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 | 699,907,909 | 15,383 | filimonov | You use obsolete ClickHouse version. Please update.
Default settings for `insert ... select ...` are quire conservative (they not use your RAM too aggressively, and formed blocks are not too big).
Try adjusting min_insert_block_size_bytes / min_insert_block_size_rows.
Try (for example)
```
set max_insert_th... | 2020-09-28T09:58:40 | 2020-09-28T09:59:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,846,945 | 15,301 | filimonov | > Usually you don't need a core dump to decode address. You can copy the following binary from the system:
> `/usr/lib/debug/lib/x86_64-linux-gnu/libc-2.*.so`
> and use `addr2line` to decode the address.
>
> PS. I did not try it but it should work.
I've tried that - w/o success :|
I've asked the guys to col... | 2020-09-28T07:59:43 | 2020-09-28T12:05:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,519,101 | 15,323 | alexey-milovidov | The dataset from 2011 until today is just about 2.5 billion records and 146 GB in size in ClickHouse. | 2020-09-26T16:45:13 | 2020-09-27T00:46:02 | {} | MEMBER |
ClickHouse | ClickHouse | 699,606,435 | 8,545 | GPF199541 | Connecting to 127.0.0.1:9000 as user default.
Connected to ClickHouse server version 20.9.2 revision 54439.
dbrtmappredb01 :) SELECT * FROM (SELECT a, runningDifference(a) rd FROM (SELECT * FROM (SELECT 1 a UNION ALL SELECT 2 a ) ORDER BY a DESC ));
SELECT *
FROM
(
SELECT
a,
runningDi... | 2020-09-27T08:51:00 | 2020-09-27T10:04:12 | {} | NONE |
ClickHouse | ClickHouse | 699,776,986 | 15,186 | knoguchi | I'm on Mac + CLion. I had to comment out this particular test. Not sure why it fails but I got the same error message at the same line. I'm using the HEAD of master.
```
% git diff src/Common/tests/int_hashes_perf.cpp
diff --git a/src/Common/tests/int_hashes_perf.cpp b/src/Common/tests/int_hashes_perf.cpp
in... | 2020-09-28T05:23:56 | 2020-09-28T05:23:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,834,776 | 15,336 | alesapin | `test_multiple_disks/test.py::test_background_move[moving_replicated_mt-ReplicatedMergeTree('/clickhouse/moving_replicated_mt', '1')]` -- flaky | 2020-09-28T07:34:23 | 2020-09-28T07:34:23 | {} | MEMBER |
ClickHouse | ClickHouse | 699,619,761 | 15,283 | tavplubix | Fixed in #15302 | 2020-09-27T10:58:26 | 2020-09-27T10:58:26 | {} | MEMBER |
ClickHouse | ClickHouse | 699,681,240 | 15,272 | alexey-milovidov | Ok. It builds nicely in our CI somehow :) But better to fix this issue. | 2020-09-27T20:00:28 | 2020-09-27T20:00:28 | {} | MEMBER |
ClickHouse | ClickHouse | 699,778,235 | 7,595 | filimonov | Please see / follow https://github.com/ClickHouse/ClickHouse/pull/11298 | 2020-09-28T05:27:51 | 2020-09-28T05:27:51 | {
"heart": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,835,636 | 15,221 | alesapin | Failed tests probably fixed in master | 2020-09-28T07:36:10 | 2020-09-28T07:36:10 | {} | MEMBER |
ClickHouse | ClickHouse | 699,881,121 | 15,063 | alesapin | Real race catch in `01370_client_autocomplete_word_break_characters` | 2020-09-28T09:06:28 | 2020-09-28T09:06:28 | {} | MEMBER |
ClickHouse | ClickHouse | 699,988,601 | 11,940 | aubweb9 | Hello,
I have kind of the same issue in version 20.8.3.18
2020.09.27 21:07:16.306908 [ 23103 ] {} <Fatal> BaseDaemon: ########################################
2020.09.27 21:07:16.316298 [ 23103 ] {} <Fatal> BaseDaemon: (version 20.8.3.18 (official build), build id: E6AA632BFA7BC9A5) (from thread 10426) (query_i... | 2020-09-28T12:55:58 | 2020-09-28T12:55:58 | {} | NONE |
ClickHouse | ClickHouse | 699,504,047 | 15,322 | den-crane | test failed. investigating. | 2020-09-26T14:39:17 | 2020-09-26T14:39:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,572,628 | 15,322 | den-crane | The reason -- CI runs in Moscow TZ | 2020-09-27T01:59:02 | 2020-09-27T01:59:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,576,940 | 15,295 | winter7 | Thank you! I think `create table xxx as remote();` is what I missed in my second attempt. | 2020-09-27T02:55:11 | 2020-09-27T02:55:11 | {} | NONE |
ClickHouse | ClickHouse | 699,655,947 | 7,595 | knoguchi | @yanggeorge here is the SELECT parser. Take a look in the tests directory for other parsers.
https://github.com/ClickHouse/ClickHouse/blob/master/src/Parsers/tests/select_parser.cpp | 2020-09-27T16:22:47 | 2020-09-27T16:22:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,680,296 | 15,341 | alexey-milovidov | First in the stream of data before limit. Usually it implies that you have ORDER BY, otherwise results will be non-deterministic. | 2020-09-27T19:51:43 | 2020-09-27T19:51:43 | {} | MEMBER |
ClickHouse | ClickHouse | 694,806,226 | 14,939 | nikitamikhaylov | Strange test failure, I will take a look.
UPD: The container with the copier didn't start. | 2020-09-18T11:09:33 | 2020-09-28T10:57:15 | {} | MEMBER |
ClickHouse | ClickHouse | 700,033,136 | 15,239 | alesapin | Let's look at perf tests! | 2020-09-28T14:11:37 | 2020-09-28T14:11:37 | {} | MEMBER |
ClickHouse | ClickHouse | 699,385,669 | 15,282 | alexey-milovidov | test_materialize_mysql_database - flaky test. | 2020-09-26T04:33:19 | 2020-09-26T04:33:19 | {} | MEMBER |
ClickHouse | ClickHouse | 699,561,680 | 15,306 | den-crane | STR:
```
$ cat t.sql
drop table if exists st;
CREATE TABLE st(sssss String, pids Array(UInt16), t DateTime)
ENGINE = MergeTree() ORDER BY tuple();
drop table if exists stv;
CREATE MATERIALIZED VIEW stv ENGINE = MergeTree() ORDER BY tuple()
AS SELECT toStartOfHour(t) AS t, pids[num] pid
FROM st ARRAY JO... | 2020-09-26T23:44:26 | 2020-09-26T23:54:24 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,569,784 | 15,278 | byx313 | thank you den! | 2020-09-27T01:27:27 | 2020-09-27T01:27:27 | {} | NONE |
ClickHouse | ClickHouse | 699,599,491 | 15,272 | hexiaoting | When I use gcc10, There is still some error:
```
FAILED: utils/db-generator/CMakeFiles/query_db_generator.dir/query_db_generator.cpp.o
/data/users/guojiantao/gentoo/usr/bin/c++ -DAWS_SDK_VERSION_MAJOR=1 -DAWS_SDK_VERSION_MINOR=7 -DAWS_SDK_VERSION_PATCH=231 -DENABLE_OPENSSL_ENCRYPTION -DPOCO_ENABLE_CPP11 -DPOCO_HAVE_... | 2020-09-27T07:46:23 | 2020-09-27T07:46:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,636,871 | 15,173 | den-crane | @keliss it's not about only ALTER. And it's about memory structures and mutexes. (CH does not database object level locks)
Deadlock was possible before 19.14.
AFAIK even for 2 concurrent queries `select ... from system.tables` & `drop table`. To solve such deadlocks a check "Possible deadlock avoided." was impleme... | 2020-09-27T13:38:30 | 2020-09-27T13:39:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,681,746 | 15,352 | alexey-milovidov | You have to create separate table as a storage for materialized view (for `TO` clause) or remove `TO` clause altogether - then it will create the inner storage automatically. | 2020-09-27T20:04:25 | 2020-09-27T20:04:25 | {} | MEMBER |
ClickHouse | ClickHouse | 699,864,509 | 15,365 | dengc367 | I have encountered the problem, and should delete that table metadata info in the clickhouse `data` directory. | 2020-09-28T08:34:11 | 2020-09-28T08:34:11 | {} | NONE |
ClickHouse | ClickHouse | 699,913,095 | 15,383 | qinzl1 | @filimonov
I/O is falling during insertion
| 2020-09-28T10:08:08 | 2020-09-28T10:08:08 | {} | NONE |
ClickHouse | ClickHouse | 699,976,151 | 15,239 | alesapin | Surprisingly it was able to build on my machine after merge with master. | 2020-09-28T12:30:23 | 2020-09-28T12:30:23 | {} | MEMBER |
ClickHouse | ClickHouse | 699,980,191 | 15,306 | den-crane | it's definitely a bug.
>Does this mean that we should stop using arrayEnumerate and from version 20.8.3.18 start using your solution?
It does mean that you should not use 20.8.
>ArrayEnumerate was very elegant solution
It's not elegant at all.
Nobody use it for this usecase. ArrayEnumerate brings excessiv... | 2020-09-28T12:39:02 | 2020-09-28T12:41:38 | {
"+1": 1,
"rocket": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 700,048,233 | 15,414 | nvartolomei | > If I'm not mistaken we can't retry the query after we got first data block so this could be used only for retrying exceptions that would be thrown prior to query execution (too many simultaneous queries is ok to retry but io errors during data receiving are not ok to retry).
Should ClickHouse in this case retry th... | 2020-09-28T14:35:56 | 2020-09-28T14:35:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 700,095,561 | 14,760 | tavplubix | Stress test (thread) — Sanitizer assert - known race in openssl | 2020-09-28T15:34:12 | 2020-09-28T15:34:12 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 699,719,308 | 15,318 | alexey-milovidov | ```
CREATE TABLE wikistat
(
`time` DateTime,
`project` LowCardinality(String),
`subproject` LowCardinality(String),
`path` String,
`hits` UInt64,
`size` UInt64
)
ENGINE = MergeTree
ORDER BY (path, time)
```
```
ls -1 pageviews*.gz | xargs -P16 -I{} bash -c "
echo '{}';
... | 2020-09-28T01:14:38 | 2020-09-28T01:14:38 | {} | MEMBER |
ClickHouse | ClickHouse | 699,535,109 | 15,301 | alexey-milovidov | The details of this bug:
1. We are using asynchronous unwiding for query profiler.
Workaround: setting `query_profiler_real_time_period_ns` and `query_profiler_cpu_time_period_ns` to 0 will solve the issue.
2. ClickHouse is almost statically linked. The only part of ClickHouse that is out of our control is `lib... | 2020-09-26T19:00:35 | 2020-09-26T19:02:33 | {} | MEMBER |
ClickHouse | ClickHouse | 699,540,254 | 15,323 | alexey-milovidov | I will try to clone all repositories with at least one star:
```
~/github$ cut -f1 repos.tsv | while read repo; do echo "$repo" && mkdir -p "$repo" && git clone --single-branch --no-tags "git@github.com:$repo" "$repo"; done
996icu/996.ICU
Cloning into '996icu/996.ICU'...
remote: Enumerating objects: 9733, done.
r... | 2020-09-26T19:49:30 | 2020-09-27T03:12:07 | {
"eyes": 1
} | MEMBER |
ClickHouse | ClickHouse | 699,913,061 | 15,370 | filimonov | Duplicate of https://github.com/ClickHouse/ClickHouse/issues/10212#issuecomment-612676750 | 2020-09-28T10:08:04 | 2020-09-28T10:08:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,438,821 | 15,323 | alexey-milovidov | 1. `wget --continue https://data.gharchive.org/{2015..2020}-{01..12}-{01..31}-{0..23}.json.gz`
Download will take multiple days.
2.
```
CREATE TABLE github_events
(
event_type Enum('CommitCommentEvent' = 1, 'CreateEvent' = 2, 'DeleteEvent' = 3, 'ForkEvent' = 4,
'GollumEvent' = 5, 'Iss... | 2020-09-26T06:31:02 | 2020-09-27T00:47:23 | {} | MEMBER |
ClickHouse | ClickHouse | 699,751,904 | 14,893 | guanlisheng | many thanks and look forward | 2020-09-28T03:39:19 | 2020-09-28T03:39:19 | {} | NONE |
ClickHouse | ClickHouse | 699,534,200 | 15,301 | alexey-milovidov | Usually you don't need a core dump to decode address. You can copy the following binary from the system:
`/usr/lib/debug/lib/x86_64-linux-gnu/libc-2.*.so`
and use `addr2line` to decode the address.
PS. I did not try it but it should work. | 2020-09-26T18:53:03 | 2020-09-26T18:53:38 | {} | MEMBER |
ClickHouse | ClickHouse | 699,609,356 | 14,436 | peaksnail | hi
the lock still exists on the version 20.6
ch version 20.6.6.7
log
```
Received exception from server (version 20.6.6):
Code: 473. DB::Exception: Received from 172.18.122.58:7000. DB::Exception: WRITE locking attempt on "taotie.abacus_merchant_desktop" has timed out! (120000ms) Possible deadlock avoided. C... | 2020-09-27T09:16:25 | 2020-09-27T09:16:25 | {
"+1": 3
} | NONE |
ClickHouse | ClickHouse | 699,636,178 | 8,545 | den-crane | @GPF199541 No solution. 2 blocks. runningDifference works only in block.
You can use arrays
```
SELECT arrayJoin(arrayDifference(groupArray(a))) from
(select * FROM (SELECT 1 a UNION ALL SELECT 2 a ) ORDER BY a DESC)
``` | 2020-09-27T13:32:28 | 2020-09-27T13:32:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,758,745 | 15,365 | zhang2014 | I will try to fix it. | 2020-09-28T04:11:34 | 2020-09-28T04:17:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,910,352 | 15,375 | filimonov | This is by intent (not cleat what type of result is exected when operation between float and decimal happens).
You need to cast your Float to Decimal first:
```
select price - toDecimal32(0.0,4) from default.num;
``` | 2020-09-28T10:02:57 | 2020-09-28T10:02:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,924,188 | 15,369 | BernieJiangDL | > To be able to help you we need more details:
>
> 1. clickhouse version
> 2. table schema
> 3. result of `select * from system.mutations where is_done=0`
1. ClickHouse server version 20.3.8 revision 54433
2. CREATE TABLE test.scene_cnt_table (`day` Int64 , `hour` Int64 ', `minute` Int64 , `scene` Int64 , `c... | 2020-09-28T10:31:57 | 2020-09-28T10:32:20 | {} | NONE |
ClickHouse | ClickHouse | 699,455,121 | 15,317 | amosbird | you can also do
```
WITH tuple(COLUMNS('tags_')) AS o
select
COLUMNS('tags_'),
intDiv(toUInt64(timestamp), 30) * 30 AS t,
max(value)
from
table
GROUP BY
t,
COLUMNS('tags_')
ORDER BY o, t
``` | 2020-09-26T08:05:27 | 2020-09-26T08:05:27 | {
"+1": 1
} | COLLABORATOR |
ClickHouse | ClickHouse | 699,540,056 | 15,323 | alexey-milovidov | This dataset can solve most of the subtasks from https://github.com/ClickHouse/ClickHouse/issues/10616
Even more, it can do it for all existing GitHub repositories! | 2020-09-26T19:47:31 | 2020-09-27T03:05:39 | {} | MEMBER |
ClickHouse | ClickHouse | 699,626,869 | 15,341 | amosbird | Interesting. What's the definition of "first" here? | 2020-09-27T12:06:52 | 2020-09-27T12:06:52 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 699,681,532 | 15,352 | alexey-milovidov | > CREATE MATERIALIZED VIEW IF NOT EXISTS test_view_materialized
**to test_view** as select execution_time,count() **from test_view** group by execution_time;
It means that you are going to store the data of the materialized view in the same table - infinite loop. | 2020-09-27T20:02:50 | 2020-09-27T20:02:50 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 699,833,627 | 15,339 | alesapin | `test_adaptive_granularity/test.py::test_version_update_two_nodes` falky again | 2020-09-28T07:32:03 | 2020-09-28T07:32:03 | {} | MEMBER |
ClickHouse | ClickHouse | 699,845,344 | 15,306 | cybertjofi | Den, thank you for this workaround and for your fast response. We will try to implement this in our system.
Tell me if I am wrong, as I understand, this is clearly a bug? Does this mean that we should stop using arrayEnumerate and from version 20.8.3.18 start using your solution?
ArrayEnumerate was very elegant sol... | 2020-09-28T07:56:11 | 2020-09-28T07:56:11 | {} | NONE |
ClickHouse | ClickHouse | 699,864,920 | 15,200 | yiquanzhou | Hi @den-crane,
Although the workaround looks easy by just wrapping the column name with backquotes like for an alias, it's not trivial to do so in our code because we have built a DSL in Java to build and render the queries. It may impact many other queries and our unit tests so it's pretty annoying and it's blocking ... | 2020-09-28T08:35:02 | 2020-09-28T08:35:27 | {} | NONE |
ClickHouse | ClickHouse | 699,899,064 | 13,593 | toby080508050805 | /etc/security/limit.config
add clickhouse user thread limit. | 2020-09-28T09:41:03 | 2020-09-28T09:41:03 | {} | NONE |
ClickHouse | ClickHouse | 699,999,029 | 15,411 | SergeyMirvoda | Here is the stack from another server (Ubuntu 18.04) with the same query as in the issue.
Also, it reproduced on ClickHouse 20.6.6.7 (CentOS 7)
```
2020.09.28 12:26:38.560516 [ 11396 ] {0738b390-f230-47e5-8a42-76e0222b89b7} <Error> executeQuery: Code: 43, e.displayText() = DB::Exception: Second argument to tupleElem... | 2020-09-28T13:14:53 | 2020-09-28T13:17:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,389,612 | 14,977 | alexey-milovidov | I'm going to review in a few days... | 2020-09-26T04:36:58 | 2020-09-26T04:36:58 | {
"eyes": 1
} | MEMBER |
ClickHouse | ClickHouse | 699,445,711 | 15,321 | amosbird | btw, we can also feed the output obfuscated query to the formatter. It may generates better output and can catch some syntax errors. | 2020-09-26T07:09:38 | 2020-09-26T07:09:38 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 699,549,212 | 15,312 | alexey-milovidov | Will rerun CI checks... | 2020-09-26T21:19:22 | 2020-09-26T21:19:22 | {} | MEMBER |
ClickHouse | ClickHouse | 698,941,298 | 15,306 | den-crane | sidenote:
you don't need arrayEnumerate.
These work
```
partner_id
...
ARRAY JOIN partner_ids AS partner_id
```
or not simply arrayJoin
```
select .... arrayJoin(partner_ids) as partner_id
```
without ARRAY JOIN | 2020-09-25T13:51:59 | 2020-09-26T23:46:56 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,590,870 | 15,351 | alexey-milovidov | Some tests require some configuration tunes to run successfully. They are located in `tests/config`.
Here is the log of the latest test run in CI: https://clickhouse-test-reports.s3.yandex.net/0/bbbe51033dfd5b8c3d54e168475ca707ac7ec0b4/functional_stateless_tests_(release)/test_run.txt.out.log
| 2020-09-27T06:06:09 | 2020-09-27T06:06:36 | {} | MEMBER |
ClickHouse | ClickHouse | 699,639,844 | 15,173 | keliss | So you're saying that different ALTER behavior on the newer versions won't solve the issue? Or that it is still possible in theory but with fixed ALTER it will most probably be gone? What I'm trying to understand is how we can avoid or mitigate this next time and if this can be reproduced and fixed. | 2020-09-27T14:06:02 | 2020-09-27T14:29:51 | {} | NONE |
ClickHouse | ClickHouse | 699,947,439 | 15,063 | alesapin | Waiting only for Functional stateless with thread sanitizer. | 2020-09-28T11:25:45 | 2020-09-28T11:25:45 | {} | MEMBER |
ClickHouse | ClickHouse | 699,958,120 | 1,841 | Tasselmi | hey guys, any progress on this topic?
It really helpful to use map type to store key-values.
When I dump hive data into clickhouse, this made me mad. `Arrya/Map/Struct` THEY ARE SO fundamental ~ | 2020-09-28T11:49:49 | 2020-09-28T11:49:49 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 699,976,619 | 15,301 | filimonov | Wondering also - can't it be caused by some of security hardening flags, used by CentOS 8?
https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/8/html-single/8.0_release_notes/index
> The following security hardening flags are applied:
>
> - For detection of buffer-overflow errors: D_FORTIFY... | 2020-09-28T12:31:21 | 2020-09-28T12:31:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 700,046,367 | 15,414 | nvartolomei | Looks like these retries could be added to `RemoteQueryExecutor` to the read method https://github.com/ClickHouse/ClickHouse/blob/c48d212320bab7fb48257165d6e24f81e005f491/src/DataStreams/RemoteQueryExecutor.cpp#L169-L245.
If I'm not mistaken we can't retry the query after we got first data block so this could be use... | 2020-09-28T14:32:51 | 2020-09-28T14:33:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,482,988 | 15,265 | bohutang | A1:
`REPLICATION` privilege is need, doc:
https://dev.mysql.com/doc/refman/8.0/en/replication-howto-repuser.html
A2:
MaterializeMySQL supports ALTER DDL(including ADD_COLUMN/DROP_COLUMN/MODIFY_COLUMN/RENAME_TABLE, it's not documented yet) here:
https://github.com/ClickHouse/ClickHouse/blob/master/src/Interpreter... | 2020-09-26T11:28:24 | 2020-09-26T11:28:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,575,642 | 15,317 | detailyang | @amosbird Many thanks:) It works.
Consider it's solved | 2020-09-27T02:38:03 | 2020-09-27T02:38:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,575,712 | 15,351 | elderbig | **CPU INFO**
```
[ClickHouse]$ cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 166
model name : Intel(R) Core(TM) i7-10710U CPU @ 1.10GHz
stepping : 0
microcode : 0xca
cpu MHz : 3881.913
cache size : 12288 KB
physical id : 0
siblings : 12
core id : 0
cpu cores : 6
apic... | 2020-09-27T02:38:56 | 2020-09-27T02:40:45 | {} | NONE |
ClickHouse | ClickHouse | 699,592,344 | 10,745 | Ernestyj | I also frequently met below error:
<Error> executeQuery: Code: 210, e.displayText() = DB::NetException: I/O error: Broken pipe, while writing to socket
<Error> ServerErrorHandler: Poco::Exception. Code: 1000, e.code() = 104, e.displayText() = Connection reset by peer, Stack trace (when copying this message, always ... | 2020-09-27T06:25:20 | 2020-09-27T06:25:20 | {} | NONE |
ClickHouse | ClickHouse | 699,761,747 | 11,512 | alexey-milovidov | @gopala000 Yes, it is scheduled for 20.10 release.
| 2020-09-28T04:24:12 | 2020-09-28T04:24:12 | {} | MEMBER |
ClickHouse | ClickHouse | 699,834,687 | 15,336 | alesapin | `test_storage_rabbitmq` -- disabled in master | 2020-09-28T07:34:13 | 2020-09-28T07:34:13 | {} | MEMBER |
ClickHouse | ClickHouse | 699,924,187 | 15,384 | zhang2014 | It should make sense to set `mysql_datatypes_support_level` to immutable. because the table cache will be inconsistent when it changes. | 2020-09-28T10:31:56 | 2020-09-28T10:32:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,965,293 | 15,306 | cybertjofi | I tried to use this logic on production table where we have 5 arrays of data. And when i insert one row with 5 arrays in it, and every array has 4 elements, I end up having 20 rows in my MV table...hmmm
Am I using this wrong? I put this logic in select just like you told me: arrayJoin(partner_ids) as partner_id, array... | 2020-09-28T12:06:02 | 2020-09-28T12:06:02 | {} | NONE |
ClickHouse | ClickHouse | 699,625,530 | 15,270 | filimonov | See also #12188 #13403 | 2020-09-27T11:54:38 | 2020-09-27T11:54:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 698,086,132 | 10,954 | amosbird | > Will dictGet(..) be performed by DT and query sent to the data nodes be re-written as WITH 10 as _ver SELECT ... ?
That's the vanilla ClickHouse WITH, which introduces some aliases to the current query scope. I'd like to name it as CSE (common scalar expression) because it can only hold scalar values. The expressi... | 2020-09-24T03:16:08 | 2020-09-28T03:54:14 | {
"+1": 3
} | COLLABORATOR |
ClickHouse | ClickHouse | 699,835,958 | 15,223 | alesapin | Fixed in master? | 2020-09-28T07:36:52 | 2020-09-28T07:36:52 | {} | MEMBER |
ClickHouse | ClickHouse | 699,921,068 | 15,352 | filimonov | https://altinity.com/blog/clickhouse-materialized-views-illuminated-part-1 | 2020-09-28T10:24:59 | 2020-09-28T10:24:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,924,014 | 15,384 | zhang2014 | CC: @tavplubix | 2020-09-28T10:31:34 | 2020-09-28T10:31:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,959,856 | 15,007 | alesapin | Check `Integration tests flaky check (asan)` can be ignored. We have a lot of flaky integration tests and we will fix them gradually. | 2020-09-28T11:53:57 | 2020-09-28T11:53:57 | {} | MEMBER |
ClickHouse | ClickHouse | 700,073,777 | 15,409 | zhang2014 | Maybe use pub/sub to refactor `MaterializeMySQLSyncThread` is better ?(database and table name is topic name, DDL will be pushed in the empty string name topic). topic consumer should be dynamically added. | 2020-09-28T15:17:38 | 2020-09-28T15:17:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,518,889 | 15,321 | alexey-milovidov | Yes, it will be more convenient... only need to add "no format" option to `clickhouse-format` to also allow the current behaviour.
I will merge this now as it's 100% green. | 2020-09-26T16:43:07 | 2020-09-26T16:43:21 | {} | MEMBER |
ClickHouse | ClickHouse | 699,556,131 | 15,173 | den-crane | @simo-lineate "Possible deadlock avoided." -- not a bug.
It's a special safe-limiter to prevent deadlocks. This limiter throws the alert "Possible deadlock avoided."
This issue was solved in 20.6. No alerts anymore.
It will be never solved in 20.3 (LTS). Because it's not a bug.
Currently supported versions are 2... | 2020-09-26T22:25:38 | 2020-09-26T22:31:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 699,539,984 | 15,323 | alexey-milovidov | Some first queries:
```
SELECT repo_name, count() FROM github_events GROUP BY repo_name ORDER BY count() DESC LIMIT 50
SELECT repo_name, count() FROM github_events WHERE repo_name ILIKE '%ClickHouse%' GROUP BY repo_name ORDER BY count() DESC LIMIT 50
SELECT actor_login AS k, count() FROM github_events WHERE repo_... | 2020-09-26T19:46:42 | 2020-09-27T00:48:33 | {} | MEMBER |
ClickHouse | ClickHouse | 699,574,060 | 15,270 | BernieJiangDL | > @BernieJiangDL for now clickhouse touches every part for any mutation (alter update/delete).
> This behavior is by design. It needs to do consistent mutations.
>
> Mutations has been implemented for rare admin operations (twice a year) not for daily business tasks.
Got u! Thanks | 2020-09-27T02:17:46 | 2020-09-27T02:17:46 | {} | NONE |
ClickHouse | ClickHouse | 699,680,914 | 14,893 | alexey-milovidov | This task looks isolated, it fits nicely to ClickHouse architecture, but there is no high demand for it.
I consider it as an experimental task for external contributors. I just listed it here: https://github.com/ClickHouse/ClickHouse/issues/15065
I expect that someone will be interested in this task (it should be con... | 2020-09-27T19:57:32 | 2020-09-27T19:57:32 | {} | MEMBER |
ClickHouse | ClickHouse | 699,914,132 | 15,369 | filimonov | To be able to help you we need more details:
1) clickhouse version
2) table schema
3) result of `select * from system.mutations where is_done=0` | 2020-09-28T10:10:11 | 2020-09-28T10:10:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 700,027,812 | 2,352 | ZhangzheBJUT | > > What do these a1, c1, a2, c2... parameters represent
>
> It's just a simple demo equation. In your case, if you want to sort the events by sequential comparison by (event_time, pid), you can create an equation based on your data.
it can not use equation. the equation will make the window size not work!
i t... | 2020-09-28T14:03:01 | 2020-09-28T14:05:46 | {} | NONE |
ClickHouse | ClickHouse | 699,383,425 | 15,285 | alexey-milovidov | `SELECT now(NULL, NULL, NULL);` triggers assertion failure in debug build. | 2020-09-26T04:31:05 | 2020-09-26T04:31:05 | {} | MEMBER |
ClickHouse | ClickHouse | 699,593,100 | 14,833 | Ernestyj | I also meet lots of error like below (version 20.4.4.18 (official build)):
<Error> ServerErrorHandler: Code: 101, e.displayText() = DB::NetException: Unexpected packet from client, Stack trace (when copying this message, always include the lines below):
executeQuery: Code: 210, e.displayText() = DB::NetException:... | 2020-09-27T06:33:54 | 2020-09-27T06:33:54 | {} | NONE |
ClickHouse | ClickHouse | 699,676,821 | 15,361 | alexey-milovidov | Let's introduce a setting in config but don't change the default.
People often forget to update anything on their servers (including tzdata) and only update ClickHouse. | 2020-09-27T19:21:13 | 2020-09-27T19:21:13 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 699,860,953 | 15,367 | hexiaoting | I want this. | 2020-09-28T08:26:45 | 2020-09-28T08:26:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 700,053,437 | 15,063 | alesapin | Stress test failure is well known. | 2020-09-28T14:44:33 | 2020-09-28T14:44:33 | {} | MEMBER |
ClickHouse | ClickHouse | 699,380,758 | 15,267 | alexey-milovidov | AST Fuzzer - already known issue. | 2020-09-26T04:28:30 | 2020-09-26T04:28:30 | {} | MEMBER |
ClickHouse | ClickHouse | 699,387,962 | 15,048 | alexey-milovidov | > test_rename_column
Was temporarily broken at the moment of this PR. | 2020-09-26T04:35:27 | 2020-09-26T04:35:27 | {} | MEMBER |
ClickHouse | ClickHouse | 699,605,386 | 14,893 | guanlisheng | hey @alexey-milovidov, do we have a rough plan to integrate this c++ library? | 2020-09-27T08:40:47 | 2020-09-27T08:40:47 | {} | NONE |
ClickHouse | ClickHouse | 699,615,089 | 15,301 | spihiker | I'm afraid of Segmentation fault.Which glibc version do you recommend on centos7_64?What can be done to prevent this kind of error? Thanks. | 2020-09-27T10:12:37 | 2020-09-27T10:12:37 | {} | NONE |
ClickHouse | ClickHouse | 699,768,476 | 15,296 | yaro-137 | Thanks @den-crane that's worked for me. | 2020-09-28T04:53:32 | 2020-09-28T04:53:32 | {} | NONE |
ClickHouse | ClickHouse | 699,834,206 | 15,335 | alesapin | `test_concurrent_ttl_merges/test.py::test_no_ttl_merges_in_busy_pool` falky | 2020-09-28T07:33:16 | 2020-09-28T07:33:16 | {} | MEMBER |
ClickHouse | ClickHouse | 699,835,247 | 15,224 | alesapin | `test_materialize_mysql_database/test.py::test_materialize_database_ddl_with_empty_transaction_8_0` -- flaky, should be fixed in master | 2020-09-28T07:35:20 | 2020-09-28T07:35:20 | {} | MEMBER |
ClickHouse | ClickHouse | 699,838,774 | 5,562 | phhcer | when I select subquery like this
```clickhouse
create materialized view mv2 as
select
t1.ds,
t2.ds
from (
select
ds
from `.inner.mv1`
) t1
left join (
select
ds
from `.inner.mv1`
) t2
ON t1.ds = t2.ds+1
```
Will it use full data of mv1 or one batch of mv1?
| 2020-09-28T07:42:31 | 2020-09-28T07:42:31 | {} | NONE |
ClickHouse | ClickHouse | 700,053,896 | 15,314 | stcharitak | > You have no choice except to move un_id to the beginning of PRIMARYINDEX.
> And try to clean un_id from 'us-' and save UUID as UUID (or as hex/FIXED String) not a String.
>
> ```
>
> create table product (
> country String,
> un_id UUID,
> organization_id LowCardinality(String),... | 2020-09-28T14:45:17 | 2020-09-28T14:45:17 | {} | NONE |
ClickHouse | ClickHouse | 698,672,277 | 15,179 | chang0391 | @zhang2014
Thanks | 2020-09-25T01:35:57 | 2020-09-25T01:35:57 | {} | NONE |
ClickHouse | ClickHouse | 698,720,370 | 15,272 | alexey-milovidov | We have switched to gcc-10 long time ago. | 2020-09-25T04:59:31 | 2020-09-25T04:59:31 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.