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 | 424,741,003 | 3,223 | alexey-milovidov | We should test for the maximum performance degradation. I think, this optimization makes sense. | 2018-09-26T14:40:10 | 2018-09-26T14:40:10 | {} | MEMBER |
ClickHouse | ClickHouse | 425,193,004 | 1,981 | healiseu | I got the same problem with @jgshzy, but for a different reason. The first time I installed clickhouse the service started without any error. But then I decided to move clickhouse data repository to an SSD disk. I modified paths at
/etc/clickhouse-server/config.xml
```xml
<!-- Path to data directory, with trailing s... | 2018-09-27T18:17:18 | 2018-09-27T21:37:27 | {} | NONE |
ClickHouse | ClickHouse | 425,349,711 | 3,242 | filimonov | Javascript uses Floats for big numbers. And it's completely ok to use floats for percent calculations:
```javascript
x = JSON.parse('{"read_rows":"12292599800000000000000","read_bytes":"1884341","total_rows":"122925998000000000000000"}');
percent = 100 * x.read_rows / x.total_rows;
```
| 2018-09-28T07:37:27 | 2018-09-28T07:37:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,491,064 | 3,196 | tkroman | I was looking at the code and attempted a fix but unfortunately didn't succeed.
What I've discovered so far:
at this point:
https://github.com/yandex/ClickHouse/blob/master/dbms/src/Interpreters/InJoinSubqueriesPreprocessor.cpp#L152
there is an assertion that the db is still distributed. then we go into `forEachN... | 2018-09-28T16:26:41 | 2018-09-28T16:26:41 | {} | NONE |
ClickHouse | ClickHouse | 425,305,987 | 3,142 | alexey-milovidov | In master. | 2018-09-28T03:01:25 | 2018-09-28T03:01:25 | {} | MEMBER |
ClickHouse | ClickHouse | 425,839,987 | 3,259 | 4ertus2 | There're no plans for real Update for now. At least for MergeTree engines.
ClickHouse support ALTER UPDATE syntax for rare data manipulations. Look at block Mutations at this page https://clickhouse.yandex/docs/en/query_language/alter/
| 2018-10-01T09:13:37 | 2018-10-01T09:13:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 426,489,869 | 3,271 | vavrusa | Do you see any errors in the log for the INSERT statements or materialization?
The materialization into views is not atomic and there's no rollback, so if you materialization succeeds for some views and fails for some others, you might end up double inserting to the views that succeeded on insert retry. | 2018-10-03T02:13:19 | 2018-10-03T02:13:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,374,869 | 3,242 | filimonov | Please check that: https://github.com/yandex/ClickHouse/issues/208#issuecomment-422406339
https://github.com/yandex/ClickHouse/issues/2825 | 2018-09-28T09:15:07 | 2018-09-28T09:17:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,770,233 | 3,208 | alexey-milovidov | Missing test that shows that it works with both values of setting.
Could you please make a separate PR with this test? | 2018-10-01T01:47:39 | 2018-10-01T01:47:39 | {} | MEMBER |
ClickHouse | ClickHouse | 425,867,870 | 3,260 | filimonov | Can you check if /var/lib/clickhouse/metadata/mobile//events_773292_20180926.sql is accessible in docker, and contain proper ATTACH TABLE statement? | 2018-10-01T10:59:44 | 2018-10-01T11:00:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 426,578,539 | 3,277 | filimonov | ```sql
CREATE TABLE sum_map_string(
date Date,
timeslot DateTime,
statusMap Nested(
status String,
requests UInt64
)
) ENGINE = Log;
INSERT INTO sum_map_string VALUES
('2000-01-01', '2000-01-01 00:00:00', ['a', 'b', 'c'], [10, 10, 10]),
('2000-01-01', '2000-01-01 00:... | 2018-10-03T09:59:36 | 2018-10-03T10:00:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,200,651 | 3,239 | RomanGorbatko | > @RomanGorbatko you could implement this by explicitly JOINing these subqueries
As i know i cant use mysql-like syntax around `JOIN`.
So, i cant do something like this:
```sql
select
count(v.Page) as current,
v.Page,
count(next)
from
test as v
join test as next on next.PrevPage = v.Pag... | 2018-09-27T18:41:48 | 2018-09-27T18:41:48 | {} | NONE |
ClickHouse | ClickHouse | 425,521,494 | 3,244 | jimmykuo | @filimonov
thank you! help me a lot | 2018-09-28T18:11:36 | 2018-09-28T18:11:36 | {} | NONE |
ClickHouse | ClickHouse | 425,545,672 | 3,198 | snorecone | We are also seeing this issue. Clickhouse currently has 456,357 files open; they are mostly netlink sockets:
```
clickhous 5174 clickhouse *377u sock 0,8 0t0 41551004 protocol: NETLINK
clickhous 5174 clickhouse *378u sock 0,8 0t0 41551005 protocol: NETLINK
clickho... | 2018-09-28T19:45:11 | 2018-09-28T19:45:11 | {
"+1": 5
} | NONE |
ClickHouse | ClickHouse | 424,848,507 | 3,208 | alexey-milovidov | Almost Ok. | 2018-09-26T19:55:34 | 2018-09-26T19:55:34 | {} | MEMBER |
ClickHouse | ClickHouse | 425,251,703 | 2,788 | vavrusa | This should be fixed by https://github.com/yandex/ClickHouse/commit/7eb25ef55b9f016d2bf440a277251d2d3d8b4c3e | 2018-09-27T21:39:29 | 2018-09-27T21:39:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,547,774 | 2,904 | champtar | done in #3138 | 2018-09-28T19:53:41 | 2018-09-28T19:53:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,631,694 | 3,248 | qietingfengling | thanks | 2018-09-29T09:38:07 | 2018-09-29T09:38:07 | {} | NONE |
ClickHouse | ClickHouse | 425,693,503 | 3,257 | zhang2014 | @amosbird FYI, I hope that'll be helpful : )
| 2018-09-30T04:18:19 | 2018-09-30T04:18:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,704,645 | 3,257 | amosbird | hmm, simple copy paste of that `ё` just works. Looks a bit weird in terminal though
| 2018-09-30T08:36:59 | 2018-09-30T08:36:59 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 426,016,408 | 3,257 | alexey-milovidov | And finally... look for the test
00298_enum_width_and_cast
https://travis-ci.org/yandex/ClickHouse/builds/435426678#L8971 | 2018-10-01T18:38:13 | 2018-10-01T18:38:13 | {} | MEMBER |
ClickHouse | ClickHouse | 426,147,234 | 888 | filimonov | #3257 | 2018-10-02T04:42:29 | 2018-10-02T04:42:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 426,269,749 | 3,266 | 4ertus2 | Hello,
You could make a pull request for it. It's quite simple to add a new aggregation function in ClickHouse. Look at https://github.com/yandex/ClickHouse/blob/master/dbms/src/AggregateFunctions/AggregateFunctionAvg.cpp#L44
You need add analogs for AggregateFunctionAvg.cpp and AggregateFunctionAvg.h files for y... | 2018-10-02T13:17:44 | 2018-10-02T13:17:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 424,835,490 | 3,212 | alexey-milovidov | Did you mean 18.14.0-testing? | 2018-09-26T19:11:09 | 2018-09-26T19:11:09 | {} | MEMBER |
ClickHouse | ClickHouse | 425,115,314 | 3,209 | topvisor-admin | Аналогичная проблема с ReplacingMergeTree, FINAL, PREWHERE.
Добавлю, что если date_check запрашивается в SELECT или используется в GROUP BY, то ошибки нет.
см. #2827
v18.12.17 | 2018-09-27T14:32:13 | 2018-09-27T14:46:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,196,267 | 3,239 | blinkov | @RomanGorbatko you could implement this by explicitly JOINing these subqueries | 2018-09-27T18:27:54 | 2018-09-27T18:27:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,354,096 | 3,243 | perez1987 | I prefer first option | 2018-09-28T07:55:56 | 2018-09-28T07:55:56 | {} | NONE |
ClickHouse | ClickHouse | 425,767,163 | 3,237 | alexey-milovidov | Submodules have leaked. | 2018-10-01T01:15:42 | 2018-10-01T01:15:42 | {} | MEMBER |
ClickHouse | ClickHouse | 426,419,913 | 29 | ibrarahmad | What are the native options available? I really don't want to use third party solution. | 2018-10-02T20:33:28 | 2018-10-02T20:33:28 | {} | NONE |
ClickHouse | ClickHouse | 424,889,007 | 3,208 | vavrusa | Updated, it's now behind `allow_concurrent_view_processing` setting, and disabled by default. | 2018-09-26T22:22:55 | 2018-09-26T22:22:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,506,810 | 253 | 4ertus2 | There's experimental support in current release.
https://clickhouse.yandex/docs/en/data_types/decimal/ | 2018-09-28T17:21:03 | 2018-09-28T17:21:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,608,673 | 3,256 | champtar | @KochetovNicolai | 2018-09-29T02:28:58 | 2018-09-29T02:28:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,659,440 | 3,257 | alexey-milovidov | Missing test.
Test should have full-width characters (Chinese), zero-width characters,
two variants of Russian letter ё:
ё - d191 in UTF-8
ё - d0b5cc88 in UTF-8
as long as some invalid UTF-8 sequence,
and a string with zero bytes in the middle. | 2018-09-29T16:54:31 | 2018-09-29T17:01:56 | {} | MEMBER |
ClickHouse | ClickHouse | 426,122,700 | 3,257 | amosbird | OK, updated 298's result | 2018-10-02T02:07:18 | 2018-10-02T02:07:18 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 426,589,194 | 3,210 | alesapin | Style Erros:
```
dbms/src/TableFunctions/ITableFunctionXDBC.h:38: const std::string & connection_string_) const override {
dbms/src/TableFunctions/ITableFunctionXDBC.h:55: const std::string & connection_string_) const override {
dbms/src/Common/XDBCBridgeHelper.h:238:struct ODBCBridgeMixin {
```
Build logs:
h... | 2018-10-03T10:38:33 | 2018-10-03T10:40:44 | {} | MEMBER |
ClickHouse | ClickHouse | 426,602,008 | 3,279 | proller | interserver_http_host not used for listen address.
Hostname that is used by other replicas to request this server.
If not specified, than it is determined analoguous to 'hostname -f' command.
This setting could be used to switch replication to another network interface.
| 2018-10-03T11:27:18 | 2018-10-03T11:30:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 426,605,976 | 3,279 | proller | its only for sending to remote replicas as our hostname. | 2018-10-03T11:42:20 | 2018-10-03T11:42:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 424,790,250 | 3,221 | e11it | Спасибо!
У нас эта проблема появилась после того, как мы перевели часть колонок в Decimal и натравили Табло. Для вычисляемых полей он генерит код, например:
```sql
SELECT multiIf((sold_14_days_rub AS sold_14_days_rub) = 0, NULL, ((stock AS stock) * (sebestoimost AS sebestoimost)) / sold_14_days_rub) AS Calculation... | 2018-09-26T16:51:38 | 2018-09-26T16:51:38 | {} | NONE |
ClickHouse | ClickHouse | 424,863,671 | 3,179 | alexey-milovidov | In master. | 2018-09-26T20:45:12 | 2018-09-26T20:45:12 | {} | MEMBER |
ClickHouse | ClickHouse | 425,178,398 | 3,232 | CurtizJ | Баг при прокидывании условия внутрь подзапроса. Потому что данный запрос работает при enable_optimize_predicate_expression=0. В новой версии эта настройка включена по умолчанию. | 2018-09-27T17:34:10 | 2018-09-27T17:37:33 | {} | MEMBER |
ClickHouse | ClickHouse | 425,251,151 | 2,796 | vavrusa | This should be fixed in https://github.com/yandex/ClickHouse/pull/3215 | 2018-09-27T21:37:04 | 2018-09-27T21:37:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,346,776 | 3,242 | alex-zaitsev | @Enmk is going to add percents. | 2018-09-28T07:24:59 | 2018-09-28T07:24:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,400,027 | 3,245 | alexey-milovidov | These stack traces look like the result of double free. We had fixed some similar bugs recently. I doubt that I can identify the exact case that was fixed. | 2018-09-28T10:56:48 | 2018-09-28T10:56:48 | {} | MEMBER |
ClickHouse | ClickHouse | 425,699,400 | 3,257 | amosbird | @zhang2014 Thanks. Those `ё`s are all of width 1 from what I can tell. I assume there is a variant that need special treatment. | 2018-09-30T06:55:23 | 2018-09-30T06:55:23 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 425,770,660 | 3,258 | alexey-milovidov | Looks like a flap. It is not passed on your commit, but it is passed on my merge commit with your pull request. | 2018-10-01T01:52:05 | 2018-10-01T01:52:05 | {} | MEMBER |
ClickHouse | ClickHouse | 426,322,253 | 3,267 | alesapin | Changes doesn't affect failed test. | 2018-10-02T15:39:24 | 2018-10-02T15:39:24 | {} | MEMBER |
ClickHouse | ClickHouse | 426,581,458 | 3,277 | filimonov | And simpler/more general:
```
-- those work:
select sumMap(val, cnt) from ( SELECT [ CAST(1, 'UInt64') ] as val, [1] as cnt );
select sumMap(val, cnt) from ( SELECT [ CAST(1, 'Float64') ] as val, [1] as cnt );
select sumMap(val, cnt) from ( SELECT [ CAST('a', 'Enum16(\'a\'=1)') ] as val, [1] as cnt );
-- t... | 2018-10-03T10:10:58 | 2018-10-03T10:10:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 426,588,289 | 3,276 | 4ertus2 | Does it work with LLVM disabled?
> cmake -DENABLE_EMBEDDED_COMPILER=0 -DUSE_INTERNAL_LLVM_LIBRARY=0
Note: you need to remove build directory or CMakeCache.txt to apply them. | 2018-10-03T10:34:18 | 2018-10-03T10:34:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 424,454,930 | 3,212 | champtar | see #3227
This fails in 18.14.0:
```
SELECT
geoip_latitude AS latitude,
geoip_longitude AS longitude,
dictGetString('geoname', 'city_name', toUInt64(geoip_geoname_id)) AS city,
uniq(clientip) AS uniqueip
FROM analytics.vme1
WHERE (date >= toDate(1530122160)) AND (datetime >= toDateTime(... | 2018-09-25T18:38:22 | 2018-09-26T19:22:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 424,839,124 | 3,212 | champtar | yep (i've just edited all comments) | 2018-09-26T19:23:17 | 2018-09-26T19:23:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,374,503 | 3,242 | filimonov | Actually, there are different problems with progress in HTTP.
For example when you get a streaming output (like 'SELECT columns FROM big_table PREWHERE condition_meet_thousands_of_rows') and server buffer is full server start sending data and stop sending progress (even if you still need to wait long).
Or ther... | 2018-09-28T09:13:42 | 2018-09-28T09:20:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,487,619 | 3,245 | dqminh | i went back our logs and there are more segmentation faults. https://gist.github.com/dqminh/d0bbac6ef713e68495309b1a4fa59f62 | 2018-09-28T16:15:04 | 2018-09-28T16:15:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,667,275 | 3,255 | champtar | If you don't known patch version, might I suggest not displaying it, ie showing "18.14" or "18.14.x" or "18.14.?" | 2018-09-29T18:46:53 | 2018-09-29T18:46:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,741,023 | 3,233 | josepowera | At the moment we think, that the error was caused in that we have not added field in PARTITION BY fields to ORDER BY fields list.
| 2018-09-30T18:29:27 | 2018-09-30T18:29:27 | {} | NONE |
ClickHouse | ClickHouse | 425,973,810 | 3,208 | vavrusa | Will do @alexey-milovidov | 2018-10-01T16:31:55 | 2018-10-01T16:31:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 426,602,000 | 3,276 | SlavaNovikov | Thanks, I ran the command, but now I ran into this:
` ninja -v
[68/3352] Linking CXX static library contrib/poco/Foundation/libPocoFoundation.a
/Library/Developer/CommandLineTools/usr/bin/ranlib: file: contrib/poco/Foundation/libPocoFoundation.a(ByteOrder.cpp.o) has no symbols
/Library/Developer/CommandLineTool... | 2018-10-03T11:27:15 | 2018-10-03T11:30:21 | {} | NONE |
ClickHouse | ClickHouse | 425,149,397 | 2,373 | Ouriel | Seems duplicate of this one : #2156 | 2018-09-27T16:03:58 | 2018-09-27T16:03:58 | {} | NONE |
ClickHouse | ClickHouse | 425,339,886 | 3,242 | perez1987 | +1 | 2018-09-28T06:52:11 | 2018-09-28T06:52:11 | {} | NONE |
ClickHouse | ClickHouse | 425,347,019 | 3,242 | filimonov | Why making more complicated? It's as easy as:
```
percent = read_rows / total_rows
``` | 2018-09-28T07:26:03 | 2018-09-28T07:26:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,446,042 | 3,244 | filimonov | With higher-order functions.
Try that
```sql
select
number,
bitmaskToArray(number) as bitmask,
arrayReduce('groupBitOr', bitmask) as reverse
from numbers(100);
``` | 2018-09-28T14:04:12 | 2018-09-28T14:04:31 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,459,243 | 3,228 | filimonov | Not sure if i understand you correctly, but if you want to access remote table like to local table, without using credentials all the time you should use corresponding table engine (instead of table functions)
Access to postgres / oracle is possible via odbc.
Samples:
```
CREATE TABLE table_from_mysql (column... | 2018-09-28T14:45:59 | 2018-09-28T14:53:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,465,512 | 3,248 | filimonov | Clickhouse uses re2 engine for regexp matching. It support only some subset of PCRE.
You can find docs about re2 syntax here: https://github.com/google/re2/wiki/Syntax
As you can see lookaround expressions are not supported.
In your case you actually don't need lookbehind / lookahead expressions. Just mark capture g... | 2018-09-28T15:04:18 | 2018-09-28T15:06:38 | {
"+1": 4
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,658,111 | 3,257 | alexey-milovidov | Great, we always wanted it! | 2018-09-29T16:36:37 | 2018-09-29T16:36:37 | {} | MEMBER |
ClickHouse | ClickHouse | 425,800,730 | 3,228 | cmdares | thanks! let me try. | 2018-10-01T06:20:27 | 2018-10-01T06:20:27 | {} | NONE |
ClickHouse | ClickHouse | 424,750,984 | 3,218 | 4ertus2 | #3224 | 2018-09-26T15:05:46 | 2018-09-26T15:05:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 424,783,986 | 3,223 | champtar | I confirm this fix #3144, thanks a lot @veloman-yunkan
@veloman-yunkan are you able to do a test case that reproduce the breakage ? | 2018-09-26T16:32:44 | 2018-09-26T16:32:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 424,837,504 | 3,223 | champtar | Ok so just repeating the same json a lot of time should be enough
I'll try, so we have a public repro case | 2018-09-26T19:17:53 | 2018-09-26T19:17:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,168,245 | 1,777 | sleshJdev | Any updates?
I faced the same issue, the following query works fine if I run it using clickhouse-client, but it fails with a error message: ru.yandex.clickhouse.except.ClickHouseException:
_ru.yandex.clickhouse.except.ClickHouseException: ClickHouse exception, code: 179, host: localhost, port: 7799; Code: 179, e... | 2018-09-27T17:02:21 | 2018-09-27T17:18:55 | {} | NONE |
ClickHouse | ClickHouse | 425,231,102 | 3,239 | blinkov | @RomanGorbatko `if(next, 1, 0)` — next is whole table here, not some valid expression. Also `count(v.Page)` given there's GROUP BY same column doesn't make much sense. | 2018-09-27T20:26:28 | 2018-09-27T20:26:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,230,082 | 1,981 | healiseu | Faulty package installation was corrected by running again the [quick start][1] commands
Then I modified `/etc/clickhouse-server/config.xml` again and
**BINGO**
The clickhouse server started ok and the location of `clickhouse/data` changed to my SSD
There is one thing left. How to bring back the tables from the o... | 2018-09-27T20:22:58 | 2018-09-27T21:36:32 | {} | NONE |
ClickHouse | ClickHouse | 425,460,550 | 3,226 | filimonov | Just for the record: it's better to use `today()` when you need current date instead of `now()` / `toDate(now())`. | 2018-09-28T14:49:57 | 2018-09-28T14:49:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,545,987 | 3,178 | SaltTan | I don't know if it's connected, but log_queries=1 does not work in SETTINGS either | 2018-09-28T19:46:30 | 2018-09-28T19:46:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,663,437 | 3,255 | alexey-milovidov | This is known issue. Client cannot send its patch version in a handshake, because protocol compatibility is required. For that reason, its version is received by server as x.y.0. | 2018-09-29T17:49:12 | 2018-09-29T17:49:53 | {} | MEMBER |
ClickHouse | ClickHouse | 425,664,174 | 2,898 | alexey-milovidov | We are currently merging this PR.
Some difficulties are due to it was forked from old version and not rebased.
We have almost got through it. | 2018-09-29T18:00:34 | 2018-09-29T18:00:34 | {} | MEMBER |
ClickHouse | ClickHouse | 425,691,606 | 3,257 | amosbird | @alexey-milovidov almost done. I don't find anything related to client testing. What should one look like?
> two variants of Russian letter ё:
Can you post a csv file with that letter? I'm not sure if I can generate it correctly. | 2018-09-30T03:32:54 | 2018-09-30T03:32:54 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 424,712,569 | 2,156 | vmanyushin | Same problem with latest stable version 18.12.14, 18.12.17
steps to reproduce:
```
CREATE DATABASE tmp ON CLUSTER cluster_r4s2;
CREATE TABLE tmp.t1_local ON CLUSTER cluster_r4s2
(
event_date Date DEFAULT toDate(event_time),
event_time DateTime,
log_type UInt32,
account_id String
)
ENGINE ... | 2018-09-26T13:21:26 | 2018-09-26T13:27:31 | {} | NONE |
ClickHouse | ClickHouse | 425,347,874 | 3,242 | alex-zaitsev | If it is used from some JS, for example, it may be complicated, since JS does not support Int64, while total rows may easily exceed that. | 2018-09-28T07:29:29 | 2018-09-28T07:29:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,473,500 | 3,226 | champtar | Thanks @filimonov, and thanks @alexey-milovidov for the fix | 2018-09-28T15:29:17 | 2018-09-28T15:29:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,843,509 | 3,210 | alex-krash | Looks like I've finished with this PR.
1. A new API endpoint /identifier_quote has been added to odbc bridge. Now queries are escaped with appropriate character, and should work for MySQL ODBC (when segfault there disappears :) )
2. A support for Nullable was not implemented due to conflict with ODBCBlockInputStream ... | 2018-10-01T09:26:16 | 2018-10-01T09:26:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,867,417 | 3,259 | filimonov | Also: CollapsingMergeTree / ReplacingMergeTree engines allow implementing some kind of updates by overwriting row with the newer version (usage of those Engines have some important limitations). | 2018-10-01T10:57:52 | 2018-10-01T10:57:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,885,590 | 3,260 | quarrazzella | > Can you check if /var/lib/clickhouse/metadata/mobile//events_773292_20180926.sql is accessible in docker, and contain proper ATTACH TABLE statement?
Thanks for the reply. I wish I knew how to do that? Any direct instructions? | 2018-10-01T12:14:31 | 2018-10-01T12:14:31 | {} | NONE |
ClickHouse | ClickHouse | 426,570,921 | 3,277 | 4ertus2 | Можете модифицировать исходный пример из доки? Хочется понять, что сейчас и что Вы ожидаете после доработки.
```
CREATE TABLE sum_map(
date Date,
timeslot DateTime,
statusMap Nested(
status UInt16,
requests UInt64
)
) ENGINE = Log;
INSERT INTO sum_map VALUES
('2000-01-01... | 2018-10-03T09:30:37 | 2018-10-03T09:30:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 426,605,513 | 3,279 | MasterGroosha | @proller Thank you for quoting comment from config file, however this is not what I'm asking.
I'll try again:
**Previous** configuration:
```xml
<interserver_http_port>9009</interserver_http_port>
<interserver_http_host>192.168.2.92</interserver_http_host>
<tcp_port>9000</tcp_port>
<listen_host>192.168.2.92</l... | 2018-10-03T11:40:32 | 2018-10-03T11:41:18 | {} | NONE |
ClickHouse | ClickHouse | 424,835,106 | 3,223 | veloman-yunkan | @champtar Do you mean a small test case? The bug occurs when the colon delimiter is on the buffer boundary. The probability of this happening increases with the size of the data. Of course it is possible to carefully craft a regression test that places the colon on the known offset matching the used buffer size, howeve... | 2018-09-26T19:09:50 | 2018-09-26T19:16:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,223,280 | 3,239 | RomanGorbatko | @blinkov
sounds nice :)
I wrote next query:
```sql
select
count(v.Page) as current,
v.Page,
if(next, 1, 0)
from
test as v
any left join test as next on next.PrevPage = v.Page
group by v.Page
```
Looks like the best variation.
But I have column name error as u say: `Unknown id... | 2018-09-27T19:59:00 | 2018-09-27T19:59:00 | {} | NONE |
ClickHouse | ClickHouse | 425,573,409 | 3,223 | champtar | Ok here some simple repro steps:
```
CREATE TABLE test_json_parse (aaa String, bbb String) ENGINE = Memory;
for n in {1..1000000}; do echo '{"aaa":"aaa","bbb":"bbb"}'; done | curl 'http://127.0.0.1:8123/?query=INSERT%20INTO%20test_json_parse%20FORMAT%20JSONEachRow' -0 -v --data-binary @-
```
Without the patch
... | 2018-09-28T21:38:05 | 2018-09-28T21:38:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,770,372 | 3,258 | alexey-milovidov | ```
FAIL test_random_inserts/test.py::test_random_inserts
FAIL test_random_inserts/test.py::test_insert_multithreaded
PASSED test_delayed_replica_failover/test.py::test
PASSED test_dictionaries/test.py::test_select_all[clickhouse_hashed]
PASSED test_dictionaries/test.py::test_select_all[clickhouse_flat]
PASSED te... | 2018-10-01T01:49:06 | 2018-10-01T01:50:02 | {} | MEMBER |
ClickHouse | ClickHouse | 424,852,959 | 3,088 | alexey-milovidov | #3226
#3229 | 2018-09-26T20:10:06 | 2018-09-26T20:10:06 | {} | MEMBER |
ClickHouse | ClickHouse | 424,948,031 | 3,225 | zhang2014 | 👍 Thank you : )
It is a minimal example reproducing the problem, I will try to fix :
``` SQL
SELECT toString(value) AS value FROM (SELECT 1 AS value) WHERE value = '1';
```
BTW, This error can be circumvented this way:
``` SQL
SELECT id, round(total_tmp, 1) AS total FROM (
SELECT id, sum(amount) AS tota... | 2018-09-27T03:37:38 | 2018-09-27T05:06:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,329,518 | 2,788 | millin | But this issue still happens on each server restart. I'm on latest stable v18.12.17. | 2018-09-28T05:56:50 | 2018-09-28T05:56:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,663,881 | 3,250 | alexey-milovidov | What table engine do you use?
This issue should not happed if you use `MergeTree` family of table engines. Appends are atomic and table should be always in physically consistent state; moreover, it will do automatic repair if data is rendered broken in filesystem after restart.
But it may be considered normal if ... | 2018-09-29T17:56:57 | 2018-09-29T17:57:22 | {} | MEMBER |
ClickHouse | ClickHouse | 425,713,650 | 3,258 | zhang2014 | > Integration tests — fail:2, passed:73, error:0
I can't open https://nda.ya.ru/3UVxnP
Can I see more details of the error ? | 2018-09-30T11:22:01 | 2018-09-30T11:22:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 424,798,233 | 3,212 | champtar | I've reproduced the wrong results in #3227 | 2018-09-26T17:15:21 | 2018-09-26T17:15:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 424,847,175 | 3,208 | alexey-milovidov | > I could add it as a setting, but I'm not sure what would be a use case for that. Do you have a preference @ztlpn @alexey-milovidov ?
Better to add a setting just to allow switch to previous behaviour in some edge cases. | 2018-09-26T19:51:11 | 2018-09-26T19:51:11 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 425,217,284 | 3,239 | blinkov | Recent ClickHouse releases support ON and do not require subquery on right side, though there are still some known issues with full column names. | 2018-09-27T19:37:48 | 2018-09-27T19:37:48 | {
"heart": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,342,665 | 3,242 | blinkov | Console client basically just receives the same numbers and divides `read_rows` by `total_rows` with couple additional tricks.
See https://github.com/yandex/ClickHouse/blob/master/dbms/programs/client/Client.cpp#L1382 | 2018-09-28T07:06:06 | 2018-09-28T07:06:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,366,242 | 3,242 | alex-zaitsev | There is always a workaround. This is a usability question. | 2018-09-28T08:43:05 | 2018-09-28T08:43:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 425,484,337 | 3,196 | tkroman | Also note that there is a regression wrt `ON` vs using:
```
⇒ curl localhost:8123\?distributed_product_mode=global -d "select d from dstr all inner join (select d from dstr) using d"
```
**^^^ Returns OK**
```
⇒ curl localhost:8123\?distributed_product_mode=global -d "select d from dstr as l all inner join ... | 2018-09-28T16:04:12 | 2018-09-28T16:04:12 | {} | NONE |
ClickHouse | ClickHouse | 425,888,578 | 3,167 | simPod | Looks like it's converted internally:
```
:) SELECT *
:-] FROM test
:-] WHERE created_at BETWEEN toStartOfFiveMinute('2018-09-01 00:00:00') AND now()
SELECT *
FROM test
WHERE (created_at >= toStartOfFiveMinute('2018-09-01 00:00:00')) AND (created_at <= now())
```
Thanks for help! | 2018-10-01T12:25:24 | 2018-10-01T12:25:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 426,603,092 | 3,276 | proller | it still tries to build contrib/llvm.
please delete CMakeCache.txt ,
run ```cmake -DENABLE_EMBEDDED_COMPILER=0 -DUSE_INTERNAL_LLVM_LIBRARY=0```
and show full console output | 2018-10-03T11:31:44 | 2018-10-03T11:31:44 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.