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 | 617,486,553 | 10,419 | st3rv04ka | This command
`cat data.csv | clickhouse-client --query="INSERT INTO db.database_test (product,id,date,type) SELECT col1, col2, '2020-01-04' AS date,'cheese' AS type FROM input('col1 String, col2 String') FORMAT CSV" --format_csv_delimiter=":" --compression 0 --format_csv_allow_double_quotes=0 --format_csv_allow_sing... | 2020-04-22T01:00:11 | 2020-04-22T01:00:11 | {} | NONE |
ClickHouse | ClickHouse | 617,692,170 | 10,200 | AlexanderJLiu | @KochetovNicolai
> Warnings with Connection failed probably say that remote server have also crashed.
Yes, it did. But it crashed with the different query executed. I'll explain below:
1. One `Grafana Dashboard` has several `Panels` point to different `db.distributed_table` in `Node A` (This cluster has 10 Node... | 2020-04-22T10:26:44 | 2020-04-22T10:28:13 | {} | NONE |
ClickHouse | ClickHouse | 617,935,982 | 10,444 | alexey-milovidov | Maybe we need to sample allocations and deallocations randomly (weighted by bytes) instead of using the change in peak memory usage. | 2020-04-22T17:57:15 | 2020-04-22T17:58:18 | {} | MEMBER |
ClickHouse | ClickHouse | 618,052,483 | 10,420 | alexey-milovidov | 16/16 builds are OK | 2020-04-22T21:33:33 | 2020-04-22T21:33:33 | {} | MEMBER |
ClickHouse | ClickHouse | 618,287,646 | 3,196 | 4ertus2 | No updates yet. Could you make more simple sample without replicated merge tree and configs? There're some samples for distributed queries in tests/queries/0_stateless. They use `Distributed(test_cluster_two_shards, currentDatabase(), ...)` pattern in tests. Could you make something like this to allow reproduce it in o... | 2020-04-23T09:21:37 | 2020-04-23T09:21:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,574,760 | 9,040 | filimonov | Fix #10071 | 2020-04-22T06:18:24 | 2020-04-22T06:18:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,643,752 | 10,432 | pic-Nick | Thank you! Did not find it in docs. | 2020-04-22T08:49:23 | 2020-04-22T08:49:39 | {} | NONE |
ClickHouse | ClickHouse | 617,755,266 | 10,410 | blinkov | @BayoNet lets fix other languages as well? | 2020-04-22T12:41:42 | 2020-04-22T12:41:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,790,545 | 10,438 | filimonov | Done intentionally in v20.1.2.4, due to some race condition found during switching log files.
> Require server to be restarted to apply the changes in logging configuration. This is a temporary workaround to avoid the bug where the server logs to a deleted log file (see #8696). #8707 (Alexander Kuzmenkov)
>
> htt... | 2020-04-22T13:48:56 | 2020-04-22T13:49:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,842,688 | 7,665 | Axelfoley85 | Thank you for checking so quickly!
I would hope exactly for the results you are getting for the three queries. If I run the `SEMI LEFT JOIN` on clickhouse version 19.15.3.6 I get the following result:
```
SELECT
c,
count()
FROM
(
SELECT
timestamp,
count() AS c
FROM
... | 2020-04-22T15:16:39 | 2020-04-22T15:16:39 | {} | NONE |
ClickHouse | ClickHouse | 617,546,522 | 9,099 | alexey-milovidov | Duplicate of #9030 | 2020-04-22T04:45:06 | 2020-04-22T04:45:06 | {} | MEMBER |
ClickHouse | ClickHouse | 617,774,823 | 10,440 | nikitamikhaylov | Added as many as possible options from
- https://github.com/google/sanitizers/wiki/AddressSanitizerFlags
- https://github.com/google/sanitizers/wiki/SanitizerCommonFlags
- (maybe useful) http://lists.llvm.org/pipermail/cfe-dev/2017-December/056307.html | 2020-04-22T13:20:26 | 2020-04-22T13:20:26 | {} | MEMBER |
ClickHouse | ClickHouse | 617,782,782 | 10,436 | filimonov | Do you use ALTER UPDATE / ALTER DELETE?
(There is a known memory leak in your version, please update to the latest release. see #9850 #9873) | 2020-04-22T13:34:41 | 2020-04-22T13:42:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 618,097,296 | 10,362 | alexey-milovidov | Why it grows?
Because memory is deallocated in a context where there is no CurrentThread and total_memory_tracker. I planned try to make it used in every context. It should be a simple change but care should be taken... Maybe you can also try to make this change?
Also worth to note that we cannot calculate 100% c... | 2020-04-22T23:51:22 | 2020-04-22T23:51:48 | {} | MEMBER |
ClickHouse | ClickHouse | 618,098,202 | 10,362 | azat | >Maybe you can also try to make this change?
Sure, I will take a look if you will not fix it by that time | 2020-04-22T23:54:57 | 2020-04-22T23:54:57 | {} | MEMBER |
ClickHouse | ClickHouse | 618,155,757 | 8,696 | akuzm | This particular phenomenon is fixed by applying logging configuration changes only on server restart, so I'll close this issue. | 2020-04-23T03:29:48 | 2020-04-23T03:29:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 618,172,031 | 10,357 | YiuRULE | @alexey-milovidov Yep, I could work on it, I cannot work on it right now at the moment, but when I have some free time, I can take a look :) | 2020-04-23T04:35:11 | 2020-04-23T04:35:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 618,218,971 | 4,959 | simonhou | ClickHouse version 19.11.9.52, I met the same error:
```
2020.04.23 14:48:43.994496 [ 48 ] {02a71eb2-5843-41b9-8d0a-8ffcd5de70d1} <Error> executeQuery: Code: 33, e.displayText() = DB::Exception: Cannot read all data. Bytes read: 19854. Bytes expected: 211790. (version 19.11.9.52) (from 10.62.184.177:29845) (in quer... | 2020-04-23T07:01:18 | 2020-04-23T07:01:18 | {} | NONE |
ClickHouse | ClickHouse | 617,480,911 | 9,843 | alexey-milovidov | All done (as of now).
More issues to be found. | 2020-04-22T00:35:26 | 2020-04-22T00:35:26 | {} | MEMBER |
ClickHouse | ClickHouse | 617,481,827 | 10,419 | st3rv04ka | Where are these settings installed? I tried adding them, but it didn't fix this error. | 2020-04-22T00:39:20 | 2020-04-22T00:39:20 | {} | NONE |
ClickHouse | ClickHouse | 617,537,087 | 10,226 | alexey-milovidov | Fun fact: This is very old issue, see CLICKHOUSE-1605.
I reported it on 2015-10-22, so it exists in all ClickHouse versions, even before open-source. | 2020-04-22T04:08:33 | 2020-04-22T04:09:14 | {} | MEMBER |
ClickHouse | ClickHouse | 617,547,850 | 10,176 | alexey-milovidov | @excitoon Why it is draft? We should either merge or close.
| 2020-04-22T04:49:58 | 2020-04-22T04:49:58 | {} | MEMBER |
ClickHouse | ClickHouse | 617,618,380 | 1,966 | alexey-milovidov | @filimonov В таком случае надо просто поместить ссылку в документацию? | 2020-04-22T08:00:44 | 2020-04-22T08:00:44 | {} | MEMBER |
ClickHouse | ClickHouse | 617,937,481 | 10,435 | alexey-milovidov | @abyss7 Maybe you can help? | 2020-04-22T18:00:03 | 2020-04-22T18:00:03 | {} | MEMBER |
ClickHouse | ClickHouse | 618,079,793 | 10,362 | azat | This PR makes ClickHouseMetrics_MemoryTracking grows on and on (after some time it >10TB, and it does not stops growing), while ClickHouseAsyncMetrics_MemoryResident is ok (after some time it is in between 300-600GB), any thoughts? | 2020-04-22T22:50:01 | 2020-04-22T22:50:01 | {} | MEMBER |
ClickHouse | ClickHouse | 618,263,651 | 9,545 | alexey-milovidov | @akuzm Is it possible to avoid initialization/deinitialization all the time? | 2020-04-23T08:36:54 | 2020-04-23T08:36:54 | {} | MEMBER |
ClickHouse | ClickHouse | 617,479,610 | 10,267 | filimonov | Once i tried, and was not able to make it work too. | 2020-04-22T00:30:20 | 2020-04-22T00:30:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,535,319 | 8,550 | alexey-milovidov | @alesapin Do you remember, was it possible that the expression for DELETE was serialized or deserialized in a wrong way, so the datetime is parsed as `((2019 - 12) - 3)`? | 2020-04-22T04:01:11 | 2020-04-22T04:01:11 | {} | MEMBER |
ClickHouse | ClickHouse | 617,640,738 | 10,246 | ficapy | thank you,now it's working~~~ | 2020-04-22T08:43:54 | 2020-04-22T08:43:54 | {} | NONE |
ClickHouse | ClickHouse | 617,658,574 | 10,409 | 4ertus2 | The case when we have most data in hash table slightly improved here #10416. It can be improved further only with another kind of hash table: [Perfect hashing](https://en.wikipedia.org/wiki/Perfect_hash_function) or sparse_hash_map. We could make StorageJoin with perfect hashing and then make JOIN with it.
Another p... | 2020-04-22T09:17:27 | 2020-04-22T09:28:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,686,316 | 9,336 | bohutang | Comments:
workbench 8.0:
```
SHOW SESSION VARIABLES LIKE 'lower_case_table_names'
```
```
2020.04.22 18:05:14.519488 [ 19356 ] {} <Error> executeQuery: Code: 62, e.displayText() = DB::Exception: Syntax error: failed at position 6: SESSION VARIABLES LIKE 'lower_case_table_names'. Expected one of: PROCESSLIST, C... | 2020-04-22T10:13:39 | 2020-04-22T10:13:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,767,325 | 7,665 | 4ertus2 | You are able to test old ANY JOIN with `SETTINGS any_join_distinct_right_table_keys = 1` set.
I've run 3 queries on current master build with your data:
```
clickhouse-client -q 'insert into resolution300 format TSV' < GeneralTests.resolution300.tsv
clickhouse-client -q 'insert into resolution3600 format TSV' < G... | 2020-04-22T13:06:06 | 2020-04-22T13:17:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,920,438 | 9,597 | blinkov | So the current entry point is https://clickhouse.tech/codebrowser/html_report/ClickHouse/index.html
Some parts of it have started showing up in Google search index, but it'll take a while to be relatively complete because there are not so many inbound links and user activity is low.
Here's a query to check curre... | 2020-04-22T17:28:06 | 2020-04-22T17:28:06 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,825,371 | 10,441 | alesapin | Actually it's an unfinished part of https://github.com/ClickHouse/ClickHouse/pull/9227. We already don't add `CAST` in alters, after this, we will not add them in create/attach queries.
It's a bug fix because if the user added column with default with `ALTER` query to the replicated table and executed detach/attach ... | 2020-04-22T14:47:43 | 2020-04-22T17:29:35 | {} | MEMBER |
ClickHouse | ClickHouse | 618,065,986 | 10,418 | alexey-milovidov | @4ertus2 Could you please help to clarify? | 2020-04-22T22:08:04 | 2020-04-22T22:08:04 | {} | MEMBER |
ClickHouse | ClickHouse | 617,476,349 | 1,467 | alexey-milovidov | Steps:
1. Download the data from Kaggle.
2. `unzip kkbox-churn-prediction-challenge.zip`
3.
```
CREATE TABLE user_logs
(
`msno` FixedString(44),
`date_raw` String,
`date` Date DEFAULT toDate(concat(concat(concat(substring(date_raw, 1, 4), '-'), concat(substring(date_raw, 5, 2), '-')), substrin... | 2020-04-22T00:17:56 | 2020-04-22T00:17:56 | {} | MEMBER |
ClickHouse | ClickHouse | 617,530,540 | 9,360 | LiemLin | thinks | 2020-04-22T03:41:03 | 2020-04-22T03:41:03 | {} | NONE |
ClickHouse | ClickHouse | 617,585,620 | 8,438 | qoega | actually there was a fix and now test passes
```
clickhouse-client -n < 00632_get_sample_block_cache.sql
Still alive
65535
``` | 2020-04-22T06:48:36 | 2020-04-22T06:48:36 | {} | MEMBER |
ClickHouse | ClickHouse | 617,638,805 | 10,432 | alexey-milovidov | It's not possible. Totals are represented as a separate part of the resultset and this is the intended behaviour. When you use formats like Pretty, TabSeparated, JSON... you can see that this part of result is semantically separated. | 2020-04-22T08:40:30 | 2020-04-22T08:40:30 | {} | MEMBER |
ClickHouse | ClickHouse | 617,828,138 | 9,514 | nikitamikhaylov | Hello! We are currently rewriting "Cluster copier" to let it be better. Can you try to use copier from our last stable release? Note: You don't have to update your server. | 2020-04-22T14:52:23 | 2020-04-22T14:52:23 | {} | MEMBER |
ClickHouse | ClickHouse | 618,291,918 | 10,398 | 4ertus2 | It's strange that you have JSONRowOutputFormat in stack trace, cause your query has no output format specified. Does the crash appear with default output format? | 2020-04-23T09:29:39 | 2020-04-23T09:29:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,478,865 | 10,278 | filimonov | > I am currently running ClickHouse version 19.5.3
Consider update.
> It is expected that after data is written to this Distributed table, based on sharding key, then data will be written to local MergeTree table or transferred to others MergeTree tables in 2 remaining nodes.
By default inserts to distributed... | 2020-04-22T00:27:34 | 2020-04-22T00:27:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,537,442 | 10,226 | alexey-milovidov | Unfinished PR: #3334 | 2020-04-22T04:10:05 | 2020-04-22T04:10:05 | {} | MEMBER |
ClickHouse | ClickHouse | 617,576,827 | 8,438 | qoega | `
2020.04.22 09:23:49.426193 [ 3200331 ] {c2fda1bf-5f18-4a2f-b318-686fcd9c84e6} <Error> executeQuery: Code: 306, e.displayText() = DB::Exception: Stack size too large. Stack address: 0x70000314a000, frame address: 0x700003186980, stack size: 276096, maximum stack size: 524288 (version 20.4.1.1) (from 127.0.0.1:58956) ... | 2020-04-22T06:24:40 | 2020-04-22T06:25:07 | {} | MEMBER |
ClickHouse | ClickHouse | 617,586,779 | 1,966 | filimonov | Эта статья частично решает проблему недостаточной документации:
https://www.altinity.com/blog/2019/6/11/clickhouse-local-the-power-of-clickhouse-sql-in-a-single-command | 2020-04-22T06:51:19 | 2020-04-22T06:51:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,643,888 | 10,433 | alexey-milovidov | You can wrap your column in `toNullable` function:
```
select toNullable(k) AS grp, 'R' gk, sum(v) s from (
select 0 as k, 5 as v
union all
select 2, 5
union all
select 0, 3
union all
select 4, 9
)
group by grp, gk with totals
```
```
┌─grp─┬─gk─┬─s─┐
│ 0 │ R │ 8 │
│ 2 │ R │ 5 │
... | 2020-04-22T08:49:36 | 2020-04-22T08:49:36 | {} | MEMBER |
ClickHouse | ClickHouse | 617,673,894 | 10,392 | alexey-milovidov | Ok, great. | 2020-04-22T09:47:49 | 2020-04-22T09:47:49 | {} | MEMBER |
ClickHouse | ClickHouse | 618,133,748 | 10,436 | gglinux | > Do you use ALTER UPDATE / ALTER DELETE?
>
> (There is a known memory leak in your version, please update to the latest release. see #9850 #9873)
Yes, frequently used. And my version 20.3.5.21, is the latest release. | 2020-04-23T02:06:18 | 2020-04-23T02:12:44 | {} | NONE |
ClickHouse | ClickHouse | 618,157,296 | 8,696 | akuzm | Ah, also, for posterity -- we've seen similar behavior on some Yandex servers, but for a different reason: the ЫТЬ dictionary library overwrites our signal handler. This is being tracked here: https://nda.ya.ru/t/LgxdmboX3Vx4Cu | 2020-04-23T03:35:49 | 2020-04-23T03:35:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,574,962 | 8,352 | filimonov | Fix #10071 | 2020-04-22T06:19:01 | 2020-04-22T06:19:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,585,053 | 8,438 | alexey-milovidov | Ok. Now we can see that the initial issue (`bus error`) was due to stack overflow and now we have a check for this case and server continue to work.
And the test "sample block cache" is not going to pass on Mac OS, that's normal. | 2020-04-22T06:47:13 | 2020-04-22T06:47:13 | {} | MEMBER |
ClickHouse | ClickHouse | 617,639,425 | 10,432 | alexey-milovidov | But you can emulate the desired behaviour with `group by arrayJoin([k, 0])` | 2020-04-22T08:41:31 | 2020-04-22T08:41:31 | {} | MEMBER |
ClickHouse | ClickHouse | 617,965,667 | 10,425 | alexey-milovidov | Side note: multiple time zones are identical. We can coalesce them at build time.
And the library CCTZ is doing that at runtime (duplicate work).
But as they are small it's ok as is. | 2020-04-22T18:53:25 | 2020-04-22T18:54:06 | {} | MEMBER |
ClickHouse | ClickHouse | 618,048,922 | 10,446 | alexey-milovidov | Tests were not run... we can add `java` to the list of file extensions with code. | 2020-04-22T21:25:22 | 2020-04-22T21:25:22 | {} | MEMBER |
ClickHouse | ClickHouse | 618,156,249 | 10,438 | akuzm | I closed the old issue, because technically it does not reproduce anymore. We can keep this one, sure, though I don't know when we'll get to fixing it properly. | 2020-04-23T03:31:38 | 2020-04-23T03:31:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 618,235,354 | 9,545 | akuzm | The performance tests look bad. It seems that initializing these events is really slow: https://sendeyo.com/up/d/ba99cdd31d
(flame graph from the per fest output, you can download it using the link at the end of the report) | 2020-04-23T07:38:31 | 2020-04-23T07:38:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,946,643 | 10,340 | akuzm | The tests were broken after merge with master and don't look related to this PR. | 2020-04-22T18:17:53 | 2020-04-22T18:17:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,974,027 | 10,000 | ildus | Ok, I have no idea how to remove this regression :) Just adding these conditions like `if (tuple)` makes performance going down. | 2020-04-22T19:09:27 | 2020-04-22T19:09:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 618,018,887 | 10,420 | alexey-milovidov | > I hope you applied all these recommendations automatically?
No, everything was done manually.
It is possible to ask clang-tidy to apply some modifications automatically, but not 100% of them will be correct and it requires careful look into the code. | 2020-04-22T20:22:15 | 2020-04-22T20:22:15 | {} | MEMBER |
ClickHouse | ClickHouse | 617,485,572 | 9,936 | alexey-milovidov | > it could be some bad code there
I hope that it was removed... because we are removing old code ruthlessly. | 2020-04-22T00:55:49 | 2020-04-22T00:55:49 | {} | MEMBER |
ClickHouse | ClickHouse | 617,487,631 | 4,061 | alexey-milovidov | @vitlibar Do you remember that SELECT FROM system.tables recently was forcing unnecessarily loading of dictionaries? | 2020-04-22T01:04:56 | 2020-04-22T01:05:08 | {} | MEMBER |
ClickHouse | ClickHouse | 617,515,871 | 10,385 | akuzm | > Waiting for fix of performance test by @akuzm
No need to wait -- download test results, open set_index-report.tsv, see that the query improved. The layout is exactly the same as for the table with changed performance -- time old, time new, difference in%, quantiles of randomization distribution (5%, 50%, 95%, 99%)... | 2020-04-22T02:46:13 | 2020-04-22T02:46:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,747,900 | 10,285 | filimonov | Similar for meric_log: #8670
Similar for dictionaries: https://github.com/ClickHouse/ClickHouse/issues/9932#issuecomment-605938668
Similar issue for Buffer tables: #10437
Similar (fixed) for Distributed tables: #5142 #4064
Similar (fixed) in LIVE VIEW: #7773
Similar (fixed) in Kafka: #8180 #5243
Similar for mutat... | 2020-04-22T12:26:25 | 2020-04-22T12:29:08 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,793,717 | 10,429 | filimonov | Can i ask you why? What terminal do you use?
```
laptop5591.local :) select * from system.formats where name like '%Pretty%' format PrettyCompactNoEscapes;
SELECT *
FROM system.formats
WHERE name LIKE '%Pretty%'
FORMAT PrettyCompactNoEscapes
┌─name───────────────────┬─is_input─┬─is_output─┐
│ PrettyNoEsc... | 2020-04-22T13:54:26 | 2020-04-22T13:54:26 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,903,945 | 10,429 | alexey-milovidov | > PrettyCompactASCII can be easily added.
Or better to do it with a setting. | 2020-04-22T16:58:27 | 2020-04-22T16:58:27 | {} | MEMBER |
ClickHouse | ClickHouse | 618,169,384 | 10,299 | ltylty | use this sql
```sql
SELECT
warehouse, product, SUM(quantity)
FROM
inventory
GROUP BY warehouse, product with ROLLUP having warehouse != '';
``` | 2020-04-23T04:25:08 | 2020-04-23T04:25:08 | {} | NONE |
ClickHouse | ClickHouse | 617,536,043 | 9,117 | alexey-milovidov | The use case is unclear.
But it may be a nice and easy task to do. | 2020-04-22T04:03:55 | 2020-04-22T04:03:55 | {} | MEMBER |
ClickHouse | ClickHouse | 617,557,010 | 10,405 | hombit | Server log for `SELECT min(mag) FROM default.tmp2` query
```
2020.04.22 05:16:47.341792 [ 56 ] {854d7bbf-3b07-4202-b692-88c126e0eec9} <Debug> executeQuery: (from 127.0.0.1:36182) SELECT min(mag) FROM default.tmp2
2020.04.22 05:16:47.342167 [ 56 ] {854d7bbf-3b07-4202-b692-88c126e0eec9} <Debug> default.tmp2 (SelectExe... | 2020-04-22T05:21:11 | 2020-04-22T05:21:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,565,532 | 10,405 | hombit | Bug reproduces on 20.3.7 revision 54433 | 2020-04-22T05:50:16 | 2020-04-22T05:50:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,918,347 | 10,441 | alesapin | If old version created a table with `CAST` expressions in defaults we will read them from `.sql` and from the zookeeper. For new tables, we will store defaults without casts also in both places. It should work well. | 2020-04-22T17:24:18 | 2020-04-22T17:24:18 | {} | MEMBER |
ClickHouse | ClickHouse | 617,520,275 | 1,467 | alexey-milovidov | Everything is Ok.
The issue was fixed long time ago. | 2020-04-22T03:02:14 | 2020-04-22T03:02:14 | {} | MEMBER |
ClickHouse | ClickHouse | 617,649,026 | 10,433 | pic-Nick | Yes, this is suitable solution. This way I don't need dummy string column at all.
But... This column still has value 'R' in totals. Not default empty string. Is it right? | 2020-04-22T08:59:17 | 2020-04-22T08:59:17 | {} | NONE |
ClickHouse | ClickHouse | 617,663,669 | 10,435 | alexey-milovidov | Something is wrong with link order:
```
: && /usr/bin/ccache /usr/bin/clang++-10 -fdiagnostics-color=always -std=c++2a -fsized-deallocation -gdwarf-aranges -pipe -msse4.1 -msse4.2 -mpopcnt -Wall -Wno-unused-command-line-argument -Wnon-virtual-dtor -fno-omit-frame-pointer -Werror -w -std=c++0x -Wall -pedantic -We... | 2020-04-22T09:27:14 | 2020-04-22T09:27:14 | {} | MEMBER |
ClickHouse | ClickHouse | 618,093,896 | 10,362 | alexey-milovidov | We cannot longer calculate this metric in incremental fashion - should reset it in the same way as total_memory_tracker. | 2020-04-22T23:38:58 | 2020-04-22T23:38:58 | {} | MEMBER |
ClickHouse | ClickHouse | 618,282,148 | 3,196 | sanyfan | Hi @tkroman @blinkov , is there any update about this issue? | 2020-04-23T09:11:26 | 2020-04-23T09:11:26 | {} | NONE |
ClickHouse | ClickHouse | 617,480,084 | 10,385 | alexey-milovidov | Waiting for fix of performance test by @akuzm | 2020-04-22T00:31:57 | 2020-04-22T00:31:57 | {} | MEMBER |
ClickHouse | ClickHouse | 617,985,818 | 10,433 | alexey-milovidov | This is not quite right. It's ok from the implementation point of view. We ensure that main result and totals have the same structure and "structure" means data types and values of constant expressions... | 2020-04-22T19:33:04 | 2020-04-22T19:33:04 | {} | MEMBER |
ClickHouse | ClickHouse | 618,050,672 | 10,447 | mikhno-s | Resolved.
Maybe it can be helpful to somebody, because I did not understand that from docs.
clickhouse has disk `default`, that can be set via `<path>` option.
Then, if you have multiple storage policy - earlier you can just set separately `<path>` and then describe disk in `storage_configuration` - but now yo... | 2020-04-22T21:29:22 | 2020-04-22T21:29:22 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 618,174,162 | 10,436 | gglinux | After 20 hours, the clickhouse crashed. Get this log, it may be useful
```
{} <Error> void DB::StorageBuffer::flushThread(): Code: 252, e.displayText() = DB::Exception: Too many parts (300). Merges are processing significantly slower than inserts., Stack trace (when copying this message, always include the lines belo... | 2020-04-23T04:43:27 | 2020-04-23T06:12:35 | {} | NONE |
ClickHouse | ClickHouse | 617,672,263 | 10,392 | marising | I found the cause of the problem, because the network speed is not very good, contrib /libcxx is not updated completely, I delete the empty directory, and then use git submodule update --init --recursive
| 2020-04-22T09:44:37 | 2020-04-22T09:44:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,678,557 | 10,391 | gglinux | The 20.3.5.21 version fixed. Thx~
But there's another problem in the latest 20.3. #10436 | 2020-04-22T09:57:04 | 2020-04-22T09:57:04 | {} | NONE |
ClickHouse | ClickHouse | 617,978,644 | 9,499 | alexey-milovidov | Continued in #10388 | 2020-04-22T19:18:16 | 2020-04-22T19:18:16 | {} | MEMBER |
ClickHouse | ClickHouse | 618,275,069 | 10,452 | alexey-milovidov | First reccommendation is to update to 19.14.9. | 2020-04-23T08:58:15 | 2020-04-23T08:58:23 | {} | MEMBER |
ClickHouse | ClickHouse | 617,489,790 | 658 | alexey-milovidov | Artem @4ertus2 is currently doing a similar task: transitive constant propagation from WHERE via JOIN ON for pushdown: #9733
| 2020-04-22T01:13:18 | 2020-04-22T01:13:53 | {} | MEMBER |
ClickHouse | ClickHouse | 617,546,214 | 9,623 | alexey-milovidov | @zhang2014 maybe you can help with this issue? | 2020-04-22T04:43:55 | 2020-04-22T04:43:55 | {} | MEMBER |
ClickHouse | ClickHouse | 617,574,635 | 9,270 | filimonov | Fix #10071 | 2020-04-22T06:18:03 | 2020-04-22T06:18:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 617,640,140 | 10,432 | alexey-milovidov | Also you can use `GROUP BY k WITH ROLLUP` (this is probably the best variant) | 2020-04-22T08:42:51 | 2020-04-22T08:42:51 | {} | MEMBER |
ClickHouse | ClickHouse | 617,641,981 | 10,432 | alexey-milovidov | This is exactly what you need:
```
SELECT *
FROM
(
SELECT
k,
sum(v)
FROM
(
SELECT
'1' AS k,
5 AS v
UNION ALL
SELECT
'2',
5
UNION ALL
SELECT
'1',
3
... | 2020-04-22T08:46:10 | 2020-04-22T08:46:27 | {} | MEMBER |
ClickHouse | ClickHouse | 617,717,894 | 10,200 | AlexanderJLiu | @KochetovNicolai
Today, a **no query** `Segment Fault` occur. StackTraces are shown as below:
```shell
2020.04.22 12:10:24.392982 [ 40652 ] {} <Fatal> BaseDaemon: ########################################
2020.04.22 12:10:24.393077 [ 40652 ] {} <Fatal> BaseDaemon: (version 20.3.7.46 (official build)) (from thread... | 2020-04-22T11:24:13 | 2020-04-22T11:35:08 | {} | NONE |
ClickHouse | ClickHouse | 617,786,759 | 9,873 | filimonov | Working properly now. (latest 20.3, 20.1, 20.4) | 2020-04-22T13:42:09 | 2020-04-22T13:42:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 618,013,522 | 10,188 | nikitamikhaylov | `test_mysql_protocol/test.py::test_java_client` container failed to start - not related | 2020-04-22T20:11:35 | 2020-04-22T20:11:35 | {} | MEMBER |
ClickHouse | ClickHouse | 618,239,967 | 10,326 | BernieJiangDL | > Yes, it does support but:
> MV is an insert trigger. It watches inserts into LEFT FROM TABLE only.
> So you can create MV `as select ... from SOMELEFT join SOMERIGHT`, but MV will be triggered by inserts into SOMELEFT.
> Plus another problem that JOINs are slow and MV with JOIN with large SOMERIGHT will slow down ... | 2020-04-23T07:48:15 | 2020-04-23T07:48:15 | {} | NONE |
ClickHouse | ClickHouse | 617,480,509 | 10,383 | alexey-milovidov | Need to adjust description, because it's a bugfix and it should be in changelog. | 2020-04-22T00:33:37 | 2020-04-22T00:33:37 | {} | MEMBER |
ClickHouse | ClickHouse | 617,536,534 | 9,099 | alexey-milovidov | We decided to extract tables in AST level. It will work for tables that are mentioned in SQL query, not for tables that are actually used. | 2020-04-22T04:05:55 | 2020-04-22T04:05:55 | {} | MEMBER |
ClickHouse | ClickHouse | 617,545,175 | 10,414 | alexey-milovidov | Your changelog entry is not suitable for the changelog.
I'll fix it...
(done). | 2020-04-22T04:40:07 | 2020-04-22T04:41:05 | {} | MEMBER |
ClickHouse | ClickHouse | 617,741,465 | 7,665 | Axelfoley85 | SEMI LEFT JOIN does not just replace the old behaviour of ANY INNER JOIN. Assuming you have two tables with different time resolutions and want to join them.
STEPS TO REPRODUCE
Ran this on version 19.15.3.6
1. create tables
```sql
'DROP TABLE IF EXISTS GeneralTests.resolution300',
`CREATE TABLE Gen... | 2020-04-22T12:13:04 | 2020-04-22T12:23:18 | {} | NONE |
ClickHouse | ClickHouse | 617,854,895 | 7,665 | 4ertus2 | > Is that correct?
No. `SEMI LEFT JOIN` sould be the same as `ANY INNER JOIN` with `any_join_distinct_right_table_keys=1` and `ANY INNER JOIN` in old versions where we have no `any_join_distinct_right_table_keys` setting. But there could be bugs. I.e. bugs with column names: you have columns with same names in both ... | 2020-04-22T15:36:38 | 2020-04-22T15:47:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 618,225,604 | 4,959 | simonhou | @alexey-milovidov Can you help to verify this exception? | 2020-04-23T07:17:04 | 2020-04-23T07:17:04 | {} | NONE |
ClickHouse | ClickHouse | 625,191,933 | 10,725 | sundy-li | @wenwen2019 please use English to submit the issue. | 2020-05-07T11:16:52 | 2020-05-07T11:16:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 625,230,135 | 10,686 | CurtizJ | Failed tests are not related. | 2020-05-07T12:39:31 | 2020-05-07T12:39:31 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.