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 | 624,244,245 | 7,006 | alexey-milovidov | > At last I found that when the table has many records in system.replication_queue it will not works.
Yes, but these records are related to Replicated tables (and indicate that the Replicated tables may have some issues with replication) not to Distibuted table (that is actually just a view on top of your cluster). | 2020-05-05T18:58:52 | 2020-05-05T18:58:52 | {} | MEMBER |
ClickHouse | ClickHouse | 624,472,554 | 10,680 | 4ertus2 | Do you have a sample file? Columns of which type are missed? | 2020-05-06T06:47:19 | 2020-05-06T06:47:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,832,039 | 10,671 | KochetovNicolai | Have backported to 20.3 | 2020-05-06T19:02:37 | 2020-05-06T19:02:37 | {} | MEMBER |
ClickHouse | ClickHouse | 624,934,072 | 7,091 | den-crane | @silviucpp In what CH version? It was fixed recently https://github.com/ClickHouse/ClickHouse/issues/8505 | 2020-05-06T22:59:24 | 2020-05-06T22:59:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 625,108,303 | 10,672 | alex-zaitsev | Let's add those functions for UInt64 <-> DateTime64 conversion:
toUnixTimeMillies
fromUnixTimeMillies | 2020-05-07T08:25:34 | 2020-05-07T08:25:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,246,816 | 5,604 | alexey-milovidov | @CurtizJ Changing the type to LowCardinality indeed causes column rewrite.
But the `primary.idx` is not modified by this transformation.
It means that now we can support it. | 2020-05-05T19:04:00 | 2020-05-05T19:04:00 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 624,252,031 | 4,670 | alexey-milovidov | > Thanks den, what about to select distinct values of multiple columns? I guess there is no optimization for such case.
You're right, there is no such optimization.
> To my understanding, these low cardinality columns are stored with compressed reference to the raw values.
The details are a bit more complicate... | 2020-05-05T19:14:02 | 2020-05-05T19:14:02 | {} | MEMBER |
ClickHouse | ClickHouse | 624,401,739 | 10,684 | champtar | 00834_cancel_http_readonly_queries_on_client_close failed :( | 2020-05-06T01:48:59 | 2020-05-06T01:48:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,479,037 | 10,663 | qoega | Client failed with `terminating with uncaught exception of type DB::NetException: Connection refused (localhost:9000)` | 2020-05-06T07:05:49 | 2020-05-06T07:05:49 | {} | MEMBER |
ClickHouse | ClickHouse | 624,578,982 | 10,680 | Crazylearner30 | Parquet file generated in hive have the below schema
Create table test
(
id BigInt,
company String,
Email String,
Date1 String,
Date2 String
)
Destination table in clickhouse have the below schema
create table temp
(
id UInt64,
Company String,
Email String,
Date1 Date,
Date2 Date
ENGINE = MergeT... | 2020-05-06T10:53:36 | 2020-05-06T10:53:36 | {} | NONE |
ClickHouse | ClickHouse | 624,556,989 | 9,887 | leoones | When can the with statement return multiple lines ？？
Window functions are not supported (OVER ... PARTITION BY ... ) & analytical functions RANK, FIRST, LAST, LAG ， Do you have a plan ？？ | 2020-05-06T10:04:54 | 2020-05-06T11:07:57 | {} | NONE |
ClickHouse | ClickHouse | 624,586,364 | 9,887 | alex-zaitsev | @leoones , 'WITH' returning multiple lines is hopefully for Q3 and window functions for Q4 earliest. Those are complex projects, and we are looking for contributors and sponsors to get it done.
| 2020-05-06T11:10:51 | 2020-05-06T11:10:51 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,854,807 | 10,534 | alexey-milovidov | Every function must be in its own file. | 2020-05-06T19:51:08 | 2020-05-06T19:51:08 | {} | MEMBER |
ClickHouse | ClickHouse | 624,233,930 | 7,957 | alexey-milovidov | @MahmoudGoda0
The logic is correct. But this scenario will not happen because largest parts appeared after merges and merges of larger parts appeared when parts are old enough and when they are already merged from smaller parts. So, largest parts cannot be the most fresh. | 2020-05-05T18:38:37 | 2020-05-05T18:38:37 | {} | MEMBER |
ClickHouse | ClickHouse | 624,712,235 | 10,690 | XavierQuiniou | Very disappointed with your answer, I draw your attention to the fact that this goes against the interests of Altinity, which is trying to promote the use of CLICKHOUSE among users of TABLEAU.
| 2020-05-06T15:18:02 | 2020-05-06T15:18:02 | {} | NONE |
ClickHouse | ClickHouse | 624,864,826 | 10,679 | filimonov | It should not be possible to remove active replica this way, only stale/inactive. | 2020-05-06T20:12:08 | 2020-05-06T20:12:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 625,079,612 | 2,718 | filimonov | https://github.com/ClickHouse/ClickHouse/pull/10622 | 2020-05-07T07:25:33 | 2020-05-07T07:25:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 625,101,509 | 10,709 | KochetovNicolai | Cool! | 2020-05-07T08:11:26 | 2020-05-07T08:11:26 | {} | MEMBER |
ClickHouse | ClickHouse | 624,247,481 | 5,604 | alexey-milovidov | I assign to @alesapin because he has recently done a big rewrite of ALTER mechanics and now this task looks easy to implement. | 2020-05-05T19:05:16 | 2020-05-05T19:05:16 | {} | MEMBER |
ClickHouse | ClickHouse | 624,263,937 | 6,534 | alexey-milovidov | > I need to update this row_state for a set of random Rows.for that also I used Alter query with row_id IN ( 1,2,3,..etc. unfortunately, I need to update for a large number of rows.which can't be executed by single query so I split it by some chunk of rows and suddenly CH crashed.
can anyone help me!!
The proper so... | 2020-05-05T19:37:27 | 2020-05-05T19:37:27 | {} | MEMBER |
ClickHouse | ClickHouse | 624,659,018 | 10,690 | den-crane | Because CH does not support such joins now.
Actually CH able to support `JOIN ON a <=> b` but it is not implemented https://github.com/ClickHouse/ClickHouse/issues/3889 | 2020-05-06T13:45:58 | 2020-05-06T13:45:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,662,714 | 9,786 | den-crane | I agree that behavior is inconsistent. Probably CH should not support this in the Values as well. | 2020-05-06T13:52:30 | 2020-05-06T13:52:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,731,499 | 10,583 | qoega | Same problem
``` sql
SELECT `WatchID`, `JavaEnable`,`Title`,`GoodEvent`,`EventTime`,(today() - 2 + EventDate % 5) as EventDate,`CounterID`, `ClientIP`, `ClientIP6` ,`RegionID`, `UserID`, `CounterClass`,`OS`,`UserAgent`,`URL`,`Referer`,`URLDomain`,`RefererDomain`,`Refresh`,`IsRobot`,`RefererCategories` ,`URLCategories... | 2020-05-06T15:51:05 | 2020-05-06T15:51:05 | {} | MEMBER |
ClickHouse | ClickHouse | 624,805,521 | 7,246 | alexey-milovidov | Does not reproduce on 19.16+. | 2020-05-06T18:10:24 | 2020-05-06T18:10:24 | {} | MEMBER |
ClickHouse | ClickHouse | 624,836,246 | 10,707 | alexey-milovidov | Checked with `docker run --rm -it centos:6 /bin/bash` | 2020-05-06T19:11:36 | 2020-05-06T19:11:36 | {} | MEMBER |
ClickHouse | ClickHouse | 624,845,610 | 10,684 | alexey-milovidov | Please check how it corresponds with https://github.com/ClickHouse/ClickHouse/pull/10687
Maybe the reference to Poco submodule has to be updated once again? | 2020-05-06T19:31:50 | 2020-05-06T19:31:50 | {} | MEMBER |
ClickHouse | ClickHouse | 624,225,333 | 10,682 | alexey-milovidov | We can recognize numbers and check for symbolic values if the value is not a number.
It can be done even automatically without any settings if the Enum type does not have any symbolic values that can be ambiguously parsed as numbers. | 2020-05-05T18:21:53 | 2020-05-05T18:21:53 | {} | MEMBER |
ClickHouse | ClickHouse | 624,514,612 | 10,590 | 4ertus2 | It's not clear how much memory uses each query. I.e. first one could need 18.5G and second one needs 17.5G. Could ypu create a sample with smaller data set that would successed for both queries and show max memory usage and spent time for them? | 2020-05-06T08:33:35 | 2020-05-06T08:33:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,948,833 | 10,717 | alexey-milovidov | I cannot find `pika` package in Ubuntu repository.
```
$ apt search pika
Sorting... Done
Full Text Search... Done
node-pikaday/eoan,eoan 1.8.0-1 all
Lightweight JavaScript datepicker
python-pika/eoan,eoan 0.11.0-4 all
AMQP client library for Python 2
python-pika-doc/eoan,eoan 0.11.0-4 all
AMQP cli... | 2020-05-06T23:46:42 | 2020-05-06T23:46:42 | {} | MEMBER |
ClickHouse | ClickHouse | 625,051,364 | 7,091 | silviucpp | @den-crane on the master from sept 2019 :) . You are right next time I will put the commit id as well
Silviu | 2020-05-07T06:15:28 | 2020-05-07T06:15:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,286,508 | 10,539 | CurtizJ | In older verison this bug leads to non-usage of primary key.
```
select version()
┌─version()─┐
│ 19.17.3.7 │
└───────────┘
set force_primary_key = 1;
select sum(project in ('val1', 'val2')) from test where project in ('val1', 'val2')
Received exception from server (version 20.5.1):
Code: 277. DB::Exce... | 2020-05-05T20:24:06 | 2020-05-05T20:24:06 | {} | MEMBER |
ClickHouse | ClickHouse | 624,480,004 | 10,674 | 4ertus2 | 20.3.7 already has some related patch https://github.com/ClickHouse/ClickHouse/commit/e8cec7688c60330da9bbe5520cf7c66eac9d44e8
Looks like it's not enough. | 2020-05-06T07:08:34 | 2020-05-06T07:08:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,633,454 | 10,694 | filimonov | Thanks for good report. It somehow relates to where vs prewhere.
With `set optimize_move_to_prewhere=0` the original query works.
Was broken between 19.13 and 19.14. | 2020-05-06T12:58:33 | 2020-05-06T13:09:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 623,655,801 | 10,664 | KochetovNicolai | The idea of the fix: we need to set `in->setNeeded();` for all inputs at first call. Otherwise, inputs will be consecutively read and remote queries will be consecutively executed in case of single level aggregation. | 2020-05-04T19:22:01 | 2020-05-06T18:34:35 | {} | MEMBER |
ClickHouse | ClickHouse | 624,220,669 | 7,778 | alexey-milovidov | @hustnn We already have CPU profiler for ClickHouse queries.
https://clickhouse.tech/docs/en/operations/optimizing-performance/sampling-query-profiler/ | 2020-05-05T18:13:02 | 2020-05-05T18:13:02 | {
"heart": 1
} | MEMBER |
ClickHouse | ClickHouse | 624,221,555 | 7,778 | alexey-milovidov | > For example, it seems impossible to understand how much memory the tables with the Join / Set engine use.
Yes, Join and Set are not instrumented enough. | 2020-05-05T18:14:41 | 2020-05-05T18:14:41 | {} | MEMBER |
ClickHouse | ClickHouse | 624,397,612 | 10,671 | hoseiney | id UInt32
created_at , updated_at DateTime
created_date Date
as you know 8192 is default GRANULARITY, you are saying the default value is not appropriate ? | 2020-05-06T01:30:02 | 2020-05-06T01:30:02 | {} | NONE |
ClickHouse | ClickHouse | 624,522,596 | 6,503 | gubinjie | @lamber-ken
http_send_timeout and http_receive_timeout. Are these two parameters configured in config.xml or users.xml? The unit of the parameter is seconds? | 2020-05-06T08:51:19 | 2020-05-06T08:51:19 | {} | NONE |
ClickHouse | ClickHouse | 624,726,219 | 10,520 | qoega | @astudnev Can you approve changes from Ivan and me? Or add rights to change PR.
And we will merge it | 2020-05-06T15:42:14 | 2020-05-06T15:42:14 | {} | MEMBER |
ClickHouse | ClickHouse | 624,842,277 | 9,990 | alexey-milovidov | Reopen, because the function `extractGroups` is not implemented.
It's pointless to have `extractAllGroups` without `extractGroups` function. | 2020-05-06T19:24:35 | 2020-05-06T19:24:35 | {} | MEMBER |
ClickHouse | ClickHouse | 624,221,210 | 7,778 | alexey-milovidov | > Also: do all memory allocations happen through jemalloc?
No, some allocations goes to mmap directly.
| 2020-05-05T18:14:03 | 2020-05-05T18:14:03 | {} | MEMBER |
ClickHouse | ClickHouse | 624,222,686 | 7,971 | alexey-milovidov | > Can you please confirm that inserting int for enum does not work at all for CSV import?
Yes, Enums are only parsed as their symbolic values - in all text formats like CSV, TSV, JSONEachRow...
But It works for numeric values for Values format because it does automatic implicit type conversions for SQL compatibil... | 2020-05-05T18:16:52 | 2020-05-05T18:16:52 | {} | MEMBER |
ClickHouse | ClickHouse | 624,230,190 | 4,035 | alexey-milovidov | @fancyqlx Yes, the reason is to get deterministic and well defined subset of data.
You can execute a query with SAMPLE 1/2 and another query with SAMPLE 1/2 OFFSET 1/2 and be sure that the subsets of data have no intersections and cover 100% of data together.
Even more, you can execute the same SAMPLE expression ... | 2020-05-05T18:31:21 | 2020-05-05T18:31:40 | {} | MEMBER |
ClickHouse | ClickHouse | 624,358,095 | 10,539 | orloffv | @alexey-milovidov https://nda.ya.ru/t/69aO58Sa3VxkoF | 2020-05-05T23:26:01 | 2020-05-05T23:26:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,472,797 | 10,680 | filimonov | We need details to be able to help you. Schemas on both end for example | 2020-05-06T06:47:57 | 2020-05-06T06:47:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,853,962 | 10,534 | alexey-milovidov | I found that re2 is used instead of re2_st and it is slow. | 2020-05-06T19:49:23 | 2020-05-06T19:49:23 | {} | MEMBER |
ClickHouse | ClickHouse | 624,917,125 | 10,622 | nikitamikhaylov | Could you please provide us a query structure for your performance test? | 2020-05-06T22:08:50 | 2020-05-06T22:08:50 | {} | MEMBER |
ClickHouse | ClickHouse | 624,232,653 | 4,833 | alexey-milovidov | @baldurgaldur The old version 1.1.54394 had some bugs that may relate to this behaviour.
@yuzhichang You should not worry about this issue if it happens only on system tables.
It may happen due to incorrect set of parts in filesystem that may appear after server restart.
The system will continue to work and system... | 2020-05-05T18:36:04 | 2020-05-05T18:36:04 | {} | MEMBER |
ClickHouse | ClickHouse | 624,245,613 | 2,998 | alexey-milovidov | > Probably, compression size won't be less than for numeric column.
Compressed size for LowCardinality column will be somewhat less than for ordinary column and this column can be selected for PREWHERE optimization automatically.
We should not implement any additional heuristics specifically for LowCardinality co... | 2020-05-05T19:01:41 | 2020-05-05T19:01:41 | {} | MEMBER |
ClickHouse | ClickHouse | 624,410,511 | 10,671 | den-crane | approach to STR:
412.32 MiB VS 64.02 MiB
20.3.8.53
```
CREATE TABLE xxxt (
id UInt32,
col_name1 Int32,
col_name2 Int32,
col_name3 Float64,
col_name4 Float64,
col_name5 Int32,
col_name6 String,
col_name7 String,
col_name8 Float64,
col_name9 Float64,
col_name10 Int32,
col_name11 String,
... | 2020-05-06T02:28:20 | 2020-05-06T02:35:47 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,415,240 | 9,786 | den-crane | @mdelvecchio-wasabi
I closed this (my) issue because it was not a bug at all.
I had a wrong expectation that it should work but it should not.
WRONG: select toDateTime64(1584009186942540518, 9)
RIGHT: select toDateTime64('1584009186.942540518', 9)
| 2020-05-06T02:51:01 | 2020-05-06T02:51:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,487,655 | 10,667 | qoega | There is only one place where `max_threads_!=temporary_data_merge_threads_`? Or you have found more? https://github.com/ClickHouse/ClickHouse/blob/e34a9457e664fc6ea05936cd8d7a97a6759456e1/src/Interpreters/InterpreterSelectQuery.cpp#L1734 | 2020-05-06T07:29:29 | 2020-05-06T07:29:29 | {} | MEMBER |
ClickHouse | ClickHouse | 624,708,643 | 10,694 | XavierQuiniou | I generalized the solution you gave us.
Thank you for this quick and efficient response.
| 2020-05-06T15:12:04 | 2020-05-06T15:12:04 | {} | NONE |
ClickHouse | ClickHouse | 624,802,277 | 10,671 | KochetovNicolai | Was fixed in 4837257bca175f52ad9c2b5530229b095a44c05c | 2020-05-06T18:03:58 | 2020-05-06T18:03:58 | {} | MEMBER |
ClickHouse | ClickHouse | 624,821,324 | 10,667 | palasonic1 | > There is only one place where `max_threads_!=temporary_data_merge_threads_`? Or you have found more?
>
> https://github.com/ClickHouse/ClickHouse/blob/e34a9457e664fc6ea05936cd8d7a97a6759456e1/src/Interpreters/InterpreterSelectQuery.cpp#L1734
Yes, only one. | 2020-05-06T18:41:55 | 2020-05-06T18:41:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,887,573 | 9,117 | icoleslaw | Just ran into this while testing out some queries on a temporary table. It would be nice to have. | 2020-05-06T20:58:21 | 2020-05-06T20:58:40 | {} | NONE |
ClickHouse | ClickHouse | 624,965,415 | 10,534 | alexey-milovidov | Note: if you compare re2 with re2_st, the difference looks like this:
```
2,33% libpthread-2.30.so [.] __pthread_rwlock_unlock
1,56% libpthread-2.30.so [.] __pthread_rwlock_rdlock
``` | 2020-05-07T00:47:04 | 2020-05-07T00:47:04 | {} | MEMBER |
ClickHouse | ClickHouse | 624,252,297 | 10,539 | orloffv | it's critical bug for us. we can't update clickhouse for new version( | 2020-05-05T19:14:32 | 2020-05-05T19:59:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,287,538 | 10,539 | alexey-milovidov | @orloffv Could you please share your use case (queries)? Maybe it is slightly different than the original. | 2020-05-05T20:26:20 | 2020-05-05T20:26:20 | {} | MEMBER |
ClickHouse | ClickHouse | 624,395,050 | 10,671 | den-crane | What are the types of created_date / updated_at/ created_at / id ?
>Is it ok to upgrade to v20.3 with experimental_use_processors = 0 config ?
yes
----
> INDEX min_max created_date TYPE minmax GRANULARITY 8192
SKIP INDEX GRANULARITY 8192 -- has no sense. Try GRANULARITY 1..5.
| 2020-05-06T01:19:00 | 2020-05-06T01:29:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,475,400 | 10,676 | 4ertus2 | The setting that was needed for roll back to 19.16 has been backported. So it should be possible to downgrade to 19.16 and upgrade to newer version. https://github.com/ClickHouse/ClickHouse/blob/19.16/dbms/src/Core/Settings.h#L383 | 2020-05-06T06:55:10 | 2020-05-06T06:55:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,592,768 | 10,684 | tavplubix | > 00834_cancel_http_readonly_queries_on_client_close failed :(
It failed only with unbundled build, because Poco library without out patches is used in this build. I will disable this test with unbundled build. | 2020-05-06T11:24:53 | 2020-05-06T11:24:53 | {} | MEMBER |
ClickHouse | ClickHouse | 624,828,422 | 9,621 | den-crane | Seems it's solved in master and 20.3.8.53
```
CREATE TABLE metrics (`Path` String, `Value` Float64, `Time` UInt32, `Date` Date, `Timestamp` UInt32)
ENGINE = MergeTree PARTITION BY toYYYYMM(Date) ORDER BY (Path, Time);
insert into metrics select number, 0, toDateTime('2000-01-01 00:00:00') + intDiv(number,100),... | 2020-05-06T18:55:18 | 2020-05-06T19:04:19 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,866,777 | 10,680 | filimonov | Can you create a parquet files with just a few rows which are failing during import and zip / attach them here? | 2020-05-06T20:16:17 | 2020-05-06T20:16:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,886,350 | 10,471 | azat | AFAICS the problem here is different, distributed engine do not account current database, IOW:
```sql
use system;
create table dist_parts as parts engine=Distributed(test_cluster_two_shards, currentDatabase(), parts);
select * from system.dist_parts l join parts r using name format Null;
```
Will trigger the ... | 2020-05-06T20:55:37 | 2020-05-06T20:55:37 | {} | MEMBER |
ClickHouse | ClickHouse | 624,889,327 | 10,690 | filimonov | Keep calm @XavierQuiniou. Altinity's here :)
Of course we aware that there are certain limitations (and workaround for them - check that issue and comments:
https://github.com/ClickHouse/ClickHouse/issues/3830 )
That issue will definitely be fixed - it was not done yet just because it's quite complicated, and ... | 2020-05-06T21:02:13 | 2020-05-06T21:08:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,224,190 | 7,971 | alexey-milovidov | It's easy to implement a feature to allow this behaviour under setting. #10682
| 2020-05-05T18:19:40 | 2020-05-05T18:19:40 | {} | MEMBER |
ClickHouse | ClickHouse | 624,254,370 | 6,445 | alexey-milovidov | I don't see how it can be related to ODBC driver because it doesn't have retries with 1 and 10 seconds for sending data. BTW, HTTP POST queries that ODBC driver is using should not be retried at all.
It may happen due to some logic on application side or in some proxy. | 2020-05-05T19:18:40 | 2020-05-05T19:18:40 | {} | MEMBER |
ClickHouse | ClickHouse | 624,399,881 | 10,671 | den-crane | > id UInt32
> created_at , updated_at DateTime
> created_date Date
Thank you.
> as you know 8192 is default GRANULARITY, you are saying the default value is not appropriate ?
8192 is default GRANULARITY for a table SPARSE INDEX and it means rows.
Unfortunately SKIP indexes GRANULARITY has the same name but ... | 2020-05-06T01:40:17 | 2020-05-06T01:40:17 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 604,799,208 | 9,887 | leoones | **When can I support this** ？？
CTE syntax is different
CTE returning a table is not supported
Window functions are not supported
RANK OVER is not supported
GROUPING aggregate function not implemented | 2020-03-27T03:43:04 | 2020-05-06T11:07:44 | {} | NONE |
ClickHouse | ClickHouse | 624,656,213 | 10,692 | den-crane | Do you use circle replication?
Can you show clickhouse server logs with this error? | 2020-05-06T13:40:39 | 2020-05-06T13:40:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,720,520 | 10,583 | qoega | Your request kills 20.4 prestable
```
2020.05.06 18:27:31.772125 [ 745456 ] {} <Trace> BaseDaemon: Received signal 11
2020.05.06 18:27:31.772330 [ 748841 ] {} <Fatal> BaseDaemon: ########################################
2020.05.06 18:27:31.772405 [ 748841 ] {} <Fatal> BaseDaemon: (version 20.4.2.2 (official build))... | 2020-05-06T15:32:13 | 2020-05-06T15:32:13 | {} | MEMBER |
ClickHouse | ClickHouse | 624,834,948 | 10,709 | KochetovNicolai | It was probably fixed here: https://github.com/ClickHouse/ClickHouse/pull/10660
You may install testing package from [here](https://clickhouse-builds.s3.yandex.net/10660/b33fd6f2ee394a7e343dde6d1cf5c86f62c908f2/report.html) to test if it really was. Or just wait for the next 20.3 release (it will be soon). | 2020-05-06T19:08:47 | 2020-05-06T19:08:47 | {} | MEMBER |
ClickHouse | ClickHouse | 624,871,189 | 10,709 | vfilatov | @KochetovNicolai looks to be working on the testing package
ClickHouse server version 20.5.1 revision 54435.
Thanks! | 2020-05-06T20:25:14 | 2020-05-06T20:25:14 | {} | NONE |
ClickHouse | ClickHouse | 624,892,582 | 3,889 | filimonov | Related #3830 | 2020-05-06T21:09:27 | 2020-05-06T21:09:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,966,107 | 10,534 | alexey-milovidov | But the difference in performance is usually very small. | 2020-05-07T00:49:47 | 2020-05-07T00:49:47 | {} | MEMBER |
ClickHouse | ClickHouse | 624,524,190 | 10,689 | BayoNet | @blinkov Looks like it should help with broken links :-) | 2020-05-06T08:54:48 | 2020-05-06T08:54:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,561,086 | 7,455 | nvartolomei | #10316 | 2020-05-06T10:14:37 | 2020-05-06T10:14:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,660,716 | 10,669 | den-crane | ALTER TABLE t RENAME COLUMN x TO old, rename COLUMN new to x
also does not work | 2020-05-06T13:48:54 | 2020-05-06T13:48:54 | {
"confused": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,862,248 | 10,684 | champtar | Will this be in next 20.3 ? | 2020-05-06T20:06:45 | 2020-05-06T20:06:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,393,465 | 10,671 | hoseiney | **it is solved** with `set experimental_use_processors=0`
currently I have downgraded it into v20.1
Is it ok to upgrade to v20.3 with `experimental_use_processors = 0` config ?
**logs** :
query : select * from table limit 1;
1- with experimental_use_processors=0 :
```
Key condition: unknown
MinMax index co... | 2020-05-06T01:12:35 | 2020-05-06T01:12:35 | {} | NONE |
ClickHouse | ClickHouse | 624,407,849 | 10,671 | den-crane | @hoseiney in users' profile
```
cat /etc/clickhouse-server/conf.d/experimental_use_processors.xml
<?xml version="1.0" ?>
<yandex>
<profiles>
<default>
<experimental_use_processors>0</experimental_use_processors>
</default>
</profiles>
</yandex>
``` | 2020-05-06T02:17:22 | 2020-05-06T02:17:47 | {
"+1": 1,
"heart": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,810,110 | 10,690 | den-crane | Who is Altinity? Why I should worry about it? Or about TABLEAU?
It is what it is. CH is an open source product driven by the community. It will be implemented eventually. I hope. | 2020-05-06T18:19:30 | 2020-05-06T18:19:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,865,688 | 10,679 | filimonov | https://github.com/ClickHouse/ClickHouse/issues/7080 | 2020-05-06T20:13:51 | 2020-05-06T20:13:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 625,140,500 | 10,680 | Crazylearner30 | Hi, I am attaching the sample failing data here
Note: Schema given above is not the exact Schema we have used its just the example.
Let me know if i have to send the schema privately if required for further investigation.
[Failing_data.zip](https://github.com/ClickHouse/ClickHouse/files/4592048/Failing_data.zi... | 2020-05-07T09:29:41 | 2020-05-07T09:29:41 | {} | NONE |
ClickHouse | ClickHouse | 624,507,411 | 10,471 | 4ertus2 | It's most probable duplicate for #4756
Need reproduction on 20.4+ | 2020-05-06T08:17:08 | 2020-05-06T08:19:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,642,526 | 10,693 | filimonov | Few questions:
1 - did you have enough RAM during insert (check free)
2 - does the issue reproduces
3 - copy log with stracktrace please. | 2020-05-06T13:16:04 | 2020-05-06T13:16:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,656,563 | 9,786 | mdelvecchio-wasabi | I guess I disagree with the "WRONG" assessment. I think this should work, as nanoseconds since the epoch is the natural value for DateTime64(9). Particularly since this notation works in the VALUES clause.
Thanks. | 2020-05-06T13:41:21 | 2020-05-06T13:41:21 | {} | NONE |
ClickHouse | ClickHouse | 624,856,032 | 10,684 | tavplubix | Both pull requests update poco submodule to `9b347d4ab71e4436d8215aace978024e93462731`, no additional updates required | 2020-05-06T19:53:50 | 2020-05-06T19:53:50 | {} | MEMBER |
ClickHouse | ClickHouse | 625,064,329 | 10,622 | evillique | Queries for our performance test look like this:
| 2020-05-07T06:48:27 | 2020-05-07T06:53:42 | {} | MEMBER |
ClickHouse | ClickHouse | 624,242,660 | 8,020 | alexey-milovidov | @orangeops You can explicitly specify the desired timezone in DateTime type to avoid this issue:
```
DateTime('Asia/Shanghai')
```
and I hope that pg2ch will understand that. | 2020-05-05T18:55:45 | 2020-05-05T18:55:45 | {} | MEMBER |
ClickHouse | ClickHouse | 624,413,839 | 9,786 | mdelvecchio-wasabi | I don't understand why this issue was closed with a fix for the problem. I am seeing the same error in https://github.com/ClickHouse/ClickHouse/issues/10661.
We hacked our way around the 1-second resolution of DateTime columns by adding a UInt64 column of nanoseconds since the epoch. But when I try to convert that... | 2020-05-06T02:43:35 | 2020-05-06T02:43:35 | {} | NONE |
ClickHouse | ClickHouse | 624,935,760 | 8,427 | CurtizJ | Sorry for long response.
About your additional question:
Optimization didn't work in the second case. Currently optimizer can optimize cases, when one of the primary columns is wrapped into monotonic function in `select` query, but not when it wrapped into monotonic function in order key of table and you drop this fu... | 2020-05-06T23:04:52 | 2020-05-06T23:04:52 | {} | MEMBER |
ClickHouse | ClickHouse | 625,021,058 | 9,899 | sundy-li | Also `getMacro('replica')` failed. | 2020-05-07T04:23:13 | 2020-05-07T04:23:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 625,120,344 | 10,588 | excitoon | This is wrong fix, which makes things even worse. | 2020-05-07T08:49:32 | 2020-05-07T08:49:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 624,480,033 | 10,666 | qoega | @alexey-milovidov `pr-build` label is about packaging, build and test scripts. It is not intended to be used in refactoring PRs. | 2020-05-06T07:08:38 | 2020-05-06T07:08:38 | {} | MEMBER |
ClickHouse | ClickHouse | 624,844,841 | 10,704 | CurtizJ | https://clickhouse-test-reports.s3.yandex.net/10704/80cf8881da9474282298345dada28d5463c706f8/performance_comparison/all-queries.html | 2020-05-06T19:30:08 | 2020-05-06T19:30:08 | {} | MEMBER |
ClickHouse | ClickHouse | 624,858,992 | 10,534 | alexey-milovidov | > Re-arrange `all_matches`
It does not correspond to the "You don't pay for what you don't need" principle. | 2020-05-06T20:00:11 | 2020-05-06T20:00:11 | {} | MEMBER |
ClickHouse | ClickHouse | 621,766,764 | 7,714 | d-ganchar | Repeated again.
There is a table with data(`engine = MergeTree() PARTITION BY toMonth(some_col)`). Data is collected every day. I need to add a column. I did:
`ALTER TABLE my_tbl ADD COLUMN col_name Nullable(Float64) DEFAULT NULL;`
Past queries works fine. But I got the error when I try to add a new column into ... | 2020-04-30T11:09:20 | 2020-04-30T11:09:20 | {} | NONE |
ClickHouse | ClickHouse | 622,437,442 | 10,585 | alexey-milovidov | Exception handling is broken: https://clickhouse-test-reports.s3.yandex.net/10585/f11148a77c1f387158fe43f3203d47402aee557f/functional_stateless_tests_(thread)/test_run.txt.out.log
But only in sanitizer builds.
Let's disable ThinLTO for builds with sanitizers. | 2020-05-01T15:36:25 | 2020-05-01T15:37:19 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.