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 | 492,515,742 | 5,276 | filimonov | ? | 2019-05-15T06:00:46 | 2019-05-15T06:00:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,571,563 | 1,394 | blinkov | @tangyong that item is not about replacing native storage with S3, it's more about import/export and on the fly processing. Actually this is already partially possible with [URL table engine](https://clickhouse.yandex/docs/en/operations/table_engines/url/) or table function, but it lacks authentication support so it wo... | 2019-05-15T09:06:32 | 2019-05-15T09:06:32 | {
"rocket": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,683,209 | 5,165 | abyss7 | Consumed by #5286 | 2019-05-15T14:38:55 | 2019-05-15T14:38:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,836,436 | 4,542 | alexey-milovidov | We have concerns about the names of these functions. Expect renaming... | 2019-05-15T21:48:50 | 2019-05-15T21:48:50 | {} | MEMBER |
ClickHouse | ClickHouse | 492,862,870 | 5,294 | kvinty | Run performance tests
Need to fix floats comparison with nan_direction_hint | 2019-05-15T23:40:16 | 2019-05-15T23:40:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,068,879 | 5,295 | filimonov | Rather not related to the mentioned issue.
Can you repeat the query with from clickhouse-client with `set send_logs_level='trace'`? It can give some extra hints. | 2019-05-16T13:32:18 | 2019-05-16T13:32:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,213,040 | 4,525 | alexey-milovidov | @proller Do you want to continue with this task or "won't fix"? | 2019-05-16T20:05:57 | 2019-05-16T20:05:57 | {} | MEMBER |
ClickHouse | ClickHouse | 492,365,509 | 5,268 | alexey-milovidov | Description for the changelog does not follow the form. | 2019-05-14T18:57:19 | 2019-05-14T18:57:19 | {} | MEMBER |
ClickHouse | ClickHouse | 492,476,368 | 5,264 | kindred77 | version:19.8.1 | 2019-05-15T02:09:10 | 2019-05-15T02:09:10 | {} | NONE |
ClickHouse | ClickHouse | 492,632,282 | 4,308 | abyss7 | From v19.8+ the shutdown and `DROP TABLE` should wait only for `stream_flush_interval_ms` time - is it fair enough? | 2019-05-15T12:24:15 | 2019-05-15T12:24:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,166,283 | 5,306 | dlesnikov | > 1. kubernetes?
No, Ubuntu 18.04 on bare metal.
> 2. try with this data:
>
> ```
> insert into default.datastore2
insert into **default.datastore**,
not datastore2 | 2019-05-16T17:46:26 | 2019-05-16T17:46:26 | {} | NONE |
ClickHouse | ClickHouse | 493,203,224 | 5,147 | yctn | cause a different group name will not pull All the data. it will just start now. and now get the entire topic. | 2019-05-16T19:35:21 | 2019-05-16T19:35:21 | {} | NONE |
ClickHouse | ClickHouse | 493,216,767 | 2,592 | alexey-milovidov | FYI. We still consider implementing this feature. | 2019-05-16T20:18:07 | 2019-05-16T20:18:07 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 492,374,834 | 5,267 | alexey-milovidov | > Perf says most of time we spend in ColumnVector<*>::replicate() on page faults.
Probably because the block became large enough and allocator stopped to cache memory chunks (started to mmap/munmap). | 2019-05-14T19:22:50 | 2019-05-14T19:22:50 | {} | MEMBER |
ClickHouse | ClickHouse | 492,535,940 | 5,277 | Chr1sxu | cheer 4 this，it helps more or less，but hope Uint64 would be supported soon | 2019-05-15T07:21:52 | 2019-05-15T07:21:52 | {} | NONE |
ClickHouse | ClickHouse | 492,715,822 | 5,177 | abyss7 | As a quick workaround you can use `ALIAS` column instead. | 2019-05-15T15:58:26 | 2019-05-15T15:58:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,114,177 | 5,308 | abyss7 | Fixes #5100 | 2019-05-16T15:24:18 | 2019-05-16T15:24:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,417,179 | 2,205 | blinkov | on recent releases all mentioned examples just always overflow | 2019-05-14T21:30:03 | 2019-05-14T21:30:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,627,279 | 3,997 | mcarbonneaux | i've done quasi the same as https://github.com/yandex/ClickHouse/pull/5253.
```
diff --git a/cmake/find_rdkafka.cmake b/cmake/find_rdkafka.cmake
index de34f37..206aa21 100644
--- a/cmake/find_rdkafka.cmake
+++ b/cmake/find_rdkafka.cmake
@@ -15,14 +15,12 @@ endif ()
if (NOT USE_INTERNAL_RDKAFKA_LIBRARY)
... | 2019-05-15T12:08:18 | 2019-05-15T12:08:18 | {} | NONE |
ClickHouse | ClickHouse | 492,780,136 | 4,889 | 4ertus2 | #5291 | 2019-05-15T18:54:46 | 2019-05-15T18:54:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,101,127 | 5,305 | lzy305 | i run in clickhouse-client.
[master] 2019.05.16 22:51:44.834426 [ 73 ] <Trace> Aggregator: Written part in 0.819 sec., 146174 rows, 90.803 MiB uncompressed, 14.800 MiB compressed, 651.374 uncompressed bytes per row, 106.170 compressed bytes per row, compression rate: 6.135 (178483.931 rows/sec., 110.874 MiB/sec. u... | 2019-05-16T14:52:40 | 2019-05-16T14:56:32 | {} | NONE |
ClickHouse | ClickHouse | 493,156,416 | 5,310 | alexey-milovidov | How to implement:
`KeyCondition.cpp`, look at `atom_map` and its usage. | 2019-05-16T17:18:43 | 2019-05-16T17:18:54 | {} | MEMBER |
ClickHouse | ClickHouse | 493,181,329 | 5,309 | alexey-milovidov | Note that many DBMS don't support LIMIT clause.
Notable examples are MS SQL Server (that has `SELECT TOP N` instead) and Oracle.
We must check if there is a method in ODBC interface to ensure the support for LIMIT.
And the SQL standard does not have LIMIT.
https://stackoverflow.com/questions/1528604/how-unive... | 2019-05-16T18:29:06 | 2019-05-16T18:33:09 | {} | MEMBER |
ClickHouse | ClickHouse | 492,417,327 | 2,199 | blinkov | @fahrettinf89 do you have any further questions? | 2019-05-14T21:30:30 | 2019-05-14T21:30:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,537,772 | 5,196 | blinkov | @zhang2014 as you wish, that was just a suggestion | 2019-05-15T07:27:47 | 2019-05-15T07:27:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,068,228 | 5,293 | filimonov | Sounds reasonable. | 2019-05-16T13:30:18 | 2019-05-16T13:30:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,094,734 | 5,295 | jameshartig | ```
velda :) set send_logs_level='trace'
SET send_logs_level = 'trace'
Ok.
0 rows in set. Elapsed: 0.001 sec.
velda :) CREATE DATABASE IF NOT EXISTS visitor_api ON CLUSTER '{cluster}';
CREATE DATABASE IF NOT EXISTS visitor_api ON CLUSTER `{cluster}`
↘ Progress: 0.00 rows, 0.00 B (0.00 rows/s., 0.00 B... | 2019-05-16T14:36:56 | 2019-05-16T14:37:18 | {} | NONE |
ClickHouse | ClickHouse | 493,102,516 | 5,306 | den-crane | 1. kubernetes?
2. try with this data:
```
insert into default.datastore2
select number, today()+number%103, toFloat64(cityHash64(number)/13244354311.0),
number%13, number%17, number%9991, toString(number)
from numbers(500000000);
```
I could not reproduce with 19.6.2 debian_version 9.2
insert i... | 2019-05-16T14:56:04 | 2019-05-16T16:18:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,168,569 | 5,306 | dlesnikov | Tomorrow I will attach the ClickHouse configuration files. | 2019-05-16T17:52:41 | 2019-05-16T17:52:41 | {} | NONE |
ClickHouse | ClickHouse | 492,418,682 | 5,196 | blinkov | @zhang2014 maybe we should merge at least partial result? | 2019-05-14T21:34:50 | 2019-05-14T21:34:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,683,716 | 2,691 | abyss7 | Consumed by #5286 | 2019-05-15T14:40:05 | 2019-05-15T14:40:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,946,676 | 5,279 | andyyzh | From the clickhouse-server.txt, Crash when this use case is executed.
> 2019.05.16 01:09:04.003917 [ 52 ] {38de2a3c-cbd9-49b7-a055-7486559955ea} <Debug> executeQuery: (from [::1]:51410) select bitmapHasAny(bitmapBuild([ 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33, ... | 2019-05-16T07:10:49 | 2019-05-16T07:10:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,075,794 | 5,305 | filimonov | 1) Can you try to run clickhouse-client, then run `set send_logs_level='trace'`, and after - run that problematic query? and attach the results here.
2) Are you able to create a reproducible example? It would make diagnosing and fixing much simpler | 2019-05-16T13:50:52 | 2019-05-16T13:50:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,476,614 | 5,256 | kindred77 | thanks very much, I should read the doc more carefully. | 2019-05-15T02:10:42 | 2019-05-15T02:10:42 | {} | NONE |
ClickHouse | ClickHouse | 492,634,918 | 5,242 | 4ertus2 | #5282 | 2019-05-15T12:32:42 | 2019-05-15T12:32:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,911,558 | 3,925 | vzakaznikov | I am working on a variant of this PR to meet the "not intrusive" requirement
so that BlockInfo and Datastreams do not have to be modified.
Unfortunately, we can't keep the full original functionality without being intrusive.
However, I believe we can get most of the functionality with some modifications which
can b... | 2019-05-16T04:25:41 | 2019-05-16T04:25:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,069,386 | 5,213 | abyss7 | You can use `ALIAS` default expression as a workaround for now. | 2019-05-16T13:33:42 | 2019-05-16T13:33:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,103,039 | 5,301 | alexey-milovidov | Assigned to @proller for reviewing `cotire`. | 2019-05-16T14:57:21 | 2019-05-16T14:57:29 | {} | MEMBER |
ClickHouse | ClickHouse | 493,122,868 | 3,925 | alex-zaitsev | @javisantana That was also my idea while ago to notify of changes. But looks more complicated comparing to select and wait for the data to appear
| 2019-05-16T15:46:21 | 2019-05-16T15:46:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,547,814 | 5,263 | YiuRULE | @alexey-milovidov Ah thanks ! Indeed some library packages concerning `llvm` were missing on my end, thanks :)
I will do a PR soon, just the xml file without the custom function right ?
@filimonov It was with the latest master branch from the repository compiled on my end in an Ubuntu 18.10 as I needed to create cu... | 2019-05-15T07:54:51 | 2019-05-15T07:54:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,616,815 | 2,551 | abyss7 | Fixed in v19.8+ | 2019-05-15T11:30:53 | 2019-05-15T11:30:53 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,690,450 | 5,208 | abyss7 | There was a problem with Kafka messages get lost even without restarts. It was fixed in 19.5 and 19.6 (you haven't provided the version you're using). Can you test against latest 19.5.* or 19.6.* versions and confirm that problem persists? | 2019-05-15T14:55:53 | 2019-05-15T14:55:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,773,810 | 5,286 | filimonov | Maybe also partition and offset? | 2019-05-15T18:37:26 | 2019-05-15T18:37:26 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,104,802 | 5,293 | alexey-milovidov | How to implement:
- add `LOG_NOTICE` to `logger_useful.h`;
- replace `LOG_INFO` to `LOG_NOTICE` in the appropriate places related to server startup. | 2019-05-16T15:01:36 | 2019-05-16T15:01:36 | {} | MEMBER |
ClickHouse | ClickHouse | 493,207,079 | 4,861 | alexey-milovidov | Where to look in code:
`DateLUTImpl.h`, `makeDayNum` and `makeDateTime` functions.
We can throw an exception for definitely invalid dates/datetimes, but we still accept something like February 31th; 25th hour; 99th minute, etc. This is documented behaviour. Introducing better checks will require careful performance... | 2019-05-16T19:47:31 | 2019-05-16T19:47:31 | {} | MEMBER |
ClickHouse | ClickHouse | 493,226,286 | 3,540 | alexey-milovidov | Confirmed. | 2019-05-16T20:48:05 | 2019-05-16T20:48:05 | {} | MEMBER |
ClickHouse | ClickHouse | 492,347,503 | 5,165 | kmatt | @abyss7 We do use a list to connect a single Kafka table to multiple topics and brokers, but what I am looking for the ability to insert the topic name from that list with each Kafka message.
This would give us the ability to use a single consumer and query messages by topic, without having to create separate Kafka ... | 2019-05-14T18:09:00 | 2019-05-14T18:09:00 | {} | NONE |
ClickHouse | ClickHouse | 492,386,116 | 4,970 | abraithwaite | Following up on this: are the replication connection pools recycled when calling `system flush dns cache`?
https://github.com/yandex/ClickHouse/blob/8814e0cda578806d2f159f2d0cbcdae11a2d6744/dbms/src/IO/HTTPCommon.cpp#L96-L166 | 2019-05-14T19:56:28 | 2019-05-14T19:56:28 | {} | NONE |
ClickHouse | ClickHouse | 492,386,210 | 3,882 | abraithwaite | Maybe related?
https://github.com/yandex/ClickHouse/issues/4970#issuecomment-492386116 | 2019-05-14T19:56:46 | 2019-05-14T19:56:46 | {} | NONE |
ClickHouse | ClickHouse | 492,664,355 | 5,221 | 4ertus2 | #5284 | 2019-05-15T13:53:43 | 2019-05-15T13:53:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,678,025 | 5,275 | abraithwaite | Another replica which hasn't come online yet. We'll be making that entrypoint script but I'm still a bit surprised that this failure would prevent startup. | 2019-05-15T14:27:07 | 2019-05-15T14:27:07 | {} | NONE |
ClickHouse | ClickHouse | 493,133,830 | 4,631 | abyss7 | Please, increase the verbosity of rdkafka in server config:
```
<kafka>
<debug>all</debug>
</kafka>
```
And provide me logs when the issue happens. So I can understand what is really happening from the client point of view. | 2019-05-16T16:14:01 | 2019-05-16T16:14:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,167,446 | 5,306 | den-crane | >insert into default.datastore,
>not datastore2
OKKK, try to fill datastore2 ( or other table with the same DDL ) with this data:
```
insert into default.datastore2
select number, today()+number%103, toFloat64(cityHash64(number)/13244354311.0),
number%13, number%17, number%9991, toString(number)
fro... | 2019-05-16T17:49:31 | 2019-05-16T17:50:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,212,571 | 4,585 | alexey-milovidov | Could you please verify that this has been fixed? | 2019-05-16T20:04:30 | 2019-05-16T20:04:30 | {} | MEMBER |
ClickHouse | ClickHouse | 493,213,708 | 4,114 | alexey-milovidov | Fixed in https://github.com/yandex/ClickHouse/pull/5238 | 2019-05-16T20:08:18 | 2019-05-16T20:08:18 | {} | MEMBER |
ClickHouse | ClickHouse | 492,347,479 | 5,253 | abyss7 | > Is it possible to add a test?
Will try to test at least manually | 2019-05-14T18:08:56 | 2019-05-14T18:08:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,576,498 | 4,518 | BayoNet | I have found the information about SYSTEM FLUSH LOGS in the system.query_log description https://clickhouse.yandex/docs/en/operations/system_tables/#system_tables-query-log. | 2019-05-15T09:21:04 | 2019-05-15T09:21:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,611,112 | 5,263 | alexey-milovidov | > I will do a PR soon, just the xml file without the custom function right ?
Yes. | 2019-05-15T11:11:38 | 2019-05-15T11:11:38 | {} | MEMBER |
ClickHouse | ClickHouse | 492,725,559 | 5,128 | abyss7 | About logging. In recent versions the Kafka debug level is set to nothing by default - and even if you use older version, you can setup the "kafka" section inside the server config xml. | 2019-05-15T16:24:53 | 2019-05-15T16:24:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,957,150 | 3,925 | javisantana | @alexey-milovidov wouldn't a mat view with a table using URL engine would be useful to notify changes, like:
```
:) create table test_maser (a Int32) Engine=MergeTree order by tuple();
:) create table test_url (a Int32) Engine = URL('http://localhost:8080', 'CSV');
# i know it would need an aggregatingmergetree b... | 2019-05-16T07:45:20 | 2019-05-16T07:45:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,214,455 | 4,070 | alexey-milovidov | PS. You can also do it for PostgreSQL: https://github.com/mkabilov/pg2ch | 2019-05-16T20:10:41 | 2019-05-16T20:10:53 | {} | MEMBER |
ClickHouse | ClickHouse | 492,360,426 | 5,259 | daledude | Thanks @den-crane. I was under the assumption that the `<id>` and `<key>` tags were different than the `<attribute>`. I removed the `<attribute>` entries where the field was also in the `<id>` or `<key>`. | 2019-05-14T18:43:17 | 2019-05-14T18:43:17 | {} | NONE |
ClickHouse | ClickHouse | 492,910,833 | 4,026 | nupursinghbaghel | Hi @wawanawna,
What is the workaround you found ? | 2019-05-16T04:20:50 | 2019-05-16T04:20:50 | {} | NONE |
ClickHouse | ClickHouse | 493,111,408 | 5,243 | abyss7 | > MV never supported select from select. AFAIK no plans to support them.
That's correct - I don't see any other Kafka-related problem here, so I close the issue. You can open new issues regarding MV behavior only. | 2019-05-16T15:17:32 | 2019-05-16T15:17:32 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,193,412 | 5,309 | Schaefer-Thomas | Wow, exhaustive answer.
Thank you, Alexey! :1st_place_medal:
p.s.
think I should do more research before posting here | 2019-05-16T19:04:53 | 2019-05-16T19:04:53 | {} | NONE |
ClickHouse | ClickHouse | 492,624,763 | 5,264 | filimonov | Are the replicas & hostname are literally matches the addresses provided in cluster configs (`<remote_servers>` section in config.xml)?
Can you show full stracktrace (available in the log, or when you run `clickhouse-client --stacktrace`)? | 2019-05-15T12:00:09 | 2019-05-15T12:00:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,626,042 | 2,551 | mcarbonneaux | ok they look good !
and i see that they use libgsasl !
i've done quasi the same with libsasl from cyrus project... | 2019-05-15T12:04:16 | 2019-05-15T12:04:16 | {} | NONE |
ClickHouse | ClickHouse | 492,832,254 | 4,470 | den-crane | The same problem with official 19.4.3.11 & debian 10 (buster)
cat /etc/debian_version
10.0
openssl 1.1.1b-2
| 2019-05-15T21:34:24 | 2019-05-15T21:34:24 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,096,158 | 5,305 | proller | Please show full error (maybe from server logs)
Do you have database dw_hr ?
| 2019-05-16T14:40:28 | 2019-05-16T14:40:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,134,802 | 4,026 | den-crane | @nupursinghbaghel it could be https://github.com/yandex/ClickHouse/issues/5274
Then workaround is to add some column to MV to break de-duplication
as select , now() as workaround -- or generateUUIDv4()
and collapse it using any
| 2019-05-16T16:16:45 | 2019-05-16T16:16:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,150,867 | 5,309 | alexey-milovidov | Implementation:
`transformQueryForExternalDatabase.h`
`transformQueryForExternalDatabase.cpp` | 2019-05-16T17:02:39 | 2019-05-16T17:02:39 | {} | MEMBER |
ClickHouse | ClickHouse | 493,137,366 | 5,274 | den-crane | CREATE TABLE testAggregateReplicate
(date Date, mcount AggregateFunction(count), **workaround DateTime**)
ENGINE = ReplicatedAggregatingMergeTree('/clickhouse/testCH/tables/testAggregateReplicate','1') order by date;
CREATE MATERIALIZED VIEW testAggregateReplicate_mv to testAggregateReplicate
AS SELECT date, co... | 2019-05-16T16:24:06 | 2019-05-16T17:16:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,220,312 | 2,384 | alexey-milovidov | Проблема со вторым запросом может быть исправлена в https://github.com/yandex/ClickHouse/pull/5313
Но мне нужен более подробный пример, чтобы воспроизвести проблему. | 2019-05-16T20:29:06 | 2019-05-16T20:29:06 | {} | MEMBER |
ClickHouse | ClickHouse | 492,477,002 | 5,196 | zhang2014 | @blinkov
There are very few parts that are not completed. maybe it is better to wait for this part to be completed before merging? And I need to review the documents at least once. | 2019-05-15T02:13:18 | 2019-05-15T02:13:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,514,935 | 5,275 | filimonov | Probably you can try to use kubernetes Services, or just add some dns-ready waiting on entrypoint of clickhouse container before starting of clickhouse-server itself as a workaround.
In the exception it can't resolve own name, or name of some other replica? | 2019-05-15T05:57:09 | 2019-05-15T06:07:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,521,275 | 5,187 | suxw8813 | Does the Windowfunnel function support datetime64? | 2019-05-15T06:25:56 | 2019-05-15T06:25:56 | {} | NONE |
ClickHouse | ClickHouse | 492,524,006 | 5,277 | ghost | As a workaround, you can try to use `UInt32` with number of fractions of second since specific date. For example, you can store high-precision timestamp as the number of milliseconds since UNIX epoch, then call `windowFunnel` as
```
windowFunnel(600000)(toUInt32(timestampMs - firstTimestampMs), cond1, cond2, ...)
``... | 2019-05-15T06:37:13 | 2019-05-15T06:57:31 | {} | NONE |
ClickHouse | ClickHouse | 493,153,164 | 5,309 | alexey-milovidov | The LIMIT clause can be sent to the remote database only in very limited cases.
When the query does not have:
- any incompatible functions in WHERE;
- GROUP BY;
- DISTINCT;
- ORDER BY;
- HAVING;
- ARRAY JOIN or arrayJoin function;
- JOIN;
- LIMIT BY;
- WITH TIES / WITH FILL (upcoming feature).
So, if it ... | 2019-05-16T17:09:42 | 2019-05-16T17:15:33 | {} | MEMBER |
ClickHouse | ClickHouse | 493,224,953 | 3,485 | alexey-milovidov | > to implement a ADD COLUMN for multiple columns
ADD COLUMN *is implemented* for multiple columns from the beginning. You can execute multiple operations on columns in a single ALTER statement:
```
ALTER TABLE t ADD COLUMN a UInt8, ADD COLUMN b UInt16, DROP COLUMN c, MODIFY COLUMN d String, ...
```
> It woul... | 2019-05-16T20:43:45 | 2019-05-16T20:43:45 | {
"heart": 1
} | MEMBER |
ClickHouse | ClickHouse | 492,416,433 | 2,214 | blinkov | @27260102 it is recommended to write data to shards directly instead of via Distributed table. | 2019-05-14T21:27:34 | 2019-05-14T21:27:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,584,300 | 847 | filimonov | Documentation is added and available on the official website:
https://clickhouse.yandex/docs/en/operations/system_tables/#system_tables-query-log | 2019-05-15T09:43:32 | 2019-05-15T09:43:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,668,508 | 5,285 | abyss7 | Fixes #5085 | 2019-05-15T14:03:44 | 2019-05-15T14:03:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,071,609 | 5,301 | filimonov | ClickHouse uses a lot of C++ templates. Also, it's has a lot of builtin dependencies. So it's quite a hard task for the compiler & linker.
Do you use ccache?
Maybe you already tried adding cotire (it looks like it should be simple)? That request will be more convincing if you will show results 'before' and 'aft... | 2019-05-16T13:39:55 | 2019-05-16T13:59:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,221,819 | 1,779 | alexey-milovidov | @filimonov Does it reproduce in recent versions? | 2019-05-16T20:33:40 | 2019-05-16T20:33:40 | {} | MEMBER |
ClickHouse | ClickHouse | 492,305,243 | 5,259 | den-crane | ```
<id>
<name>id</name>
</id>
<attribute>
<name>id</name>
....
</attribute>
```
Why do you id twice?
If you really need it just rename
```
<id>
<name>_id</name>
<expression>id</expression>
</id>
<attribute>
<name>id</name>
....
</att... | 2019-05-14T16:12:40 | 2019-05-14T17:52:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,476,970 | 5,250 | kindred77 | thanks, I have fixed with CMAKE_EXE_LINKER_FLAGS "-static-libgcc -static-libstdc++"
| 2019-05-15T02:13:06 | 2019-05-15T02:13:06 | {} | NONE |
ClickHouse | ClickHouse | 492,558,615 | 1,394 | tangyong | AFAIK: from roadmap, 2019 Q2 seems to support S3-like object storage, right? | 2019-05-15T08:27:59 | 2019-05-15T08:27:59 | {} | NONE |
ClickHouse | ClickHouse | 493,056,686 | 3,730 | 4ertus2 | #5304 | 2019-05-16T12:58:18 | 2019-05-16T12:58:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,116,488 | 4,658 | ekbfh | https://github.com/yandex/ClickHouse/pull/4718 was merged
In release: https://github.com/yandex/ClickHouse/blob/master/CHANGELOG.md#clickhouse-release-19526-2019-04-15
| 2019-05-16T15:29:58 | 2019-05-16T15:29:58 | {} | NONE |
ClickHouse | ClickHouse | 493,128,957 | 3,925 | javisantana | totally, it's a poor mans version of that | 2019-05-16T16:01:05 | 2019-05-16T16:01:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 493,130,642 | 5,307 | alexey-milovidov | No changes in docs. | 2019-05-16T16:05:36 | 2019-05-16T16:05:36 | {} | MEMBER |
ClickHouse | ClickHouse | 493,207,496 | 5,312 | den-crane | workaround https://t.me/clickhouse_ru/81231 | 2019-05-16T19:48:52 | 2019-05-16T19:48:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,351,145 | 5,165 | abyss7 | Now I understood. Will reconsider, how this can be implemented if possible. | 2019-05-14T18:18:49 | 2019-05-14T18:18:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,495,985 | 4,659 | pastatem | :+1: | 2019-05-15T04:10:57 | 2019-05-15T04:10:57 | {} | NONE |
ClickHouse | ClickHouse | 492,531,841 | 4,970 | andrei-dascalu | @abraithwaite hi, unfortunately no, `system flush dns cache` didn't help in the original issue. | 2019-05-15T07:07:11 | 2019-05-15T07:07:11 | {} | NONE |
ClickHouse | ClickHouse | 492,583,550 | 847 | filimonov | > can I safely truncate `query_log` table?
Yes, it's 100% safe. | 2019-05-15T09:41:26 | 2019-05-15T09:41:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 492,952,495 | 5,014 | github1youlc | I figure out the problem in my case, and I make a fix here. [9527](https://github.com/yandex/ClickHouse/pull/5297/) | 2019-05-16T07:30:41 | 2019-05-16T07:30:41 | {} | NONE |
ClickHouse | ClickHouse | 493,182,376 | 5,309 | alexey-milovidov | I cannot find the corresponding method in ODBC interface:
https://docs.microsoft.com/en-us/sql/odbc/reference/syntax/sqlgetinfo-function | 2019-05-16T18:32:03 | 2019-05-16T18:32:03 | {} | MEMBER |
ClickHouse | ClickHouse | 493,222,058 | 54 | alexey-milovidov | @BayoNet ??? | 2019-05-16T20:34:26 | 2019-05-16T20:34:26 | {} | MEMBER |
ClickHouse | ClickHouse | 493,222,997 | 3,893 | alexey-milovidov | Is the issue still relevant? Did you ask DBeaver developers about it? | 2019-05-16T20:37:23 | 2019-05-16T20:37:23 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.