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 | 289,815,978 | 638 | robot-metrika-test | Can one of the admins verify this patch? | 2017-03-28T15:54:03 | 2017-03-28T15:54:03 | {} | NONE |
ClickHouse | ClickHouse | 290,343,951 | 642 | sirkon | You need Date (not Data, it is Date, i.e. day, '2017-03-30' for instance), because MergeTree engine splits data into partitions by months, and the engine obviously needs a way to know what date a record belongs to and it only was taught to use Date typed field for the purpose.
So, you need a Date typed field in table... | 2017-03-30T08:46:08 | 2017-03-30T08:46:08 | {} | NONE |
ClickHouse | ClickHouse | 290,461,393 | 196 | ztlpn | Solved in v1.1.54198. | 2017-03-30T16:13:27 | 2017-03-30T16:13:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 291,479,425 | 642 | ludv1x | FYI You can create dummy date filed (which you will not use into INSERT and SELECT queries).
In that case all table's data belong to the same partition (with name `197001`), but such commands as `FREEZE PARTITION`, `DROP PARTITION`, etc does not make sense. | 2017-04-04T12:04:24 | 2017-04-04T12:04:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,303,485 | 700 | ztlpn | Sure:
```
:) create table t1(f1 Nullable(String),f2 Nullable(Float64),f3 Nullable(Int64),f4 Nullable(DateTime),f5 Nullable(Float64)) engine = TinyLog;
CREATE TABLE t1
(
f1 Nullable(String),
f2 Nullable(Float64),
f3 Nullable(Int64),
f4 Nullable(DateTime),
f5 Nullable(Float64)
) ENGINE = T... | 2017-04-15T16:27:06 | 2017-04-15T16:27:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 295,458,850 | 703 | alexey-milovidov | Confirmed, will fix. | 2017-04-19T21:48:45 | 2017-04-19T21:48:45 | {
"+1": 3
} | MEMBER |
ClickHouse | ClickHouse | 290,478,511 | 615 | George3d6 | Sorry, I had a bit of a brainfart when writing the original, I'll add a similar `show_precentage_char` bool instead of removing `progress.rows * 2 < progress.total_rows` in order for the behavior to remain the same as before my commit + % still being outputed when there is no physical terminal. | 2017-03-30T17:13:03 | 2017-03-30T17:13:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,595,817 | 684 | ztlpn | Most likely there are some problems in connecting from one server to another. Try searching in server logs for lines containing `Distributed.DirectoryMonitor`. By the way, why are you using IP addressed instead of host names? Is DNS configured properly?
Also note that recommended approach for data replication is to ... | 2017-04-12T14:30:51 | 2017-04-12T14:31:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,666,842 | 679 | ztlpn | I assume that you are talking about [this section in the docs](https://clickhouse.yandex/reference_en.html#Recovery%20after%20failures). Replicas store info about their local parts in ZooKeeper (each replica in its own ZooKeeper path). During initialization of the replicated table ClickHouse checks if this info is cons... | 2017-04-12T18:28:35 | 2017-04-12T18:28:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,303,379 | 700 | egorlitvinenko | Example data:
```
vjebywyypgqorighfaw,,898999574,,3.008874029012975E8
tjqevznxclm,,2079703005,,1.1591825521454709E9
aawethajnahzxgxzzdu,,1461070703,,7.682462476872058E8
feubrhhqaja,,1078610460,,8.908589890974745E8
hnjurnrlpcjxfa,,469686467,,6.0313867905135885E7
bqzqaphgvjj,,1133548469,,1.046040289828672E8
zcp... | 2017-04-15T16:25:27 | 2017-04-15T16:25:27 | {} | NONE |
ClickHouse | ClickHouse | 294,324,463 | 701 | alexey-milovidov | Everything is fine and ready to merge!
I will fix style and check some edge cases (when arrays have different size), after merging. | 2017-04-15T23:49:05 | 2017-04-15T23:49:05 | {} | MEMBER |
ClickHouse | ClickHouse | 294,441,307 | 215 | hatarist | Is there any way I could fix it myself? Could you describe the origin of the issue so I could dig into it? | 2017-04-17T09:42:39 | 2017-04-17T09:42:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,554,171 | 699 | alexey-milovidov | You always could specify time zone explicitly, for example:
```
toDate(1492124400, 'America/New_York')
```
```
2017-04-13
``` | 2017-04-17T18:35:59 | 2017-04-17T18:36:19 | {} | MEMBER |
ClickHouse | ClickHouse | 295,494,834 | 666 | alexey-milovidov | Please tell if our answer has helped you. | 2017-04-19T23:39:14 | 2017-04-19T23:39:14 | {} | MEMBER |
ClickHouse | ClickHouse | 290,331,370 | 644 | Jim-Raynor | Спасибо за ответ) | 2017-03-30T07:53:11 | 2017-03-30T07:53:11 | {} | NONE |
ClickHouse | ClickHouse | 290,386,578 | 645 | alexey-milovidov | Спасибо. Всё вместе - сойдёт. | 2017-03-30T11:44:51 | 2017-03-30T11:44:51 | {} | MEMBER |
ClickHouse | ClickHouse | 293,192,144 | 658 | ludv1x | Thank you for your suggestions.
We are updating our expression analyzer, it will enable to introduce more optimizations (including more powerful constants pruning).
Currently `if(const, expr1, expr2)` function supports it, you could try to rewrite the queries using `if`. | 2017-04-11T08:45:56 | 2017-04-11T08:45:56 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,488,104 | 679 | xingxing122 | Hello, I am also studying clickhouse, I built a distributed cluster, did you test the distributed cluster? | 2017-04-12T06:41:25 | 2017-04-12T06:41:25 | {} | NONE |
ClickHouse | ClickHouse | 294,001,956 | 695 | alexey-milovidov | That looks strange, but in fact, it is reasonable.
Memory usage for query shows difference between sizes of memory allocations and deallocations, that was made in context of query execution.
Usually, they are balanced: what was allocated during query execution, will be subseqently freed and so on.
But there are ... | 2017-04-13T19:40:11 | 2017-04-13T19:40:11 | {} | MEMBER |
ClickHouse | ClickHouse | 294,301,747 | 700 | ztlpn | You can use \N for NULL values. | 2017-04-15T15:55:56 | 2017-04-15T15:55:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 295,498,767 | 633 | alexey-milovidov | Did the answer has helped you? | 2017-04-19T23:51:58 | 2017-04-19T23:51:58 | {} | MEMBER |
ClickHouse | ClickHouse | 292,143,139 | 654 | valyala | This feature is orthogonal to the upcoming [ALTER TABLE DROP COLUMN FROM PARTITION](https://github.com/yandex/ClickHouse/pull/622), so the following additional feature could be implemented later:
```sql
ALTER TABLE ... DROP COLUMN ... FROM KEY RANGE ...
```
This would drop column values for the given key range only... | 2017-04-06T11:17:44 | 2017-04-06T11:17:44 | {} | NONE |
ClickHouse | ClickHouse | 292,726,457 | 662 | proller | autotests | 2017-04-08T15:50:41 | 2017-04-08T15:50:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,231,700 | 671 | ludv1x | More perfs: https://nda.ya.ru/3SqCD6
Some perfs
```bash
:~/benchmark/preferred_block_size_bytes$ clickhouse-client --preferred_block_size_bytes=0 --time -q "SELECT count() FROM hits_1000m_single WHERE URL LIKE '%metrika%' FORMAT Null"
6.894
:~/benchmark/preferred_block_size_bytes$ clickhouse-client --preferred_... | 2017-04-11T11:42:50 | 2017-04-11T11:42:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,682,704 | 690 | robot-metrika-test | Can one of the admins verify this patch? | 2017-04-12T19:29:04 | 2017-04-12T19:29:04 | {} | NONE |
ClickHouse | ClickHouse | 293,960,448 | 693 | robot-metrika-test | Can one of the admins verify this patch? | 2017-04-13T16:59:02 | 2017-04-13T16:59:02 | {} | NONE |
ClickHouse | ClickHouse | 294,313,716 | 701 | robot-metrika-test | Can one of the admins verify this patch? | 2017-04-15T19:39:02 | 2017-04-15T19:39:02 | {} | NONE |
ClickHouse | ClickHouse | 295,494,493 | 684 | alexey-milovidov | Please tell me if our answer has helped you. | 2017-04-19T23:38:26 | 2017-04-19T23:38:26 | {
"heart": 1
} | MEMBER |
ClickHouse | ClickHouse | 290,086,183 | 615 | alexey-milovidov | Sorry, your code is wrong.
Percent doesn't shows if it is greater than 50.
Look at
```
if (show_progress_bar
|| (width_of_progress_bar > 0
&& progress.total_rows
&& elapsed_ns > 500000000
&& progress.rows * 2 < progress.total_rows))
{
show_progress_bar = true;
```
| 2017-03-29T13:16:08 | 2017-03-29T13:16:08 | {} | MEMBER |
ClickHouse | ClickHouse | 290,462,769 | 242 | ztlpn | The problem was not with tzdata itself but with incorrect detection of the system timezone by ClickHouse (instead of 'Asia/Vladivostok' it resolved to '+10'). This should be fixed in v1.1.54198.
Note: system timezone can be overridden in [the server config](https://github.com/yandex/ClickHouse/blob/master/dbms/src/S... | 2017-03-30T16:18:20 | 2017-03-30T16:18:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 292,615,249 | 64 | alexey-milovidov | > What are the performance implications of doing queries spanning a month over a merge table that has 365 x 24 tables under it?
Every query, even simple one, will require to do at least 365 * 24 * num_used_columns disk seeks.
For short-running queries this will be very expensive.
> The docs for merge engine say ... | 2017-04-07T18:27:13 | 2017-04-07T18:27:13 | {} | MEMBER |
ClickHouse | ClickHouse | 293,766,672 | 684 | xingxing122 | Has added resolution in / etc / hosts, clickhouse's distributed query is clustered? Can I create multiple tables at once as a distributed query? The official website of the document description is not very detailed | 2017-04-13T03:10:53 | 2017-04-13T03:10:53 | {} | NONE |
ClickHouse | ClickHouse | 293,974,456 | 658 | hagen1778 | I did some tests and can confirm that `if` statement helps | 2017-04-13T17:53:51 | 2017-04-13T17:53:51 | {} | NONE |
ClickHouse | ClickHouse | 294,115,998 | 379 | atlas-comstock | Thanks @alexey-milovidov
It's of great help. | 2017-04-14T08:32:42 | 2017-04-14T08:32:42 | {} | NONE |
ClickHouse | ClickHouse | 294,552,760 | 701 | alexey-milovidov | > Perhaps we could add in the comments of IAggreateFunction::add and IAggregateFunction::insertResultInto that each AggregateDataPtr place will be used for at most one aggregation key (but one aggregation key could correspond to several AggregateDataPtr places)?
Aggregation state could not be shared for multiple key... | 2017-04-17T18:30:44 | 2017-04-17T18:30:44 | {} | MEMBER |
ClickHouse | ClickHouse | 290,483,050 | 647 | robot-metrika-test | Can one of the admins verify this patch? | 2017-03-30T17:29:03 | 2017-03-30T17:29:03 | {} | NONE |
ClickHouse | ClickHouse | 292,910,867 | 666 | ztlpn | Yes, you can change substitutions file in the main config.xml by setting `include_from` parameter. Like this:
```xml
<?xml version="1.0"?>
<yandex>
<include_from>/path/to/your/substitutions.xml</include_from>
...
``` | 2017-04-10T10:27:57 | 2017-04-10T10:27:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,029,646 | 520 | alexey-milovidov | New settings are default. | 2017-04-13T21:43:56 | 2017-04-13T21:43:56 | {} | MEMBER |
ClickHouse | ClickHouse | 294,333,321 | 701 | alexey-milovidov | https://github.com/yandex/ClickHouse/commit/83ed2a794bfff2fae245b647fdde9348a00aaf60
https://github.com/yandex/ClickHouse/commit/d9793bf5fea890e936c868209a71896cdd4adbe9
https://github.com/yandex/ClickHouse/commit/6b5e6104a9db556d907b4a73f25f272a20d2d4bf
https://github.com/yandex/ClickHouse/commit/3fa1639877eb94a0ce... | 2017-04-16T04:40:40 | 2017-04-16T04:40:40 | {} | MEMBER |
ClickHouse | ClickHouse | 289,734,995 | 614 | ludv1x | Note that currently 429th test uses `clickhouse-compressor` if it is found.
Maybe make sense add `clickhouse --compressor` option. | 2017-03-28T10:59:44 | 2017-03-28T10:59:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 289,839,776 | 640 | George3d6 | --input_format_allow_errors_num
just realized this exists, really sorry for the issue pollution | 2017-03-28T17:13:09 | 2017-03-28T17:13:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 289,852,769 | 617 | alexey-milovidov | You need to provide complete cmake output (now we see just few tail lines). | 2017-03-28T17:57:45 | 2017-03-28T17:57:45 | {} | MEMBER |
ClickHouse | ClickHouse | 289,877,708 | 360 | proller | try search error in cmake log:
grep -b20 -a20 HAVE_READLINE_HISTORY build/CMakeFiles/CMakeError.log
| 2017-03-28T19:25:06 | 2017-03-28T19:25:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 291,001,769 | 652 | alexey-milovidov | Manually merged: https://github.com/yandex/ClickHouse/pull/652/commits/b9c740d47a0a07bf13eb06037e58f6765ccfe56d | 2017-04-02T17:41:20 | 2017-04-02T17:41:20 | {} | MEMBER |
ClickHouse | ClickHouse | 291,991,888 | 495 | sharp-dev | clickhouse вроде российское произведение.. вы простите, но я не знаю английского, можно мне в 52 года не учить с нуля?. но вот хочу попользоваться. но запусти блин не могу. и советы по установке 'US/Central' or 'America/Chicago' мне как то не помогают. можно для не грамотного, но своего, подсказку как запустить в таймз... | 2017-04-05T20:47:38 | 2017-04-05T20:58:48 | {} | NONE |
ClickHouse | ClickHouse | 292,141,587 | 654 | valyala | Additionally the following feature could be added to be consistent with `DROP/DETACH PARTITION`, so the detached key range could be archived and/or copied to another place.
```sql
ALTER TABLE ... DETACH KEY RANGE ...
``` | 2017-04-06T11:09:50 | 2017-04-06T11:09:50 | {} | NONE |
ClickHouse | ClickHouse | 294,225,243 | 690 | alexey-milovidov | :+1: | 2017-04-14T20:01:00 | 2017-04-14T20:01:00 | {} | MEMBER |
ClickHouse | ClickHouse | 290,997,228 | 652 | robot-metrika-test | Can one of the admins verify this patch? | 2017-04-02T16:29:03 | 2017-04-02T16:29:03 | {} | NONE |
ClickHouse | ClickHouse | 292,581,897 | 661 | robot-metrika-test | Can one of the admins verify this patch? | 2017-04-07T16:19:03 | 2017-04-07T16:19:03 | {} | NONE |
ClickHouse | ClickHouse | 294,333,336 | 701 | alexey-milovidov | Syntetic performance test:
```
SELECT sumForEach([number, number]) FROM system.numbers_mt
```
(performance is Ok) | 2017-04-16T04:41:10 | 2017-04-16T04:41:10 | {} | MEMBER |
ClickHouse | ClickHouse | 294,472,523 | 704 | robot-metrika-test | Can one of the admins verify this patch? | 2017-04-17T12:04:02 | 2017-04-17T12:04:02 | {} | NONE |
ClickHouse | ClickHouse | 295,493,775 | 712 | alexey-milovidov | This is intended as "safety threshold".
With large enough difference between local data and reference data in ZK, by default, the server refuses to start, to allow manual restore. To activate automatic restore, you could set `force_restore_data` flag before starting the server. This is as simple as:
```
sudo -u cl... | 2017-04-19T23:36:01 | 2017-04-19T23:36:01 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 289,720,851 | 617 | sangli00 | 64 bit | 2017-03-28T09:54:38 | 2017-03-28T09:54:38 | {} | NONE |
ClickHouse | ClickHouse | 289,849,380 | 614 | alexey-milovidov | > Maybe make sense add clickhouse --compressor option.
Yes, this is reasonable. | 2017-03-28T17:46:32 | 2017-03-28T17:46:32 | {} | MEMBER |
ClickHouse | ClickHouse | 290,307,583 | 644 | alexey-milovidov | С PCRE слишком легко создать регексп, который будет выполняться очень долго или даже уронит сервер. Хотя мы могли бы добавить поддержку PCRE, даже во все сборки по-умолчанию, но включаемую отдельной опцией в конфиге или настройкой для пользователя.
В коде уже присутсвует PCRE (внутри Poco). Хотя если добавлять такие... | 2017-03-30T05:38:24 | 2017-03-30T05:38:24 | {} | MEMBER |
ClickHouse | ClickHouse | 292,639,357 | 663 | robot-metrika-test | Can one of the admins verify this patch? | 2017-04-07T20:09:02 | 2017-04-07T20:09:02 | {} | NONE |
ClickHouse | ClickHouse | 293,083,209 | 670 | basiliscos | Yep, confirmed. Thanks you! | 2017-04-10T21:23:12 | 2017-04-10T21:23:12 | {} | NONE |
ClickHouse | ClickHouse | 293,288,480 | 676 | proller | also here true stable branch: https://github.com/yandex/ClickHouse/commits/stable | 2017-04-11T14:51:30 | 2017-04-11T14:51:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,682,697 | 691 | robot-metrika-test | Can one of the admins verify this patch? | 2017-04-12T19:29:02 | 2017-04-12T19:29:02 | {} | NONE |
ClickHouse | ClickHouse | 293,885,754 | 684 | ztlpn | The exception message points at the cause of the problem: you have configured a cluster named `enncloud`, but used a (nonexistent) cluster named `perftest_3shards_3replicas` in the definition of the Distributed table.
Using the example from the tutorial, correct table definition in your case is
```
CREATE TABLE on... | 2017-04-13T12:48:21 | 2017-04-13T12:48:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,019,982 | 520 | alexey-milovidov | Marek, please, share your observations if will be any. | 2017-04-13T20:58:56 | 2017-04-13T20:58:56 | {} | MEMBER |
ClickHouse | ClickHouse | 294,475,679 | 214 | hatarist | Hm. It seems that it arrived again (or wasn't fixed).
Reproducible on 1.1.54180 stable and 1.1.54214 testing.
```
:) CREATE TABLE test (date Date DEFAULT today(), fs FixedString(16)) ENGINE = MergeTree(date, (date, fs), 8192);
:) INSERT INTO test (fs) VALUES ('54db0d43009d\0\0\0\0'), ('fe2b58224766cf10')
... | 2017-04-17T12:32:25 | 2017-04-17T12:51:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 290,385,507 | 645 | isublimity | К сожалению коммит залез "Incriment max_size in Dictionaries:RegionsNames & RegionsHierarchy " который хотел оформить отдельным PR, пересобрать или можно так ?
Увеличение max_size обсуждали в телеграм чате
> @garikanet
Мы обновили CH до версии: 1.1.54140.
Перестали загружаться встроенные гео-словари со след... | 2017-03-30T11:39:25 | 2017-03-30T11:39:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 290,950,323 | 64 | daledude | What are the performance implications of doing queries spanning a month over a merge table that has 365 x 24 tables under it? The docs for merge engine say queries are parallel so there will be 30 x 24 queries run at the same time? | 2017-04-01T22:08:38 | 2017-04-01T22:08:38 | {} | NONE |
ClickHouse | ClickHouse | 294,360,183 | 701 | mfridental | Wow, that was quick, thank you so much!
I was trying to understand today what's wrong with using Arena::allocContinue, could you please explain?
About too much work in insertResultInto, I was under impression that one AggregateDataPtr can point to results for several aggregation keys. Now in the hindsight, its ob... | 2017-04-16T16:18:47 | 2017-04-16T16:18:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 291,779,059 | 642 | sangli00 | so,If i create dummy date,Data is not balance in every node. | 2017-04-05T07:35:19 | 2017-04-05T07:35:19 | {} | NONE |
ClickHouse | ClickHouse | 293,683,801 | 687 | ztlpn | Yes, that would be great! Unfortunately this feature requires substantial reworking of the query execution engine. The good news is that we actually plan to do it so stay tuned. | 2017-04-12T19:33:34 | 2017-04-12T19:33:34 | {
"+1": 6
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,892,337 | 684 | ztlpn | Also, I've noticed that you have configured two different shards to reside on the same server (e.g. server `clickhouse01` will contain data for both shards). This doesn't make much sense. Sharding is needed when you want to store different parts of the data set on different servers (not when you want to store the same ... | 2017-04-13T13:16:58 | 2017-04-13T13:16:58 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,304,438 | 700 | ztlpn | > What choises are available to import it?
CSV parser does not recognize empty string as a NULL value, so you will have to manually insert '\N' where you want NULLs to be. | 2017-04-15T16:43:28 | 2017-04-15T16:43:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,330,602 | 701 | alexey-milovidov | In fact, the code is somewhat wrong. I am preparing a fix right now. | 2017-04-16T03:05:02 | 2017-04-16T03:05:28 | {} | MEMBER |
ClickHouse | ClickHouse | 294,371,929 | 702 | alexey-milovidov | Great! Now 80% translated. | 2017-04-16T20:18:34 | 2017-04-16T20:18:34 | {} | MEMBER |
ClickHouse | ClickHouse | 290,381,304 | 645 | robot-metrika-test | Can one of the admins verify this patch? | 2017-03-30T11:19:02 | 2017-03-30T11:19:02 | {} | NONE |
ClickHouse | ClickHouse | 293,810,457 | 684 | xingxing122 | Regardless of the use of the host name or ip this is a failure,
```
:) select * from system.clusters
SELECT *
FROM system.clusters
┌─cluster ─ ─ fear_shard_num ─ fear ─shard_weight─ fear ─replica_num ─ fear --host_name ─ ─ ─ ─ fear-host_address ─ ┬─port ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ... | 2017-04-13T07:07:53 | 2017-04-13T13:04:42 | {} | NONE |
ClickHouse | ClickHouse | 293,958,772 | 520 | proller | already in master:
```
<merge_tree>
<replicated_max_parallel_sends_for_table>2</replicated_max_parallel_sends_for_table>
<replicated_max_parallel_sends>4</replicated_max_parallel_sends>
<replicated_max_parallel_fe... | 2017-04-13T16:52:46 | 2017-04-13T20:59:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,307,072 | 699 | ztlpn | Yes, this is a bug/limitation of how timezones currently work. The problem is that the client cannot set the server-side session timezone. All it can change is the timezone of client-side operations (that is, parsing and formatting).
Here is a good mental model to help you avoid such surprises:
* There are two time... | 2017-04-15T17:29:48 | 2017-04-15T17:29:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,333,425 | 701 | alexey-milovidov | The combinator also works for parametric aggregate functions, example:
```
SELECT
k,
groupArrayForEach(arr),
quantilesExactForEach(0.5, 0.9)(arr)
FROM
(
SELECT
intDiv(number, 3) AS k,
arrayMap(x -> (number + x), range(number)) AS arr
FROM system.numbers
LIMIT 1... | 2017-04-16T04:44:09 | 2017-04-16T04:44:09 | {} | MEMBER |
ClickHouse | ClickHouse | 295,491,856 | 713 | alexey-milovidov | Hi.
Rollup should work regardless the choice of aggregate function.
Calculation cost of 'avg' is about the same as 'any'.
Do you have any related errors in logs?
Also, probably you use 'avg' for older data and background merges doesn't go for data of that age?
In that case, it is needed to run OPTIMIZE TABLE .... | 2017-04-19T23:30:10 | 2017-04-19T23:30:10 | {} | MEMBER |
ClickHouse | ClickHouse | 290,617,506 | 647 | alexey-milovidov | Thanks! | 2017-03-31T05:11:37 | 2017-03-31T05:11:37 | {} | MEMBER |
ClickHouse | ClickHouse | 292,009,342 | 495 | ztlpn | @sharp-dev Здравствуйте. ClickHouse поддерживает наименования таймзон из IANA timezone database, вот полный список: https://en.wikipedia.org/wiki/List_of_tz_database_time_zones. Если под RTZ2 имеется в виду московское время, то его идентификатор Europe/Moscow. В большинстве случаев достаточно правильно настроить систем... | 2017-04-05T21:57:14 | 2017-04-05T21:57:14 | {
"+1": 5
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,682,712 | 689 | robot-metrika-test | Can one of the admins verify this patch? | 2017-04-12T19:29:05 | 2017-04-12T19:29:05 | {} | NONE |
ClickHouse | ClickHouse | 294,305,606 | 700 | egorlitvinenko | @ztlpn thanks for your help :) | 2017-04-15T17:03:31 | 2017-04-15T17:03:43 | {} | NONE |
ClickHouse | ClickHouse | 294,494,495 | 215 | alexey-milovidov | It is needed to create a stream, that will apply type conversions.
Look at `InterpreterInsertQuery::execute`. There are similar things: `AddingDefaultBlockOutputStream`, `SquashingBlockOutputStream`, that do some transformations on INSERT.
I hope that https://github.com/artpaul could implement it. | 2017-04-17T14:21:33 | 2017-04-17T14:21:42 | {} | MEMBER |
ClickHouse | ClickHouse | 290,463,001 | 495 | ztlpn | Fixed in v1.1.54198. | 2017-03-30T16:19:11 | 2017-03-30T16:19:11 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,278,315 | 676 | George3d6 | I was running on an unsupported OS so I just cloned github and compiled, my bad :/
I will use the latest stable one from now, thank you very much for spotting this, I might have spend hours wondering what the hell was going on. | 2017-04-11T14:19:53 | 2017-04-11T14:19:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,356,562 | 702 | robot-metrika-test | Can one of the admins verify this patch? | 2017-04-16T15:09:02 | 2017-04-16T15:09:02 | {} | NONE |
ClickHouse | ClickHouse | 289,802,682 | 360 | smiren | Hi, why I've got this test faild on MacOS 10.12:
-- Performing Test HAVE_READLINE_HISTORY
-- Performing Test HAVE_READLINE_HISTORY - Failed
And as I can see all examples above have the same issue
Of course readline is installed | 2017-03-28T15:13:26 | 2017-03-28T15:13:26 | {} | NONE |
ClickHouse | ClickHouse | 289,837,031 | 637 | El-76 | OMFG, it works! Thank you for so prompt answer! | 2017-03-28T17:03:15 | 2017-03-28T17:03:15 | {} | NONE |
ClickHouse | ClickHouse | 290,056,924 | 643 | robot-metrika-test | Can one of the admins verify this patch? | 2017-03-29T11:04:02 | 2017-03-29T11:04:02 | {} | NONE |
ClickHouse | ClickHouse | 293,061,243 | 670 | proller | should be fixed here https://github.com/yandex/ClickHouse/commit/13e9b5c94977b5c3b9b43859d3b9d7cf1d584943 | 2017-04-10T19:58:14 | 2017-04-10T19:58:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,025,414 | 520 | vavrusa | Will do on the next replication, thanks!
Is the new merge tree settings default or should I set it explicitly? | 2017-04-13T21:23:51 | 2017-04-13T21:23:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 294,125,415 | 247 | pavel-odintsov | It would be nice to have special storage type to store IPv4 and IPv6 subnets. For IPv4 subnets, we need 32 bits for the network address and uint8 to store CIDR mask length (valid range 0-32).
Yes, I could store this data in two fields but it does not look as great solution. Also I want to have ability to extract / h... | 2017-04-14T09:35:19 | 2017-04-14T09:35:19 | {
"+1": 7
} | NONE |
ClickHouse | ClickHouse | 294,302,882 | 700 | egorlitvinenko | Do you have an example?
:) INSERT INTO hello_world.hello_world_table(f2,f5,f3,f1,f4) FORMAT TabSeparated \N 7.81651003481571E8 1886081650 zlwwqlbkvrksdt \N;
INSERT INTO hello_world.hello_world_table (f2, f5, f3, f1, f4) FORMAT TabSeparated
Exception on client:
Code: 27. DB::Exception: Cannot parse input: ex... | 2017-04-15T16:15:57 | 2017-04-15T16:16:53 | {} | NONE |
ClickHouse | ClickHouse | 294,550,246 | 701 | alexey-milovidov | > I was trying to understand today what's wrong with using Arena::allocContinue, could you please explain?
`allocContinue` is intended only for resizing just allocated memory region. The purpose is for almost free resizing in most cases, when just allocated region could be expanded inside same chunk. Example usage -... | 2017-04-17T18:21:38 | 2017-04-17T18:22:56 | {} | MEMBER |
ClickHouse | ClickHouse | 295,166,588 | 223 | bzed | https://github.com/yandex/graphouse/ | 2017-04-19T08:41:30 | 2017-04-19T08:41:30 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 289,801,100 | 637 | alexey-milovidov | Please update to latest version: 1.1.54190. | 2017-03-28T15:08:36 | 2017-03-28T15:08:36 | {} | MEMBER |
ClickHouse | ClickHouse | 293,277,069 | 676 | proller | you catch broken version.
should be fixed here - https://github.com/yandex/ClickHouse/commit/38e9254d330a5a791e84fe0a2921ac00cd0226ff
for production better to use stable tags (like https://github.com/yandex/ClickHouse/tree/v1.1.54198-stable )
| 2017-04-11T14:15:53 | 2017-04-11T14:15:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,859,096 | 679 | George3d6 | "Comparisons with other replicas are also performed - after downloading a new part or merging two parts."... Ok, so this first mechanism is independent of how many replicas you have whilst the second one, I'd assume, require 3 or more, am I right in that assumption ?
xingxing9688
Currently I'm planning on working w... | 2017-04-13T11:02:41 | 2017-04-13T11:02:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,901,411 | 630 | proller | now on my ipv6-only jail
<listen_host>::1</listen_host>
<listen_host>127.0.0.1</listen_host>
=>
[ 1 ] <Error> Application: Net Exception: Protocol not supported
```
<listen_host>::</listen_host> :
# sockstat
USER COMMAND PID FD PROTO LOCAL ADDRESS FOREIGN ADDRESS
clickhouse clickhou... | 2017-04-13T13:52:00 | 2017-04-13T13:52:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 293,961,741 | 694 | robot-metrika-test | Can one of the admins verify this patch? | 2017-04-13T17:04:02 | 2017-04-13T17:04:02 | {} | NONE |
ClickHouse | ClickHouse | 295,498,402 | 616 | alexey-milovidov | Hi.
This is possible, if some data was inserted to Distributed table before password was specified in cluster configuration. The data inserted to Distributed tabl is written asynchronously - first it is written to filesystem on local server (look at data directory for table `/path_to_clickhouse/data/db/table/`), and... | 2017-04-19T23:50:42 | 2017-04-19T23:50:42 | {
"+1": 1
} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.