I tried version 20.1.2.4 (cause could not find 20.1.4) - it has the same problem:
<pre><code>
2020.02.13 13:38:04.612055 [ 106 ] {} <Fatal> BaseDaemon: ########################################
2020.02.13 13:38:04.612147 [ 106 ] {} <Fatal> BaseDaemon: (version 20.1.2.4 (official build)) (from thread 5... | 2020-02-13T10:50:31 | 2020-02-13T10:50:31 | {} | NONE |
ClickHouse | ClickHouse | 585,889,487 | 9,098 | 4ertus2 | By default it use all shards cause we have no guaranty that could omit some of them. But there's a setting `optimize_skip_unused_shards - Assumes that data is distributed by sharding_key. Optimization to skip unused shards if SELECT query filters by sharding_key.` | 2020-02-13T17:57:15 | 2020-02-13T17:57:15 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,011,358 | 9,046 | den-crane | reproduced: 20.2.1.2362, 19.17.4 and **19.1.6** (original optimize_skip_unused_shards implementation).
STR:
at two shards:
```
CREATE TABLE table1_local (id String) ENGINE = MergeTree ORDER BY tuple();
CREATE table table1_d as table1_local ENGINE =
Distributed(test_cluster_two_shards, currentDatabase(), ... | 2020-02-13T22:51:29 | 2020-02-13T23:52:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,158,254 | 9,081 | hatrena | It works perfectly fine now. Thanks for the help | 2020-02-12T11:17:47 | 2020-02-12T11:17:47 | {} | NONE |
ClickHouse | ClickHouse | 585,245,761 | 9,086 | 4ertus2 | Most probable you use too old version. Could you reproduce it on 20.1? | 2020-02-12T14:58:26 | 2020-02-12T14:58:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,385,050 | 9,092 | den-crane | Hm,
I installed 20.1.2.4, No issue
```
ClickHouse client version 20.1.2.4 (official build).
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 20.1.2 revision 54431.
CREATE TABLE my_table
(
`purchase_id` UInt32,
`pcode_series_id` Nullable(UInt32),
`purchase_... | 2020-02-12T19:47:44 | 2020-02-12T19:48:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,781,068 | 9,075 | nikitamikhaylov | And now you can't see significant changes because of "refactoring" and big diff. It was too difficult for me to understand huge single file. I will split this PR into two. | 2020-02-13T14:27:14 | 2020-02-13T14:27:14 | {} | MEMBER |
ClickHouse | ClickHouse | 586,305,544 | 9,124 | LittleMaverick | SELECT number AS interval FROM (SELECT number FROM temp.keyword_test) - ok. But if add one more SELECT FROM - fail. Is it expected behavior? @4ertus2 | 2020-02-14T14:14:24 | 2020-02-14T14:14:24 | {} | NONE |
ClickHouse | ClickHouse | 586,433,582 | 9,132 | CurtizJ | Same as #9064. Fixed in master | 2020-02-14T19:20:26 | 2020-02-14T19:20:26 | {} | MEMBER |
ClickHouse | ClickHouse | 585,112,147 | 8,568 | amarazm | @dingxiangfei2009
You trying to freshen package from one repo with package from different repo with different distributor.
You have to uninstall old version ad install new packages.
Or you could try --force flag to replace files\packages
For example: rpm --force -Uvh *19.17*
| 2020-02-12T09:26:22 | 2020-02-12T09:26:22 | {} | NONE |
ClickHouse | ClickHouse | 585,199,675 | 9,084 | RangerWolf | The root cause is found. It is because we use AWS VM and the iops is slow. It is only 150
After increase to 2500, everything works OK | 2020-02-12T13:15:45 | 2020-02-12T13:15:45 | {} | NONE |
ClickHouse | ClickHouse | 585,254,843 | 8,936 | qoega | Feel free to reopen issue or make another one if problem remains | 2020-02-12T15:17:22 | 2020-02-12T15:17:32 | {} | MEMBER |
ClickHouse | ClickHouse | 585,265,095 | 9,086 | den-crane | @onine007 meantime you can use https://clickhouse.tech/docs/en/query_language/alter/#alter_attach-partition-from | 2020-02-12T15:37:48 | 2020-02-12T15:37:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,906,787 | 9,088 | 4ertus2 | Cannot reproduce on master build. Could you update to the last release and reproduce it? | 2020-02-13T18:37:14 | 2020-02-13T18:37:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 585,932,226 | 9,101 | banburybill | Apologies. Very dumb of me to omit that. Tested on 19.16.4.12, 19.16.8.34 and 20.1.4.14. | 2020-02-13T19:29:56 | 2020-02-13T19:29:56 | {} | NONE |
ClickHouse | ClickHouse | 585,965,803 | 9,088 | den-crane | reproduced 19.13.1.11, 19.13.6.51, 19.17.8.54 , 20.1.2, 20.2.1.2343, 20.2.1.2362 -- today nightbuilt
cat t_del.sql
```
drop table if exists default.test2;
CREATE TABLE default.test2
(
`EventDate` Date,
`CounterID` Nullable(String),
`UserID` Nullable(UInt32)
)
ENGINE = MergeTree()
ORDER BY E... | 2020-02-13T20:51:47 | 2020-02-13T20:57:57 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 586,061,821 | 9,060 | Zhile | So for this case, I didn't kill the mutation by query but wait for the mutations finish.
There's a setting `optimize_throw_if_noop` which will give me an error if the OPTIMIZE doesn't do anything. Then I wait and retry until the OPTIMIZE complete. This way helps me to do the OPTIMIZE. | 2020-02-14T02:03:30 | 2020-02-14T02:03:30 | {} | NONE |
ClickHouse | ClickHouse | 586,364,850 | 6,128 | alexey-milovidov | Yes. | 2020-02-14T16:32:31 | 2020-02-14T16:32:31 | {} | MEMBER |
ClickHouse | ClickHouse | 585,743,655 | 9,099 | alexey-milovidov | @tavplubix said that it's better to extract them in AST level. | 2020-02-13T13:01:50 | 2020-02-13T13:01:50 | {} | MEMBER |
ClickHouse | ClickHouse | 586,302,271 | 9,124 | 4ertus2 | It's OK to have special words. They should be masked by back quotes. But there's a bug with one of notation.
```
SELECT * FROM (SELECT number AS `interval` FROM (SELECT number FROM numbers(1))); -- OK
SELECT s.`interval` FROM (SELECT number AS `interval` FROM (SELECT number FROM numbers(1))) s; -- OK
SELECT `in... | 2020-02-14T14:05:58 | 2020-02-14T14:08:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 583,352,160 | 8,964 | CurtizJ | Yes, it's ok. | 2020-02-07T11:30:55 | 2020-02-07T11:30:55 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.