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 | 752,137,326 | 18,610 | den-crane | replication_queue is a virtual table. It exists for users. CH itself does not use this table. | 2020-12-29T16:05:36 | 2020-12-29T16:05:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,204,001 | 2,914 | filimonov | Do we want to make further implovements here? Or we should close it? | 2020-12-29T18:54:37 | 2020-12-29T18:54:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,220,984 | 2,914 | alexey-milovidov | Yes:
- it should be enabled automatically;
- it should support injective functions. | 2020-12-29T19:52:11 | 2020-12-29T19:52:11 | {} | MEMBER |
ClickHouse | ClickHouse | 751,739,292 | 18,559 | den-crane | @Sallery-X Do you have logs with messages about "fd14b7e6c0c6610e8422d56580fb952a_1_10_2|fd14b7e6c0c6610e8422d56580fb952a_6_30_3" before time 2020.12.28 10:29:03.486666 ? | 2020-12-28T14:58:59 | 2020-12-28T14:59:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,710,020 | 18,568 | alexey-milovidov | This is experimental task, it is assigned to @Realist007. | 2020-12-28T13:19:41 | 2020-12-28T13:19:49 | {} | MEMBER |
ClickHouse | ClickHouse | 751,714,107 | 18,464 | alexey-milovidov | @hexiaoting One test should be adjusted.
We have two ways to do it:
1. Introduce a setting `allow_floating_point_partition_key`, 0 by default. Set it to 1 in this test. The setting will be also useful for some special cases.
2. Change `/` operator to `intDiv` in the test. | 2020-12-28T13:33:50 | 2020-12-28T13:33:50 | {} | MEMBER |
ClickHouse | ClickHouse | 751,780,988 | 18,189 | alexey-milovidov | `((sipHash64(toString(toDate(p))) % NULL)` | 2020-12-28T16:43:20 | 2020-12-28T16:43:20 | {} | MEMBER |
ClickHouse | ClickHouse | 751,823,862 | 18,568 | den-crane | https://crate.io/docs/sql-99/en/latest/chapters/27.html
```sql
CREATE TYPE book_udt AS -- the UDT name will be book_udt
title CHAR(40), -- title is the first attribute
buying_price DECIMAL(9,2), -- buying_price is the second attribute
selling_price DECIMAL(9,2) ... | 2020-12-28T18:38:47 | 2020-12-28T18:38:47 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,870,398 | 17,184 | ygrek | updated
| 2020-12-28T21:30:22 | 2020-12-28T23:56:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,954,023 | 18,561 | BhavyaRajSharma | @den-crane,
Thanks for the information.
This feature is not enabled because of not enough use cases or any technical challenges? | 2020-12-29T05:40:58 | 2020-12-29T05:41:23 | {} | NONE |
ClickHouse | ClickHouse | 752,015,944 | 18,587 | alexey-milovidov | See the `part_type` column in `system.parts`. | 2020-12-29T09:49:45 | 2020-12-29T09:49:45 | {} | MEMBER |
ClickHouse | ClickHouse | 752,026,976 | 18,586 | l1t1 | thanks | 2020-12-29T10:27:58 | 2020-12-29T10:27:58 | {} | NONE |
ClickHouse | ClickHouse | 752,096,744 | 18,588 | den-crane | There is no easy tricks.
You can use clickhouse-copier -- it will re-insert everything.
| 2020-12-29T14:29:16 | 2020-12-29T14:29:16 | {
"hooray": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,106,404 | 18,589 | abyss7 | Please, provide more information on how exactly you are trying to store the data into CH. Also provide CH version. | 2020-12-29T14:52:17 | 2020-12-29T14:52:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,788,583 | 16,059 | nikitamikhaylov | I think this is the same as #14700 | 2020-12-28T17:03:20 | 2020-12-28T17:03:20 | {} | MEMBER |
ClickHouse | ClickHouse | 751,845,264 | 18,518 | alesapin | Yes, now it's fixed | 2020-12-28T19:55:42 | 2020-12-28T19:55:42 | {} | MEMBER |
ClickHouse | ClickHouse | 751,882,456 | 18,563 | filimonov | > ENGINE = MergeTree()
I guess you have 2 Engine=Kafka tables, 2 MV and 2 target tables... I need schema of all 6. Especially interested in SETTINGS of engine=kafka.
> And the errors from clickhouse-server.log
Nothing related to Kafka. I need not only errors but trace logs.
> As far as I know, this erro... | 2020-12-28T22:23:13 | 2020-12-28T22:23:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,998,110 | 18,517 | ccellado | I'm sorry, Flink indeed does not "smart select" from a list - you have pick one manually.
Closing the request. | 2020-12-29T08:44:25 | 2020-12-29T08:44:25 | {} | NONE |
ClickHouse | ClickHouse | 752,064,227 | 16,833 | tavplubix | > In my docker , i only just run one "mysql" , it means that i need to set up mysql cluster? Can the 'ch' synchronize one "mysql"?
Yes, MaterializeMySQL supports replication from single MySQL server, you don't need to setup a cluster.
> i update the data of mysql ... and the tables of clickhouse has not change
... | 2020-12-29T12:50:39 | 2020-12-29T12:50:39 | {} | MEMBER |
ClickHouse | ClickHouse | 752,109,996 | 18,188 | nikitamikhaylov | Merged | 2020-12-29T15:01:01 | 2020-12-29T15:01:01 | {} | MEMBER |
ClickHouse | ClickHouse | 751,686,625 | 18,074 | filimonov | Looking on it... (had short vacations) | 2020-12-28T11:50:46 | 2020-12-28T11:50:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,786,915 | 16,059 | nikitamikhaylov | Let's close it until next occurrence. | 2020-12-28T16:59:11 | 2020-12-28T16:59:11 | {} | MEMBER |
ClickHouse | ClickHouse | 751,993,159 | 18,465 | ghost | @alex-zaitsev
It would be nice to improve the accuracy with the TPUT algorithm.
However, as you mentioned, it requires the double roundtrip, so how about implementing that feature after this draft?
Thanks for your comment. | 2020-12-29T08:25:08 | 2020-12-29T08:25:08 | {} | NONE |
ClickHouse | ClickHouse | 752,017,665 | 18,464 | alexey-milovidov | Yes, it will be perfect. | 2020-12-29T09:55:34 | 2020-12-29T09:55:34 | {} | MEMBER |
ClickHouse | ClickHouse | 752,059,327 | 18,542 | nautaa | @alexey-milovidov PTAL. | 2020-12-29T12:31:32 | 2020-12-29T12:31:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,090,010 | 16,503 | Akazz | Note that `DB::CurrentThread::detachQueryIfNotDetached()` cannot be found as well. I guess it is a matter of where you placed this attach/detach logics. Currently as it stands there is a dependency Interpreters -> common_io (`DB::CurrentThread::attachTo()` resides in `ThreadStatusExt.cpp`, which belongs to Interpreters... | 2020-12-29T14:13:15 | 2020-12-29T14:13:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,101,576 | 18,594 | den-crane | CH is designed for instances with 32GB+ RAM.
It can work with less memory instances. But it's a constant straggle with swap / OOM / memory limit errors, super slow queries.
Probably the supported minimum is 4GB+
>In that case would that be the largest csv I can load as a hashed dictionaty on CH or can I modify t... | 2020-12-29T14:41:15 | 2020-12-29T14:41:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,843,671 | 18,561 | den-crane | >Is there any way (enabling some configs) I can use prewhere with the views?
No way.
>or is it planned for future?
AFAIK no plans. | 2020-12-28T19:49:48 | 2020-12-28T19:49:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,852,384 | 18,553 | tavplubix | AST fuzzer - #13281 | 2020-12-28T20:21:12 | 2020-12-28T20:21:12 | {} | MEMBER |
ClickHouse | ClickHouse | 751,924,975 | 18,452 | meshpaul | Hmm, I think this is best to be controlled via CH Proxy and extra infrastrucure. Where one can separate `insert` from `select` and mix of K8 operator to be merged as part of same persisted/replicated storage. I am not so sure one can split read vs write effectively but with use of replication and separating cluster i... | 2020-12-29T02:34:03 | 2020-12-29T02:53:42 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 751,954,734 | 8,703 | chenziliang | May i ask if static AST analysis and rewrite is the correct direction for this kind of optimization or it is better to tackle this problem in pipeline execution ? | 2020-12-29T05:44:07 | 2020-12-29T05:44:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,065,813 | 14,735 | tavplubix | > How about MySQL enum type?
> Exception: Unknown data type family: enum. Maybe you meant: ['Enum','Enum8']
#18310 | 2020-12-29T12:56:43 | 2020-12-29T12:56:43 | {} | MEMBER |
ClickHouse | ClickHouse | 657,486,437 | 4,006 | zhang2014 | To be compatible SQL:
- [ ] Support YEAR, TIME type
- [x] TRUNCATE database.table;
- [ ] DROP TABLE database.table, table_2, ...;
- [x] ~Support ATTACH DATABASE without engine def~ (implemented)
- [x] Add factor with DateTime type
- [x] Support MySQL DateTime factor into clickhouse
- [x] ALTER TABLE table_name M... | 2020-07-13T10:48:59 | 2020-12-29T13:12:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,175,850 | 18,610 | filimonov | Do you use insert_quorum? | 2020-12-29T17:34:37 | 2020-12-29T17:34:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,191,378 | 18,610 | simPod | It's set to 0 in `system.settings`
```
name value
insert_quorum 0
insert_quorum_parallel 0
``` | 2020-12-29T18:17:04 | 2020-12-29T18:17:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,200,282 | 18,610 | alexey-milovidov | > Do not update all the servers on a cluster simultaneously, or the cluster will be unavailable for some time.
but this is kinda what should have been done this time.
Install new version. Restart half of the replicas. After they are up, restart other replicas.
| 2020-12-29T18:43:36 | 2020-12-29T18:43:36 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 751,688,701 | 18,074 | olgarev | @filimonov Thanks! | 2020-12-28T11:59:27 | 2020-12-28T11:59:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,962,830 | 18,102 | filimonov | Да, можно просто закомментировать, но best practice - это не менять конфигурационные файлы поставляемые с дистрибутивом, а производить все изменения добавляя отдельные файлы для настройки того или иного или иного поведения.
Например, в этом случае можно создать файл
`/etc/clickhouse-server/config.d/disable_metric... | 2020-12-29T06:21:56 | 2020-12-29T06:21:56 | {
"rocket": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,966,958 | 18,559 | Sallery-X | > @Sallery-X Do you have logs with messages about "fd14b7e6c0c6610e8422d56580fb952a_1_10_2|fd14b7e6c0c6610e8422d56580fb952a_6_30_3" before time 2020.12.28 10:29:03.486666 ?
2020.12.28 11:41:11.482630 [ 65 ] {} bc_online.ck_content_basic_info_pvuv: DB::StorageReplicatedMergeTree::queueTask()::<lambda(DB::StorageRepli... | 2020-12-29T06:39:33 | 2020-12-29T06:39:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,241,351 | 18,612 | den-crane | https://dev.mysql.com/doc/internals/en/date-and-time-data-type-representation.html
```
TIME: A three-byte integer packed as DD×24×3600 + HH×3600 + MM×60 + SS
1 bit sign (1= non-negative, 0= negative)
1 bit unused (reserved for future extensions)
10 bits hour (0-838)
6 bits minute (0-59)
6 bits seco... | 2020-12-29T21:07:51 | 2020-12-29T21:07:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,718,495 | 18,102 | AnaUvarova | > @alexey-milovidov Hi, can you, please, answer the question above? Do we need to describe when and how this feature can be disabled?
@olgarev
Привет! Есть инфа от Никиты "выключить можно в config.xml просто закомментировав все что связано с metric_log", но мне кажется, что этого может не хватит для красивой инст... | 2020-12-28T13:49:24 | 2020-12-28T13:49:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,719,356 | 18,569 | alexey-milovidov | `contrib` libraries are not meant to be built separately. They are build with ClickHouse. | 2020-12-28T13:52:28 | 2020-12-28T13:52:28 | {} | MEMBER |
ClickHouse | ClickHouse | 751,739,529 | 18,524 | alexey-milovidov | > Integration tests (thread) — fail: 15, passed: 890, error: 11
Infrastructure error:
`Cannot reserve 1.00 MiB, not enough space.` | 2020-12-28T14:59:33 | 2020-12-28T14:59:33 | {} | MEMBER |
ClickHouse | ClickHouse | 751,739,649 | 18,524 | alexey-milovidov | > Stress test (thread) — Sanitizer assert (in stderr.log)
Already known issue with coroutines (unrelated). | 2020-12-28T14:59:56 | 2020-12-28T14:59:56 | {} | MEMBER |
ClickHouse | ClickHouse | 751,272,475 | 18,189 | nikitamikhaylov | This is a bit strange, because this query doesn't contain function `modulo`.
```
2020.12.17 05:35:47.576023 [ 66 ] {0aafa923-375a-4d36-94ad-b63780de159f} <Debug> executeQuery: (from [::1]:42244, using old parser) SELECT toDateTime(toDateTime(NULL), 10.000100135803223, NULL, NULL), count(-2) FROM test1 WHERE (toDate(... | 2020-12-25T17:08:58 | 2020-12-28T16:36:41 | {} | MEMBER |
ClickHouse | ClickHouse | 752,016,632 | 18,586 | alexey-milovidov | I usually paste query to text editor (e.g. Kate), then replace tabs to spaces, and only then paste to clickhouse-client.
Also, if you invoke `clickhouse-client` with `--disable_suggestion` or `-A` option, it will not use autocomplete.
PS. @amosbird maybe we can disable autocomplete during bracketed paste? | 2020-12-29T09:52:12 | 2020-12-29T09:52:12 | {} | MEMBER |
ClickHouse | ClickHouse | 752,021,613 | 18,454 | zlx19950903 | What slows down the speed of regular expression? It should be quite fast. | 2020-12-29T10:09:20 | 2020-12-29T10:09:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,179,991 | 18,595 | alexey-milovidov | TSan: #18385 | 2020-12-29T17:46:15 | 2020-12-29T17:46:15 | {} | MEMBER |
ClickHouse | ClickHouse | 751,738,530 | 18,560 | den-crane | @Sallery-X your link points to issue not to a picture. | 2020-12-28T14:57:02 | 2020-12-28T14:57:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,884,613 | 18,102 | olgarev | > @olgarev
> Привет! Есть инфа от Никиты "выключить можно в config.xml просто закомментировав все что связано с metric_log", но мне кажется, что этого может не хватит для красивой инструкции. Как думаешь?
@AnaUvarova Привет. Да нет, по-моему норм ) | 2020-12-28T22:33:58 | 2020-12-28T22:33:58 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,040,777 | 18,586 | amosbird | @alexey-milovidov https://github.com/ClickHouse/ClickHouse/issues/7757 | 2020-12-29T11:19:00 | 2020-12-29T11:19:00 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 752,051,968 | 17,114 | tavplubix | > clickHouse的数据分区不是以mysql的主键和索引分布，分区是以 all_开头分区的
> 需要怎么调整mysql的索引才能在clickHouse的数据能分区。
Machine translation:
> The data partition of clickHouse is not distributed by the primary key and index of mysql, the partition is partitioned at the beginning of all_
> How to adjust the mysql index to partition the data in cli... | 2020-12-29T12:03:24 | 2020-12-29T12:03:24 | {} | MEMBER |
ClickHouse | ClickHouse | 752,144,384 | 18,455 | akuzm | Perf test failure is OK (new grammar support added).
The integration test failure looks unrelated. | 2020-12-29T16:22:40 | 2020-12-29T16:23:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,692,556 | 18,517 | filimonov | It's just for HA? I.e. all registries are the same?
Can you share some official docs / samples of how it was implemented in Flink (didn't find, sorry). | 2020-12-28T12:13:43 | 2020-12-28T12:13:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,724,558 | 18,285 | weeds085490 | @alexey-milovidov Any suggestions about this problem? We encountered this problem in our production environment. PTAL
PTAL | 2020-12-28T14:09:40 | 2020-12-28T14:09:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,843,966 | 18,564 | den-crane | never worked. Not a bug | 2020-12-28T19:50:58 | 2020-12-28T19:50:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,900,998 | 18,568 | alexey-milovidov | @filimonov Let's drop the proposal for parametrized UDT. | 2020-12-29T00:04:50 | 2020-12-29T00:04:50 | {} | MEMBER |
ClickHouse | ClickHouse | 751,918,432 | 15,806 | meshpaul | 🥇 | 2020-12-29T01:54:29 | 2020-12-29T01:54:29 | {} | NONE |
ClickHouse | ClickHouse | 752,003,238 | 18,563 | MaryamJFard | Yes you are right there are 6 tables; both of the target tables are the same as the one I pasted here to my previous comment, so in this one, the following is 2 Engine=Kafka tables and 2 MV:
CREATE TABLE db.kafka_1
(
`clmn_1` UInt64,
`clmn_2` UInt64,
`clmn_3` Int8,
`clmn_4` Nullable(String),
... | 2020-12-29T09:03:47 | 2020-12-29T09:03:47 | {} | NONE |
ClickHouse | ClickHouse | 752,015,735 | 18,587 | alexey-milovidov | It may happen if all data parts are compact or in-memory - these data parts don't have per column statistics. | 2020-12-29T09:49:03 | 2020-12-29T09:49:03 | {} | MEMBER |
ClickHouse | ClickHouse | 752,032,433 | 18,586 | amosbird | > PS. @amosbird maybe we can disable autocomplete during bracketed paste?
It already does that. `bracketed_paste_mode` will switch to a different terminal state machine. However, given the above example, @l1t1 uses multiline mode and it doesn't have `bracketed_paste_mode` enabled. | 2020-12-29T10:48:02 | 2020-12-29T10:48:02 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 752,048,697 | 18,454 | alexey-milovidov | @zlx19950903
1. There are two invocations of regular expressions, it means two passes over memory instead of one.
2. The implementation with regular expressions is not quite correct (even worse than what is proposed).
3. Performance of regular expressions depends on the engine. ClickHouse has support for two engi... | 2020-12-29T11:50:51 | 2020-12-29T11:51:41 | {} | MEMBER |
ClickHouse | ClickHouse | 752,094,464 | 18,594 | den-crane | >Application: Setting max_server_memory_usage was set to 1.61 GiB (1.79 GiB available * 0.90 max_server_memory_usage_to_ram_ratio)
Please provide an output of `free -h`
CH thinks that your hardware has less than 2GB RAM,
I think it because of some virtual properties of Scaleway instances, memory ballooning maybe... | 2020-12-29T14:23:31 | 2020-12-29T14:26:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,148,196 | 18,610 | simPod | Ah, right. Though something was the reason of
> DB::Exception: Cannot parse input: expected 'quorum: ' before: 'merge_type: 2\n'
some structure somewhere must have changed so those versions were incompatible.
The cluster is now on 20.12 and everything is running though there was some BC change that was not men... | 2020-12-29T16:32:23 | 2020-12-29T16:32:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,159,590 | 18,613 | den-crane | > 11.34 billion rows, 90.72 GB
> 11.34 billion rows, 1.29 TB
AFAIK The cause is in Url column size. CH needs to read and uncompress files.
Try to materialize domain as a real column.
Try to use LowCardinality.
| 2020-12-29T16:59:10 | 2020-12-29T16:59:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,687,498 | 18,446 | akuzm | Why is `CAST` special? Other functions also fail when they shouldn't:
```
*2* :) EXPLAIN SYNTAX SELECT (SELECT * FROM system.numbers LIMIT 1 OFFSET 1) AS n, intDiv(10, n);
Received exception from server (version 20.13.1):
Code: 153. DB::Exception: Received from localhost:9000. DB::Exception: Division by zero: W... | 2020-12-28T11:54:25 | 2020-12-28T11:54:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,740,621 | 17,653 | EasonEdison | watch | 2020-12-28T15:02:15 | 2020-12-28T15:02:15 | {} | NONE |
ClickHouse | ClickHouse | 751,984,557 | 18,333 | Akazz | I'd rather add some example as well illustrating the effects. Like the following:
```sql
SELECT
SUM(-1),
MAX(0)
FROM system.one
WHERE 0
SETTINGS aggregate_functions_null_for_empty = 0
┌─SUM(-1)─┬─MAX(0)─┐
│ 0 │ 0 │
└─────────┴────────┘
```
```sql
SELECT
SUM(-1),
MAX(0)
FROM ... | 2020-12-29T07:49:18 | 2020-12-29T09:41:53 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,704,083 | 18,508 | filimonov | `maxIntersections` ? https://github.com/ClickHouse/ClickHouse/pull/2012 ? | 2020-12-28T12:58:34 | 2020-12-28T12:58:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,046,182 | 18,310 | tavplubix | > Maybe someone has an idea what caused this error?
Short answer: MaterializeMySQL does not support ENUM data type.
Long answer:
The error happens because data type names in ClickHouse are case-sensitive, and data type names in MySQL are case-insensitive, so ClickHouse does not have `enum` or `ENUM` data type, b... | 2020-12-29T11:40:37 | 2020-12-29T11:40:37 | {
"+1": 5,
"rocket": 1
} | MEMBER |
ClickHouse | ClickHouse | 751,885,882 | 18,568 | filimonov | It can also be a solution for #4420
> It can also be parametrized itself. Then parameters can be substituted to the parameters of the underlying type
c++ programmer way of thinking detected :)
the use case is not very clear, and I'm not sure for now if that extra complexity introduced by those 'templates' is r... | 2020-12-28T22:40:29 | 2020-12-28T22:40:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,085,854 | 18,542 | alexey-milovidov | I found one issue: if previous entity is malformed, then the entiry right after it is not decoded:
```
$ clickhouse-client
ClickHouse client version 20.13.1.1.
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 20.13.1 revision 54443.
milovidov-desktop :) SELECT decodeXMLCompo... | 2020-12-29T14:03:20 | 2020-12-29T14:03:20 | {} | MEMBER |
ClickHouse | ClickHouse | 752,090,939 | 18,609 | alexey-milovidov | You can do rolling update, but replication will stop during update; while SELECT queries will continue to work and INSERT queries continue to write data that will be replicated later after upgrade is finished.
Recommendations:
- do not keep different versions of replicas for a long time;
- use your testing environ... | 2020-12-29T14:15:21 | 2020-12-29T14:15:21 | {} | MEMBER |
ClickHouse | ClickHouse | 752,187,726 | 18,248 | sevirov | @Akazz Please, review this PR. | 2020-12-29T18:07:40 | 2020-12-29T18:07:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,679,194 | 18,518 | alesapin | Probably this failure with fuzzer is OK. Going to fix test. | 2020-12-28T11:19:54 | 2020-12-28T11:19:54 | {} | MEMBER |
ClickHouse | ClickHouse | 751,716,655 | 18,550 | alexey-milovidov | > Integration tests (thread) — fail: 1, passed: 1028, error: 1
Known flaky test, two failures on master in 10 days:
491f4817139fbe779f10a7c2ae6f36fe8bd4a47e
aba7b5f60d7285ea681111a081af0569352af990
FYI @nikitamikhaylov
| 2020-12-28T13:42:51 | 2020-12-28T13:43:10 | {} | MEMBER |
ClickHouse | ClickHouse | 751,825,661 | 18,568 | alexey-milovidov | @den-crane As we have named tuples, it will be more consistent to write:
```
CREATE TYPE book_udt AS
Tuple(
title CHAR(40),
buying_price DECIMAL(9,2),
selling_price DECIMAL(9,2)
)
``` | 2020-12-28T18:45:24 | 2020-12-28T18:45:35 | {} | MEMBER |
ClickHouse | ClickHouse | 751,959,745 | 18,587 | l1t1 | and sum(bytes) FROM system.parts returns other values
```
SELECT
table,
sum(bytes) / 1048576 AS size,
sum(rows) AS rows
FROM system.parts
WHERE active AND (database = 'default') AND (table LIKE '%Dict')
GROUP BY table
ORDER BY size DESC
┌─table──┬───────────────size─┬─────rows─┐
│ Dict │ 162.... | 2020-12-29T06:07:54 | 2020-12-29T06:07:54 | {} | NONE |
ClickHouse | ClickHouse | 751,966,965 | 18,508 | depressed-pho | > `maxIntersections` ? #2012 ?
To be honest I didn't know `maxIntersections` has existed because it was undocumented and was only mentioned in the changelog. `maxIntersections` is an aggregate function which finds a single data point whose concurrency (the number of intersections) becomes the maximum. On the other h... | 2020-12-29T06:39:34 | 2020-12-29T06:39:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,024,038 | 16,689 | zhangjmruc | I tried to investigate this issue, following are my findings:
1. From the explain plans for sum(null) and sum(cast(null as Nullable(Int32)), we can see the difference in the data type and name of AggregateFunction. In the problematic case, the name is **nothing** for only NULL case.
-- sum(null)
explain plan header=... | 2020-12-29T10:17:43 | 2020-12-29T10:17:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,036,085 | 18,333 | ka1bi4 | @Akazz Thanks for examples, I'll add it. | 2020-12-29T11:01:51 | 2020-12-29T11:01:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,104,045 | 18,594 | MaxJPowers | Thanks for the info! I'm just trying to get a better understanding of the relationship between query times, RAM, and number of rows/columns inserted for Clickhouse which forces me to work with both small and large servers. | 2020-12-29T14:46:34 | 2020-12-29T14:46:34 | {} | NONE |
ClickHouse | ClickHouse | 752,244,492 | 18,446 | azat | >Why is CAST special? Other functions also fail when they shouldn't:
CAST is "special" because is used for sub-queries results.
Your example works with this PR.
>We should probably embrace the fact that we don't know the result of a subquery sometimes, and stop performing constant folding in such cases, the same... | 2020-12-29T21:19:41 | 2020-12-29T21:19:41 | {} | MEMBER |
ClickHouse | ClickHouse | 751,688,725 | 18,446 | akuzm | Also the fact that we have `isSuitableForConstantFolding` as a separate method in addition to `isDeterministic` and `isDeterministicInScopeOfQuery` or something, and even `isStateful`, whatever that could mean, makes me sad...
We should really borrow the function volatility categories from postgres, they work well:
... | 2020-12-28T11:59:32 | 2020-12-28T12:00:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 751,982,600 | 18,465 | alex-zaitsev | The industry standard for accurate results is TPUT algorithm: http://crypto.stanford.edu/~cao/topk.pdf
But it requires double roundtrip. | 2020-12-29T07:40:09 | 2020-12-29T07:40:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,033,922 | 18,193 | tavplubix | >> `mysql> show create table ac.pp_wechatinfo;`
> ```
> CREATE TABLE pp_wechatinfo
> (
> id bigint(20) DEFAULT NULL,
> aid varchar(50) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci DEFAULT NULL,
> uid varchar(200) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci DEFAULT NULL,
> oid varchar(200) CHARA... | 2020-12-29T10:53:37 | 2020-12-29T10:53:37 | {} | MEMBER |
ClickHouse | ClickHouse | 752,046,916 | 18,403 | tavplubix | Could you please provide more information, how to reproduce the issue? | 2020-12-29T11:43:35 | 2020-12-29T11:43:35 | {} | MEMBER |
ClickHouse | ClickHouse | 752,168,206 | 11,279 | alexey-milovidov | It may happen if you edited the storage policy in configuration and removed/changed some of the disks.
Then some parts may remain on disk that no longer belongs to the storage policy of the table. | 2020-12-29T17:15:48 | 2020-12-29T17:15:48 | {} | MEMBER |
ClickHouse | ClickHouse | 752,169,251 | 18,613 | alexey-milovidov | > I thought ClickHouse would extract the domains for only 50 URLs.
There is no such optimization. It is possible to implement but it will require more complex query pipeline with two passes of the data. See https://github.com/ClickHouse/ClickHouse/issues/5329 | 2020-12-29T17:17:52 | 2020-12-29T17:17:52 | {} | MEMBER |
ClickHouse | ClickHouse | 752,208,988 | 11,617 | tavplubix | Does parallel formatting work with HTTP interface? | 2020-12-29T19:09:33 | 2020-12-29T19:09:33 | {} | MEMBER |
ClickHouse | ClickHouse | 752,239,622 | 18,612 | filimonov | https://dev.mysql.com/doc/refman/8.0/en/date-and-time-functions.html#function_timediff
It returns TIME type which does not exists in clickhouse :\ | 2020-12-29T21:01:20 | 2020-12-29T21:01:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,246,240 | 18,563 | filimonov | > I sent you all the logs for that special minute not only the errors.
It seems you decreased log level from default trace to error.
Please enable trace logging and try to reproduce the problem. | 2020-12-29T21:21:20 | 2020-12-29T21:21:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,033,904 | 18,586 | alexey-milovidov | I don't understand why multiline mode disables bracketed paste mode. | 2020-12-29T10:53:32 | 2020-12-29T10:53:32 | {} | MEMBER |
ClickHouse | ClickHouse | 752,067,516 | 14,441 | tavplubix | #16132 | 2020-12-29T13:03:09 | 2020-12-29T13:03:09 | {} | MEMBER |
ClickHouse | ClickHouse | 752,087,014 | 18,542 | alexey-milovidov | This PR is also incomplete, because it does not support decoding of numeric entities.
https://en.wikipedia.org/wiki/Numeric_character_reference | 2020-12-29T14:06:12 | 2020-12-29T14:06:12 | {} | MEMBER |
ClickHouse | ClickHouse | 752,096,210 | 18,401 | akuzm | > > Another interesting solution would be to just always set all the Loggers to 'trace' level, and then
>
> It's not even necessarily. You can have `send_logs_level` higher than global log level.
Nice, we have a special override in LOG_* macros that just completely ignores the levels set in particular loggers. It... | 2020-12-29T14:27:55 | 2020-12-29T14:27:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 752,097,659 | 18,594 | MaxJPowers | > > Application: Setting max_server_memory_usage was set to 1.61 GiB (1.79 GiB available * 0.90 max_server_memory_usage_to_ram_ratio)
>
> Please provide an output of `free -h`
> CH thinks that your hardware has less than 2GB RAM,
Here's the result:
```
total used free shared buff/cache avai... | 2020-12-29T14:31:35 | 2020-12-29T14:31:35 | {} | NONE |
ClickHouse | ClickHouse | 737,072,620 | 17,700 | alexey-milovidov | Need to update our test reference files. They are appeared to be specific for TestKeeper. | 2020-12-02T08:23:35 | 2020-12-02T08:23:35 | {} | MEMBER |
ClickHouse | ClickHouse | 737,379,841 | 17,709 | abyss7 | > I'll add a test later.
Waiting for test then. | 2020-12-02T17:29:23 | 2020-12-02T17:29:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 736,340,461 | 17,624 | helifu | I have met the same problem and my env is: v20.8.7.15-lts + gcc-10.2.0.
```
[ 56%] Building CXX object contrib/llvm/llvm/utils/TableGen/CMakeFiles/obj.llvm-tblgen.dir/CodeGenSchedule.cpp.o
In file included from /root/ClickHouse/contrib/libcxx/include/math.h:311,
from /root/ClickHouse/contrib/libc... | 2020-12-01T09:26:15 | 2020-12-03T03:40:23 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.