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 | 494,381,893 | 4,002 | alex-krash | We've faced performance issues with official driver, when working with Date/DateTime fields - it uses parsing of incoming string into object. Issue was solved, when support for RowBinary read was implemented - https://github.com/yandex/clickhouse-jdbc/blob/a86eb8a65c25788f767e51a619f62073dbbfeddb/src/main/java/ru/yande... | 2019-05-21T13:03:29 | 2019-05-21T13:03:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,448,390 | 5,369 | azat | cc: @danlark1 | 2019-05-21T15:47:59 | 2019-05-21T15:47:59 | {} | MEMBER |
ClickHouse | ClickHouse | 494,473,218 | 5,369 | danlark1 | https://github.com/yandex/ClickHouse/commit/cfdba938ce22f16efeec504f7f90206a515b1280 | 2019-05-21T16:56:38 | 2019-05-21T16:56:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,787,044 | 5,362 | 4ertus2 | Cannot reproduce the issue. Can you modify this sample into crashed variant?
```
CREATE TABLE t
(
shift_code UInt32,
shift_date UInt32,
cur_date UInt32,
begin_times UInt32,
end_times UInt32
) Engine = Memory;
INSERT INTO t SELECT number as shift_code, number as shift_date, number as cu... | 2019-05-22T12:46:47 | 2019-05-22T12:46:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 423,435,071 | 3,174 | filimonov | 1) Each insert create a folder in `/var/lib/clickhouse/.../table_name/`. Inside that folder there are 2 files per each column - one with data (compressed), second with index. Data is physically sorted by primary key inside those files. Those folders are called '**parts**'.
2) ClickHouse merges those smaller parts to ... | 2018-09-21T07:04:18 | 2019-05-24T07:16:35 | {
"+1": 73,
"hooray": 7,
"rocket": 10
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,678,633 | 5,409 | nvartolomei | Postgres? | 2019-05-24T15:30:55 | 2019-05-24T15:30:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,685,531 | 5,409 | den-crane | @WeiGangqiang https://github.com/yandex/ClickHouse/issues/5125#issuecomment-487599751 | 2019-05-24T15:50:36 | 2019-05-24T15:50:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,916,562 | 5,372 | alexey-milovidov | Now let's explicitly turn off the setting in those tests that depend on order or error counters. | 2019-05-25T12:58:30 | 2019-05-25T12:58:30 | {} | MEMBER |
ClickHouse | ClickHouse | 494,636,936 | 5,371 | zhang2014 | I have discussed such issues with @sundy-li before, and I'm not sure if this will help this issues. I will describe the process of discussion below : )
- First we can support some virtual columns for the HDFS Engine, such as :
``` SQL
CREATE TABLE hdfs(...)ENGINE = HDFS('hdfs://hdfs1:9000/path/dt=YYYYMMDD', 'TSV'... | 2019-05-22T03:20:01 | 2019-05-22T03:24:49 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,650,755 | 5,362 | filimonov | Reopening. (ClickHouse should not crash even if query is wrong) | 2019-05-22T04:48:33 | 2019-05-22T04:48:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,670,157 | 5,375 | bill1234567890abc | I checked clickhouse jdbc source. It seems an issue in PreparedStatementParser.parseSQL() | 2019-05-22T06:28:50 | 2019-05-22T06:28:50 | {} | NONE |
ClickHouse | ClickHouse | 494,695,338 | 5,153 | ghost | @blinkov
Sorry, I'm too late.
I have some other issues in my company.
I just tested the insertion speed of ClickHouse.
ClickHouse is fast enough for our purpose and I will push more heavier load to ClickHouse for the production level.
If I have another question, put a new issue.
Sincerely, thank you very much. | 2019-05-22T07:56:05 | 2019-05-22T07:56:29 | {} | NONE |
ClickHouse | ClickHouse | 495,094,451 | 4,918 | filimonov | Won't it lead to excessive part 'movements' between same disk? Won't it break anything if partX will be 'moved' from diskA to diskA? | 2019-05-23T07:05:51 | 2019-05-23T07:08:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,207,091 | 5,386 | nshah14285 | Thanks for the update.
Version 19.8 release date? | 2019-05-23T12:57:07 | 2019-05-23T12:57:07 | {} | NONE |
ClickHouse | ClickHouse | 495,335,256 | 5,349 | akuzm | This is going to take some time. Currently, there is no clear stage that parses a CSV row to an array of plain text column values. I'll need to factor out this step, so that we can make decisions based on the plain text values, e.g., substitute default values for empty plain text.
For the record, now CSV is parsed ... | 2019-05-23T18:36:23 | 2019-05-23T18:36:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,680,751 | 5,408 | alesapin | Merge with squash pls. | 2019-05-24T15:37:09 | 2019-05-24T15:37:09 | {} | MEMBER |
ClickHouse | ClickHouse | 495,913,549 | 5,390 | alexey-milovidov | Thank you! | 2019-05-25T12:11:35 | 2019-05-25T12:11:35 | {} | MEMBER |
ClickHouse | ClickHouse | 494,416,446 | 1,148 | fsanaulla | > FixedStrings as dictionary keys are still not supported.
>
> But you can load FixedString content as a String.
> Simply specify `String` as a type in dictionary structure and it will load FixedStrings from the target table.
Hi @alexey-milovidov , can you also in more details explain how conversion work/should ... | 2019-05-21T14:30:36 | 2019-05-21T14:30:52 | {} | NONE |
ClickHouse | ClickHouse | 494,751,440 | 5,371 | alexey-milovidov | This is much more simple task than you expect.
Consider if we only have File storage and want to read multiple files.
In this task we will add support for wildcards.
If the path contains wildcards, the table will be readonly.
We can add virtual columns `_path` and `_filename` to filter files in query.
Everythi... | 2019-05-22T10:46:29 | 2019-05-22T10:49:30 | {} | MEMBER |
ClickHouse | ClickHouse | 495,650,014 | 5,407 | kvap | cc @bocharov @twalwyn @Dorokhov @victor-perov | 2019-05-24T14:19:14 | 2019-05-24T14:19:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,792,360 | 5,372 | alexey-milovidov | > you're making changes in your master
It's totally fine, IMO. | 2019-05-24T21:25:36 | 2019-05-24T21:25:36 | {} | MEMBER |
ClickHouse | ClickHouse | 495,806,495 | 5,372 | overshov | I have checked stateless tests logs from sandbox. Most diffs caused by rearrangement of output lines, because of parallel reading. Except 00418_input_format_allow_errors, there is some logical error.
Yes, each ReadBuffer has unique errors counter and parameter ```input_format_allow_errors_num``` does not work. | 2019-05-24T22:41:25 | 2019-05-24T22:58:28 | {} | NONE |
ClickHouse | ClickHouse | 495,889,750 | 5,403 | hczhcz | The parameters are actually "column x and y." Arrays are transformed in `arrayReduce` function. | 2019-05-25T08:54:58 | 2019-05-25T08:55:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,913,179 | 5,394 | alexey-milovidov | The only remaining is the integration test. | 2019-05-25T12:06:12 | 2019-05-25T12:06:12 | {} | MEMBER |
ClickHouse | ClickHouse | 494,456,873 | 5,369 | danlark1 | That seems strange a bit so that number of mmaps were executed. I will look at it | 2019-05-21T16:10:48 | 2019-05-21T16:10:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,072,314 | 5,381 | abraithwaite | @nvartolomei was right, poco handles it correctly.
I believe this is what's happening now:
https://github.com/yandex/ClickHouse/issues/5287#issuecomment-495071521
@Enmk WDYT about my current theory ^ | 2019-05-23T05:25:57 | 2019-05-23T05:26:12 | {} | NONE |
ClickHouse | ClickHouse | 495,071,521 | 5,287 | abraithwaite | Alright, I believe I've found it.
Here we're creating a normal session:
https://github.com/yandex/ClickHouse/blob/98359a6a0978dfa3d7e171ad8c30db9e814a8b9d/dbms/src/Storages/MergeTree/DataPartsExchange.cpp#L189-L197
PooledReadWriteBufferFromHTTP Initializer calls makePooledHTTPSession
https://github.com/yandex/C... | 2019-05-23T05:22:10 | 2019-05-23T05:44:57 | {} | NONE |
ClickHouse | ClickHouse | 495,104,437 | 5,381 | Enmk | > @nvartolomei was right, poco handles it correctly.
>
> I believe this is what's happening now:
>
> [#5287 (comment)](https://github.com/yandex/ClickHouse/issues/5287#issuecomment-495071521)
>
> @Enmk WDYT about my current theory ^
Sounds legit, I see two options now:
* Key by IP. I believe that would cau... | 2019-05-23T07:40:18 | 2019-05-23T08:53:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,900,034 | 5,403 | BayoNet | Thanx. I've changed text a bit. | 2019-05-25T09:57:18 | 2019-05-25T09:57:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,960,074 | 5,381 | nvartolomei | irc send resets connection under the hood if last request errored | 2019-05-22T20:30:39 | 2019-05-22T20:30:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,201,209 | 5,386 | alexey-milovidov | No. The minimal version is 19.8. | 2019-05-23T12:39:03 | 2019-05-23T12:39:03 | {} | MEMBER |
ClickHouse | ClickHouse | 495,340,549 | 5,349 | alexey-milovidov | > This is going to take some time. Currently, there is no clear stage that parses a CSV row to an array of plain text column values. I'll need to factor out this step, so that we can make decisions based on the plain text values, e.g., substitute default values for empty plain text.
You cannot (without sacrifice the... | 2019-05-23T18:51:02 | 2019-05-23T18:53:52 | {} | MEMBER |
ClickHouse | ClickHouse | 495,611,356 | 3,174 | WeiGangqiang | > Hi. And what can I do, if inserts are rare and small? I want near realtime processing, So I can't form greater batches. I have up to 1k rows in a batch (3kb-2mb on disc), one insert in 30 seconds. And it seems ClickHouse doesn't merge parts, collect 300 on this table, but it hasn't reached some minimal merge size (ev... | 2019-05-24T12:47:53 | 2019-05-24T12:47:53 | {} | NONE |
ClickHouse | ClickHouse | 495,691,772 | 3,174 | WeiGangqiang | > isDone
how to check mutations is Done? | 2019-05-24T16:09:20 | 2019-05-24T16:18:15 | {} | NONE |
ClickHouse | ClickHouse | 495,915,431 | 2,389 | loyd | The same with materialized columns. | 2019-05-25T12:39:30 | 2019-05-25T12:39:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,916,488 | 5,372 | alexey-milovidov | > I have checked stateless tests logs from sandbox. Most diffs caused by rearrangement of output lines, because of parallel reading.
This is the expected result (it is fine).
BTW, we have discussed an option for order-preserving parallel parsing of formats. It is a little bit more difficult.
> Yes, each ReadBu... | 2019-05-25T12:57:21 | 2019-05-25T12:57:21 | {} | MEMBER |
ClickHouse | ClickHouse | 494,465,176 | 5,369 | danlark1 | Need to think of a better fix, for now, I will disable LFALLOC.
At Yandex we have max_map_count set to 2**20 so no service has that problem. | 2019-05-21T16:33:14 | 2019-05-21T16:43:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,632,030 | 5,362 | yxjsnj2018 | The problem was found.
Wrong SQL:
====================================================
select
....
) AS u
ANY INNER JOIN (SELECT * FROM v_cur_pre_shift ) AS v
USING (shift_code, shift_date)
WHERE (u.cur_date >= v.begin_times) AND (u.cur_date < end_times)
==============================================... | 2019-05-22T02:52:09 | 2019-05-22T02:52:30 | {} | NONE |
ClickHouse | ClickHouse | 494,732,399 | 5,375 | filimonov | Can you make a pull request with fix to https://github.com/yandex/ClickHouse-jdbc project?
Maybe it is the same: https://github.com/yandex/clickhouse-jdbc/pull/322 ?
| 2019-05-22T09:45:35 | 2019-05-22T09:47:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,842,853 | 5,380 | clemrodriguez | Already fixed since 2 days. | 2019-05-22T15:04:33 | 2019-05-22T15:04:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,091,861 | 2,810 | PhilHua | I have upgraded CH to 18.6.0 but this error still occur. | 2019-05-23T06:56:02 | 2019-05-23T06:56:02 | {} | NONE |
ClickHouse | ClickHouse | 495,209,728 | 5,381 | abraithwaite | > I believe that we still need to resolve host to IP to be consistent with rest of the codebase.
Respectfully, I think that's the wrong choice here. This is a pooled connection, meaning we're already doing our best to reuse resources and not do unnecessary DNS lookups.
This is the only place `setHost` is called ... | 2019-05-23T13:05:03 | 2019-05-23T13:05:03 | {} | NONE |
ClickHouse | ClickHouse | 495,450,464 | 5,359 | andyyzh | done. | 2019-05-24T02:30:41 | 2019-05-24T02:31:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,761,482 | 5,298 | manish53 | I've modified it myself. Able to connect atlas with clickhouse. Added updated poco project in contrib path and added securesocketstream implementation in connection.cpp to support ssl enable connection.
Also added readPreference functionality. To know more please follow pull request section. Patch files are provided ... | 2019-05-24T19:22:21 | 2019-05-24T19:22:21 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 495,909,272 | 5,395 | alexey-milovidov | LGTM | 2019-05-25T11:13:12 | 2019-05-25T11:13:12 | {} | MEMBER |
ClickHouse | ClickHouse | 494,338,578 | 5,362 | 4ertus2 | It looks like we in ANY INNER JOIN. Do you have the query that leads to the crash? Do you use both the right and the left join key columns in the query? Do you have duplicated join keys in JOIN ON section (from right table)?
P.S. Most probable we're in AddedColumns::appendFromBlock and right_indexes vector is incons... | 2019-05-21T10:41:15 | 2019-05-21T14:22:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,633,458 | 5,371 | sundy-li | - Single shard to aggregate large data may be the bottleneck, will it have a new Engine like `DistributeHDFS` storage? It will distribute the files to each shard by some scheduling algorithms.
- Index pushed to filter the files, usually we split the HDFS file by date, `hdfs://path/table_name/dt=YYYYMMDD`. So the qu... | 2019-05-22T03:00:04 | 2019-05-22T03:00:37 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,676,805 | 4,631 | vezir | I updated CH from 19.5.3.8 to 19.6.2.11 and at the same time i cleaned some kafka related table manualy because drop didnt work. May be different case but then I got same error with these config settings;
` <kafka>
<max_poll_interval_ms>60000</max_poll_interval_ms>
<session_timeout_ms>60000</sessi... | 2019-05-22T06:54:37 | 2019-05-22T07:11:07 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 495,287,368 | 5,381 | abraithwaite | Actually, the last comment that line before `SYSTEM DROP DNS CACHE` had was to cache DNS until restart:
https://github.com/yandex/ClickHouse/commit/48f5d8f1a3e0f980e1d559d0a5cd4b9ebbe29220 | 2019-05-23T16:19:21 | 2019-05-23T16:19:21 | {} | NONE |
ClickHouse | ClickHouse | 495,893,919 | 5,402 | BayoNet | Along with other changes, I removed `LEFT` from the syntax string. Is it right? Should we list all the supported join types or the `ASOF JOIN` is supported for all `JOIN` types? | 2019-05-25T09:21:12 | 2019-05-25T09:21:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,378,647 | 5,365 | Chr1sxu | I tried to modify like this, but it looks windowFunnel doesn't support to group by one dimension only based on the first step.
```sql
SELECT
exp_version,
sumIf(count, level >= 1) AS sum1,
sumIf(count, level >= 2) AS sum2
FROM
(
SELECT
level,
exp_version,
count()... | 2019-05-21T12:54:23 | 2019-05-21T14:29:40 | {} | NONE |
ClickHouse | ClickHouse | 494,758,954 | 4,631 | vezir | I removed the kafka parameters from the config.xml and redefine the create table statement with kafka engine with these parameters;
`kafka_num_consumers = 4,
kafka_row_delimiter = '\n',
kafka_skip_broken_messages = 1,
kafka_max_block_size = 1000;`
Now, It seems ok. I am monitoring. | 2019-05-22T11:13:09 | 2019-05-22T11:13:09 | {} | NONE |
ClickHouse | ClickHouse | 494,974,420 | 5,381 | abraithwaite | Digging into the Poco codebase, that's some pretty complex code. I've got strong reason to believe that in our case the connections are not being reinitialized though for whatever reason.
One thing I find interesting however is that `connected()` calls `socket.initialized()` which only checks `_sockfd != POCO_INVAL... | 2019-05-22T21:10:21 | 2019-05-22T21:11:20 | {} | NONE |
ClickHouse | ClickHouse | 495,066,358 | 5,381 | Enmk | > I'm an idiot. `close()` is called in the except blocks for both `sendRequest()` and `receiveResponse()` in Poco libraries. Dead end...
Don't blame yourself, it is complicated. In fact, I think that your fix might work, regardless of catching an exception in `HTTPClientSession::sendRequest()`.
(please help me d... | 2019-05-23T04:51:50 | 2019-05-23T04:52:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,111,076 | 5,287 | alex-zaitsev | We were able to reproduce it with Altinity Operator, @Enmk will help with the fix. | 2019-05-23T08:01:18 | 2019-05-23T08:01:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,234,030 | 5,343 | 4ertus2 | #5394 | 2019-05-23T14:07:32 | 2019-05-23T14:07:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,902,445 | 5,384 | vitlibar | @manish53 You've created two more similar PRs:
https://github.com/yandex/ClickHouse/pull/5387
https://github.com/yandex/ClickHouse/pull/5388
Are they all necessary? Can you describe what's the purpose of them? | 2019-05-25T10:13:32 | 2019-05-25T10:23:02 | {} | MEMBER |
ClickHouse | ClickHouse | 495,906,550 | 5,384 | vitlibar | Can you please provide more detailed description what you're implementing here? Currently it doesn't look very clear. Maybe some links would help. What's Mongo URI? | 2019-05-25T10:41:16 | 2019-05-25T10:41:16 | {} | MEMBER |
ClickHouse | ClickHouse | 494,637,135 | 5,373 | 16601746291 | my tools problem , solved! | 2019-05-22T03:21:05 | 2019-05-22T03:21:05 | {} | NONE |
ClickHouse | ClickHouse | 495,201,928 | 2,810 | alexey-milovidov | Please upgrade to 19.6. | 2019-05-23T12:41:31 | 2019-05-23T12:41:31 | {} | MEMBER |
ClickHouse | ClickHouse | 495,684,441 | 5,407 | victor-perov | Looks good! Thanks, @kvap! | 2019-05-24T15:47:38 | 2019-05-24T15:47:38 | {} | NONE |
ClickHouse | ClickHouse | 494,475,107 | 5,369 | azat | > At Yandex we have max_map_count set to 2**20 so no service has that problem.
AFAIK it is not great to have tons of vma, and if we are talking about this particular case this should be a problem even with max_map_count=1<<20, since this is not about what limit do you have it is about "vma for each 72K of memory" | 2019-05-21T17:02:05 | 2019-05-21T17:02:05 | {} | MEMBER |
ClickHouse | ClickHouse | 494,961,427 | 5,381 | nvartolomei | https://github.com/yandex/ClickHouse/pull/4837#issuecomment-477930639 | 2019-05-22T20:34:12 | 2019-05-22T20:34:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,130,389 | 5,254 | Dinamitarh | I upgrade cluster to 19.6.2.1 end my problem is fixed. | 2019-05-23T08:58:53 | 2019-05-23T08:58:53 | {} | NONE |
ClickHouse | ClickHouse | 495,163,891 | 5,352 | sundy-li | Python 2 is outdated, it will be great if upgraded to python 3. | 2019-05-23T10:32:05 | 2019-05-23T10:32:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,222,676 | 5,208 | abyss7 | > @abyss7 why not possible? You can commit (best effort), before shutting down the tables.
It's not possible in this specific version, because it's too old and we won't update it. For 19.8+ there is already an option that allows to commit either on every message batch or on every merged block: #5100 | 2019-05-23T13:40:05 | 2019-05-23T13:40:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,226,775 | 5,381 | Enmk | > This is a pooled connection, meaning we're already doing our best to reuse resources and not do unnecessary DNS lookups.
well, maybe... I do not know for sure what was the intention behind this. @alesapin could you please clarify ? | 2019-05-23T13:50:03 | 2019-05-23T13:50:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,475,637 | 5,400 | filimonov | Most probably oomkiller killed the main clickhouse process. Please examine `dmesg` | 2019-05-24T05:17:56 | 2019-05-24T05:18:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,486,971 | 5,208 | abyss7 | Your version is too old - I think there is no mechanism to achieve desired behavior, beside using very small `kafka_max_block_size` in the table. | 2019-05-21T17:36:12 | 2019-05-21T17:36:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,512,005 | 4,918 | ObjatieGroba | Make Schema const.
For current usage it is OK | 2019-05-21T18:47:35 | 2019-05-21T18:47:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,124,149 | 5,254 | Dinamitarh | I have the same errors with joins | 2019-05-23T08:42:42 | 2019-05-23T08:42:42 | {} | NONE |
ClickHouse | ClickHouse | 495,302,039 | 5,010 | simPod | for anyone interested
```java
public class ProtoSerializer<T> implements Serializer<T> {
@Override
public void configure(Map<String, ?> map, boolean b) {
}
@Override
public byte[] serialize(String topic, T t) {
Message message = (Message) t;
ByteArrayOutputStream outpu... | 2019-05-23T17:01:41 | 2019-05-23T17:01:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,477,403 | 5,386 | filimonov | Before 19.6 you can use old-fashioned functions to work with JSON (those having visitParam prefix). | 2019-05-24T05:28:50 | 2019-05-24T05:28:50 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,679,655 | 5,409 | WeiGangqiang | > Postgres?
clickhouse, our service's main scene is OLAP, but need support update 50 rows one second | 2019-05-24T15:33:53 | 2019-05-24T15:33:53 | {} | NONE |
ClickHouse | ClickHouse | 495,907,353 | 5,408 | alexey-milovidov | Ok. Now it's time to totally remove `metric` and `main_metric` elements. | 2019-05-25T10:46:53 | 2019-05-25T10:46:53 | {} | MEMBER |
ClickHouse | ClickHouse | 494,554,620 | 4,178 | s-mx | > Inability to create dictionary with default expression without default: >https://pastebin.com/raw/gM3Gm0V4
Hi, @alesapin. You mentioned a problem with default expression. As I understand by [docs](https://clickhouse.yandex/docs/en/query_language/dicts/external_dicts_dict_structure/#attributes), null_value option i... | 2019-05-21T20:54:38 | 2019-05-21T20:54:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,282,422 | 5,072 | 4ertus2 | #5396 | 2019-05-23T16:05:42 | 2019-05-23T16:05:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,476,138 | 2,810 | filimonov | You have 1**8**.6. Please update to latest stable release (currenly 1**9**.6. 2.11) | 2019-05-24T05:21:13 | 2019-05-24T05:23:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,379,807 | 5,363 | alex-krash | I will make a PR in 2 days. | 2019-05-21T12:57:43 | 2019-05-21T12:57:43 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,498,355 | 5,208 | nvartolomei | @abyss7 why not possible? You can commit (best effort), before shutting down the tables. | 2019-05-21T18:08:26 | 2019-05-21T18:08:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,939,392 | 5,100 | abyss7 | > Also i wan't to clarify, what should happen with block of data when server restarts(`systemctl restart`). This block must be written or it would be lost?
For now there is no any stream cancellation when we stop CH. So the server will wait until all streams finish their work in either graceful way: max_block_size, ... | 2019-05-20T11:01:27 | 2019-05-23T13:28:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,621,201 | 235 | pramitchoudhary | Recently tried building the latest on Mac (_10.13.6_) but got the following error
steps:
```
1. cmake .. -DCMAKE_CXX_COMPILER=g++-8 -DCMAKE_C_COMPILER=gcc-8
2. ninja
also tried,
cmake .. -DCMAKE_CXX_COMPILER=g++-8 -DCMAKE_C_COMPILER=gcc-8 -DENABLE_EMBEDDED_COMPILER=0 -DUSE_INTERNAL_LLVM_LIBRARY=0
```
`
[28... | 2019-05-22T01:51:20 | 2019-05-23T00:10:22 | {} | NONE |
ClickHouse | ClickHouse | 495,084,752 | 4,918 | ObjatieGroba | > Bug? Isn't it just a bad config?
No, it is bug. We have to check it when we load schema from config file.
Valid schema must contain at least one volume
Valid volume must contain at least one disk
If not, it’s bug | 2019-05-23T06:26:43 | 2019-05-23T06:42:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,285,569 | 5,381 | abraithwaite | Looking at the git history, it seems like it was an attempt to use the internal caching resolver over just resolving the hostname. Following it backwards in the git history, it seems quite old.
Maybe we should move the conversation back to #5287 just to keep everything in one place.
I really appreciate your help... | 2019-05-23T16:14:03 | 2019-05-23T16:14:03 | {} | NONE |
ClickHouse | ClickHouse | 495,299,223 | 5,208 | nvartolomei | Got it.
> On 23 May 2019, at 14:40, Ivan <notifications@github.com> wrote:
>
> @abyss7 why not possible? You can commit (best effort), before shutting down the tables.
>
> It's not possible in this specific version, because it's too old and we won't update it. For 19.8+ there is already an option that allows to com... | 2019-05-23T16:53:24 | 2019-05-23T16:53:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,781,802 | 5,372 | ivan-v-kush | you're making changes in your `master`
More convenient, if you'd create feature branch and work in it and merge it.
http://nvie.com/posts/a-successful-git-branching-model/ | 2019-05-24T20:39:41 | 2019-05-24T20:39:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,685,788 | 4,631 | vezir | I changed my kafka related settings with;
` <kafka>
<fetch_max_bytes>104857600</fetch_max_bytes>
<fetch_message_max_bytes>104857600</fetch_message_max_bytes>
<receive_message_max_bytes>105000000</receive_message_max_bytes>
<max_poll_interval_ms>300000</max_poll_interval_ms>
... | 2019-05-22T07:25:23 | 2019-05-22T07:25:23 | {} | NONE |
ClickHouse | ClickHouse | 494,866,055 | 5,362 | filimonov | On altinity build synthetic sample is also working properly. | 2019-05-22T16:00:33 | 2019-05-22T16:00:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,966,673 | 5,381 | abraithwaite | Thanks for the pointers! Reading that code now, but I'm confused as to why the Poco docs still say that resetting the connection is required when issues are encountered.
Basically trying to hunt down why we'd be connecting to a stale IP address after the host changes what it's pointing to (#5287) | 2019-05-22T20:48:03 | 2019-05-22T20:48:03 | {} | NONE |
ClickHouse | ClickHouse | 495,088,444 | 4,918 | ObjatieGroba | > BTW maybe we should check is each disk appears in schema only once. I.e. not listed twice in multiple volumes.
Bug or feature?
Example:
hdd1, hdd2
Volume 1: hdd1 with part < 1Gb
Volume 2: hdd2, hdd1
Why? Because we can.
When you have hdd1 rw speed > hdd2 rw speed
Hdd1 is 2Tb
Hdd2 is 10Gb
Now you have ... | 2019-05-23T06:42:16 | 2019-05-23T06:51:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 495,113,859 | 5,384 | alesapin | You can ignore test fails until you resolve conflicts. | 2019-05-23T08:10:35 | 2019-05-23T08:10:35 | {} | MEMBER |
ClickHouse | ClickHouse | 495,530,279 | 5,396 | alesapin | perftest flap
| 2019-05-24T08:56:36 | 2019-05-24T08:56:36 | {} | MEMBER |
ClickHouse | ClickHouse | 495,567,665 | 5,014 | 243f6a8885a308d313198a2e037 | ## I tried the fix merged above and found that the problem is NOT completely fixed.
### Environment
- OS: Ubuntu 16.04
- Database structure: 1 shard, 2 replica (to use ZooKeeper)
- ClickHouse: ea739d193229df3f7230a8f5cfd90444bfca6de0 with following modifications:
```
ClickHouse$ git diff
diff --git a/contr... | 2019-05-24T10:39:46 | 2019-05-24T10:39:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,683,212 | 5,371 | alex-krash | **Syntax-related thoughts**
In terms of data manipulation, sub-directories are partitions.
From my point of view, such tables should have DDL, similar to one created in Hive:
```
CREATE TABLE hdfs_table
(
id UInt32,
dt Date,
country_id UInt32
) ENGINE = HDFS('hdfs://hdfs1:9000/path/')
PARTITION BY ... | 2019-05-22T07:16:31 | 2019-05-22T07:16:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 494,998,607 | 5,381 | abraithwaite | I'm an idiot. `close()` is called in the except blocks for both `sendRequest()` and `receiveResponse()` in Poco libraries. Dead end... | 2019-05-22T22:32:21 | 2019-05-22T22:32:36 | {} | NONE |
ClickHouse | ClickHouse | 495,288,083 | 5,287 | abraithwaite | Where the line was last changed was when `system drop dns cache` was added:
https://github.com/yandex/ClickHouse/commit/48f5d8f1a3e0f980e1d559d0a5cd4b9ebbe29220 | 2019-05-23T16:21:22 | 2019-05-23T16:21:22 | {} | NONE |
ClickHouse | ClickHouse | 495,291,544 | 2,810 | perez1987 | @alexey-milovidov which release exactly? I have the same problem with the 18.6.0. | 2019-05-23T16:31:06 | 2019-05-23T16:31:06 | {} | NONE |
ClickHouse | ClickHouse | 495,686,748 | 5,254 | koxt | @Dinamitarh I upgraded server version to 19.6.2.1, but have the same failed tests. | 2019-05-24T15:53:56 | 2019-05-24T15:53:56 | {} | NONE |
ClickHouse | ClickHouse | 495,701,868 | 5,287 | abraithwaite | @alexey-milovidov @alex-zaitsev @Enmk , is there anything I can do to help move this along? Just say the words and I'll get on it. :-) | 2019-05-24T16:36:29 | 2019-05-24T16:36:29 | {} | NONE |
ClickHouse | ClickHouse | 495,767,124 | 5,066 | den-crane | fixed in 19.6.2.11-stable | 2019-05-24T19:43:38 | 2019-05-24T19:43:38 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.