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 | 762,960,915 | 19,272 | slimtom95 | > > output will be following, randomly..
>
> it's probably because part for b is not created yet and not visible for select.
Probably not, please check the following.
```
drop database default;
create database default;
create table test(
id UInt64,value String
) Engine=Null();
create table a as test Engi... | 2021-01-19T16:31:46 | 2021-01-19T16:31:46 | {} | NONE |
ClickHouse | ClickHouse | 763,254,589 | 19,260 | l1t1 | i run it again, and it carshed again even after i press ctrl+c
```
SELECT
name,
sum(multiIf((name = '0') AND (bitmapContains(m_bit, toUInt32(id)) = 1), sal, NULL))
FROM men
, men_bit
WHERE id < 10000
GROUP BY name
Query id: 9d5bdb18-2189-4ad4-ba8a-3dc8e09e9a10
Cancelling query.
^C
```
noh... | 2021-01-20T01:10:41 | 2021-01-20T01:10:41 | {} | NONE |
ClickHouse | ClickHouse | 762,114,160 | 18,720 | hexiaoting | @alexey-milovidov I have two questions:
Q1:
`t1 JOIN t2 ON t1.k = t2.k AND t1.size > 10` will this query throw exception or translated to: `t1 JOIN t2 ON t1.k = t2.k where t1.size > 10`
Q2: `t1 JOIN t2 ON t1.size == 10 and t2.size == 8` what kind of new sql should this query translated to ? | 2021-01-18T09:28:17 | 2021-01-20T01:49:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 763,294,098 | 12,489 | dgzdot | @alexey-milovidov Are there any plans for this issue? | 2021-01-20T02:57:20 | 2021-01-20T02:57:20 | {} | NONE |
ClickHouse | ClickHouse | 762,802,666 | 19,262 | alexey-milovidov | It can also be related to the usage of some deprecated or inconsistent settings in your self-signed sertificate.
I'm not into this topic, but look at the https://boringssl.googlesource.com/boringssl/+/d7266ecc9bf92ffad277bc39653919da79c8f42b%5E%21/ for possible explanation. | 2021-01-19T12:11:49 | 2021-01-19T12:11:49 | {} | MEMBER |
ClickHouse | ClickHouse | 762,908,272 | 19,277 | tisonkun | @alexey-milovidov how can I download the source code for this build? I am hacking clickhouse itself and the symbols conflict in darwin platform seem reasonable as above and exactly happened on my environment. I'd like to find the root cause. | 2021-01-19T15:16:06 | 2021-01-19T15:16:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,956,188 | 19,272 | slimtom95 | @den-crane Any thoughts about the log above? | 2021-01-19T16:24:28 | 2021-01-19T16:24:28 | {} | NONE |
ClickHouse | ClickHouse | 763,018,972 | 19,272 | den-crane | ```
18:53:13.602554 [ 4700 ] <Debug> executeQuery: (from [::1]:33126, using production parser) insert into a (id) values
18:53:13.603055 [ 4700 ] <Debug> HashJoin: Right sample block: b.value String String(size = 0), b.id UInt64 UInt64(size = 0)
18:53:13.604774 [ 4700 ] <Trace> default.a Renaming temporary part ... | 2021-01-19T18:01:45 | 2021-01-19T18:18:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 763,066,998 | 19,288 | alexey-milovidov | Otherwise more complications will arise, e.g. RIGHT JOIN ... WHERE column IS NULL. | 2021-01-19T19:12:25 | 2021-01-19T19:12:25 | {} | MEMBER |
ClickHouse | ClickHouse | 762,596,933 | 17,623 | zhicwu | >
>
> > I'm very interested in lightweight updates/deletes, and would like to contribute. Not sure how far @myrrc has gotten in planning this?
>
> Hi @stigsb, @myrrc works for Altinity. This is one of the features we plan to work on. Please reach us if you need more detail, we have preliminary design already ava... | 2021-01-19T04:40:53 | 2021-01-19T04:40:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,653,564 | 18,945 | alesapin | `01442_merge_detach_attach` -- https://clickhouse-test-reports.s3.yandex.net/19237/4ee96869a2500fa3e75575c767631aa9fbfef732/functional_stateless_tests_(debug).html#fail1.
https://github.com/ClickHouse/ClickHouse/pull/19237 | 2021-01-19T07:22:33 | 2021-01-19T07:22:33 | {} | MEMBER |
ClickHouse | ClickHouse | 762,800,634 | 19,262 | alexey-milovidov | Self-signed sertificates are supported but pay attention to configuration here:
```
<invalidCertificateHandler>
<!-- Use for self-signed: <name>AcceptCertificateHandler</name> -->
<name>RejectCertificateHandler</name>
</invalidCertificateHandler>
``` | 2021-01-19T12:07:22 | 2021-01-19T12:12:10 | {} | MEMBER |
ClickHouse | ClickHouse | 762,844,825 | 19,278 | den-crane | >What are the reasons for this design?
Basically: it is what it is
https://github.com/ClickHouse/ClickHouse/issues/12489#issuecomment-729825118 | 2021-01-19T13:39:24 | 2021-01-19T13:40:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,930,023 | 19,280 | den-crane | that sentence is about result, not about performance. | 2021-01-19T15:47:14 | 2021-01-19T15:47:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 763,031,787 | 19,288 | UnamedRus | If left table would be the clickhouse table, everything is working.
```
SELECT *
FROM system.one
, mysql.members AS m
INNER JOIN mysql.committees AS c ON m.name = c.name
WHERE (m.name = 'John') AND (committee_id = 1)
Query id: 7a31a655-d240-457d-a613-86c214c7553f
┌─one.dummy─┬─m.member_id─┬─m.name─┬─c.com... | 2021-01-19T18:22:10 | 2021-01-19T18:24:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,767,904 | 12,390 | alexey-milovidov | @etah000 Sorry for late answer:
The setting should not modify parsing behaviour.
The parser should be able to parse simple UPDATE/DELETE queries unconditionally into separate AST.
If the setting is enabled, this AST will be converted to AST for ALTER UPDATE/DELETE and then query will be processed as usual. We ca... | 2021-01-19T11:00:03 | 2021-01-19T11:00:03 | {} | MEMBER |
ClickHouse | ClickHouse | 762,849,111 | 19,280 | den-crane | yes, because optimize_move_to_prewhere does not work against view
related https://github.com/ClickHouse/ClickHouse/issues/18561 | 2021-01-19T13:47:14 | 2021-01-19T13:47:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,876,521 | 19,259 | den-crane | typo:
`avg_ `
`agv_ ` | 2021-01-19T14:30:02 | 2021-01-19T14:30:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,648,701 | 19,260 | sundy-li | See `bitmapContains` | 2021-01-19T07:11:35 | 2021-01-19T07:11:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,670,723 | 19,028 | KochetovNicolai | For `/data/compile/ClickHouse/contrib/croaring/include/roaring/roaring.h:13:10: fatal error: 'roaring/roaring_version.h' file not found` error try to run `git submodule update --init --recursive`
For `Cannot enable LDAP support` error try to add `-DENABLE_LDAP=0` | 2021-01-19T08:02:08 | 2021-01-19T08:02:08 | {} | MEMBER |
ClickHouse | ClickHouse | 762,781,772 | 19,274 | alexey-milovidov | It's most likely will be a separate task as the implementation is different. | 2021-01-19T11:26:59 | 2021-01-19T11:26:59 | {} | MEMBER |
ClickHouse | ClickHouse | 762,826,952 | 18,554 | tisonkun | After this commit darwin build failed. See also #19277 | 2021-01-19T13:04:11 | 2021-01-19T13:04:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,840,220 | 19,278 | dgzdot | > CH uses default database in this case by design.
>
> #12489
> #15898
> #10471
>
> you need to specify database: `dgz.test_query_local`
>
> SELECT uniq(id) FROM test_query_all WHERE repo = '100' AND id IN (SELECT id FROM dgz.test_query_local WHERE repo = '200');
What are the reasons for this design? In ... | 2021-01-19T13:30:41 | 2021-01-19T13:30:41 | {} | NONE |
ClickHouse | ClickHouse | 763,018,062 | 19,022 | akuzm | Fuzzer https://github.com/ClickHouse/ClickHouse/issues/19287 | 2021-01-19T18:00:15 | 2021-01-19T18:00:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 763,309,736 | 19,277 | tisonkun | @alexey-milovidov what do you think about adding conditional compilation in libpq contrib as mentioned above? It should give a perfect build, too. But I'm not sure how to update a submodule and create a pull request now. Maybe try it later. | 2021-01-20T03:46:40 | 2021-01-20T03:46:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 763,397,257 | 19,284 | alesapin | No related failures, but I need checks with thread sanitizer. Will run them manually. | 2021-01-20T07:26:57 | 2021-01-20T07:26:57 | {} | MEMBER |
ClickHouse | ClickHouse | 763,422,021 | 17,623 | mathfool | Any plan to refactor the query parsers especially when including joins and with column aliaes. It seems there are quite some issues reported related to this part. May be worth doing something to improve the underlying model. | 2021-01-20T08:16:36 | 2021-01-20T08:16:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,604,346 | 17,623 | filimonov | > Hello @alex-zaitsev / @myrrc, is there any information available in public?
Not yet, but probably we should/will publish RFC soon.
> Does the lightweight update/delete execute asynchronously like before? Will you respect mutations_async setting?
No and no (but eventual replication is still there, and other ... | 2021-01-19T05:07:30 | 2021-01-19T05:07:53 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,778,128 | 19,274 | zhicwu | Can we also return updated/deleted rows? | 2021-01-19T11:19:21 | 2021-01-19T11:19:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,870,470 | 10,177 | erste | Looks like renaming **move ttl** to just **ttl** in zookeeper metadata fixes the problem:
```
[zk: localhost:2181(CONNECTED) 16] get /clickhouse/tables/database/1/table_rpl/metadata
metadata format version: 1
date column:
sampling expression:
index granularity: 8192
mode: 0
sign column:
primary key: mdate
dat... | 2021-01-19T14:21:58 | 2021-01-19T14:21:58 | {} | NONE |
ClickHouse | ClickHouse | 762,894,899 | 19,282 | dmftaras | @den-crane thanks! | 2021-01-19T14:57:00 | 2021-01-19T14:57:00 | {} | NONE |
ClickHouse | ClickHouse | 763,049,056 | 19,283 | alexey-milovidov | It is only related to the behaviour of interactive clickhouse-client, not to the format itself: the line was cleared to render progress.
Hence minor priority. | 2021-01-19T18:50:38 | 2021-01-19T18:50:38 | {} | MEMBER |
ClickHouse | ClickHouse | 763,065,971 | 19,288 | alexey-milovidov | That's true. When preparing a query to be sent to external database, it does not analyze for what tables the columns from WHERE belong. | 2021-01-19T19:10:40 | 2021-01-19T19:10:40 | {} | MEMBER |
ClickHouse | ClickHouse | 762,741,881 | 19,260 | l1t1 | thanks. @sundy-li | 2021-01-19T10:11:22 | 2021-01-19T10:11:22 | {} | NONE |
ClickHouse | ClickHouse | 762,800,136 | 19,255 | ivanprostran | Hi,
Confirmed - latest version works OK.
We have not found any issues so far.
` (version 21.1.2.15 (official build))`
However migration from "20.8.3.18-1" to "21.1.2.15" needed some minor changes:
Default value (uuid field type) toUUID(0) was raising an error.
toUUID('00000000-0000-0000-0000-000000000000')... | 2021-01-19T12:06:14 | 2021-01-19T12:06:14 | {
"+1": 1,
"hooray": 1
} | NONE |
ClickHouse | ClickHouse | 762,827,651 | 19,272 | den-crane | alphabetical, until parallel_view_processing is enabled
--parallel_view_processing arg Enables pushing to attached views concurrently instead of sequentially.
https://den-crane.github.io/Everything_you_should_know_about_materialized_views_commented.pdf
https://youtu.be/ckChUkC3P... | 2021-01-19T13:05:39 | 2021-01-19T13:05:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,879,896 | 19,277 | alexey-milovidov | @TisonKun it builds perfectly fine in our CI, see the https://clickhouse-builds.s3.yandex.net/0/ee4cc94e6c47a48f491c34f4844345a16a37379a/clickhouse_special_build_check/report.html
(clang-darwin)
You can download this build and run on your machine. | 2021-01-19T14:35:15 | 2021-01-19T14:35:15 | {} | MEMBER |
ClickHouse | ClickHouse | 762,896,827 | 18,495 | kssenii | Taking this task. | 2021-01-19T14:59:50 | 2021-01-19T14:59:50 | {} | MEMBER |
ClickHouse | ClickHouse | 763,014,861 | 19,229 | chiragguptadtu | >Can you confirm that clickhouse is actually uses the all available threads (within limit of max_threads) during query execution with top/htop?
<img width="1422" alt="Screen Shot 2021-01-19 at 11 19 09 PM" src="https://user-images.githubusercontent.com/20680224/105073603-477abc00-5aad-11eb-88d5-e20cc926f350.png">
<... | 2021-01-19T17:54:51 | 2021-01-19T17:54:51 | {} | NONE |
ClickHouse | ClickHouse | 763,066,644 | 19,288 | alexey-milovidov | Maybe it's better to not push down any conditions in presense of JOIN. | 2021-01-19T19:11:49 | 2021-01-19T19:11:49 | {} | MEMBER |
ClickHouse | ClickHouse | 763,091,026 | 19,290 | azat | >and I don't see how ATTEMPT_TO_READ_AFTER_EOF can be received while reading local file
Ok, I see them now (tons of `throwReadAfterEOF` in `ReadHelpers.h` and `ReadBuffer.h`), will rewrite the patch a little | 2021-01-19T19:56:51 | 2021-01-19T19:56:51 | {} | MEMBER |
ClickHouse | ClickHouse | 763,157,656 | 17,466 | CurtizJ | Fixed in #18188 #18516. | 2021-01-19T21:38:02 | 2021-01-19T21:38:02 | {} | MEMBER |
ClickHouse | ClickHouse | 763,397,937 | 9,887 | UnamedRus | @ucasFL
It's more about not using keyword INTERVAL in queries.
> Interval syntax like '+ 30 days' is not supported
```
postgres=# SELECT now() + '30 day';
?column?
-------------------------------
2021-02-19 04:01:01.766836+03
(1 row)
postgres=# SELECT now() + '30 days';
?column?
-... | 2021-01-20T07:28:33 | 2021-01-20T07:28:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,650,953 | 19,250 | alesapin | ??? | 2021-01-19T07:16:54 | 2021-01-19T07:16:54 | {} | MEMBER |
ClickHouse | ClickHouse | 763,017,873 | 19,287 | akuzm | https://clickhouse-test-reports.s3.yandex.net/19022/d0922e2985516fa230b7537e4daadde0101b78da/fuzzer_ubsan/report.html#fail1 | 2021-01-19T17:59:56 | 2021-01-19T17:59:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 763,187,542 | 15,938 | den-crane | hm, I don't see much improvement:
3.611s VS 2.063s.
```sql
CREATE TABLE T (p Int64, d Date, some_col String)
Engine=ReplacingMergeTree
Partition by toYYYYMM(d)
Order by (p, d);
insert into T select number, '2020-01-01', '' from numbers(100000000);
insert into T select number, '2020-01-01', 'updated' fr... | 2021-01-19T22:36:01 | 2021-01-19T22:37:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 763,271,493 | 18,720 | hexiaoting | @alexey-milovidov I need some help, related to above questions. Thank you. | 2021-01-20T01:49:20 | 2021-01-20T01:49:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,831,201 | 19,272 | den-crane | >output will be following, randomly..
it's probably because part for b is not created yet and not visible for select. | 2021-01-19T13:13:22 | 2021-01-19T13:13:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,850,827 | 19,225 | CurtizJ | Please, can you show your table schema? | 2021-01-19T13:49:54 | 2021-01-19T13:49:54 | {} | MEMBER |
ClickHouse | ClickHouse | 762,868,950 | 19,277 | tisonkun | I set `HAVE_DECL_STRLCAT` & `HAVE_DECL_STRLCPY` in `contrib/libpq/include/port.h` to `1` and comment out `contrib/libpq/port/strlcpy.c` and it seems works.
The problem is symbol `strlcpy` & `strlcat` conflict with darwin builtin headers but I don't have idea to make it portable & compatible. | 2021-01-19T14:19:40 | 2021-01-19T14:19:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,907,058 | 19,280 | den-crane | >So this is inconsistent with the actual situation described in the document.
What document?
>Is there any plan to optimize this
Yes, someday. | 2021-01-19T15:14:23 | 2021-01-19T15:14:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 763,225,997 | 19,229 | alexey-milovidov | > Maybe you 28 threads reached the maximum possible IO speed? So your disk can not deliver data fast enough to fully load > 30 threads?
Let's look at the ProfileEvents - they will tell what is the bottleneck. | 2021-01-20T00:00:38 | 2021-01-20T00:00:38 | {} | MEMBER |
ClickHouse | ClickHouse | 763,250,635 | 19,260 | l1t1 | > See `bitmapContains`
the follwing query is not fast
```
SELECT sum(sal)
FROM men
WHERE bitmapContains(
(
SELECT m_bit
FROM men_bit
WHERE name = '0'
), toUInt32(id)) = 1
Query id: e4a161e3-1f4f-40e4-8fe3-5108c521f542
┌────sum(sal)─┐
│ 27000000000 │
└─────────────┘
1 rows in set. Ela... | 2021-01-20T00:59:50 | 2021-01-20T00:59:50 | {} | NONE |
ClickHouse | ClickHouse | 763,274,990 | 19,272 | slimtom95 | > ```
> 18:53:13.602554 [ 4700 ] <Debug> executeQuery: (from [::1]:33126, using production parser) insert into a (id) values
> 18:53:13.603055 [ 4700 ] <Debug> HashJoin: Right sample block: b.value String String(size = 0), b.id UInt64 UInt64(size = 0)
> 18:53:13.604774 [ 4700 ] <Trace> default.a Renaming tempora... | 2021-01-20T02:00:27 | 2021-01-20T02:00:27 | {} | NONE |
ClickHouse | ClickHouse | 763,387,359 | 18,824 | tangxueming | It works well when I set timestamp_hour as primary key . Thanks!
But it's slowly when I execute ' select xxx from tableA order by timestamp_hour, timestamp desc limit 10' . It is so slowly that I don't think I will use this method.
In my case, I will execute two queries, one is ' select xxx from tableA order by time... | 2021-01-20T07:05:52 | 2021-01-20T07:05:52 | {} | NONE |
ClickHouse | ClickHouse | 762,633,547 | 16,925 | vvhungy | @den-crane oh, I've never known as there're no update on this issue thread. We just updated our CH server to 20.12.5.14 and it worked. Tks. | 2021-01-19T06:35:31 | 2021-01-19T06:35:31 | {} | NONE |
ClickHouse | ClickHouse | 762,700,433 | 15,929 | UnamedRus | Fixed with https://github.com/ClickHouse/ClickHouse/pull/19221
| 2021-01-19T08:59:47 | 2021-01-19T08:59:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,752,689 | 18,671 | alexey-milovidov | > Cannot reserve 1.00 MiB, not enough space.
Not enough disk space in CI. | 2021-01-19T10:31:12 | 2021-01-19T10:31:12 | {} | MEMBER |
ClickHouse | ClickHouse | 762,949,731 | 19,077 | qoega | only docker build should be tested in this PR. | 2021-01-19T16:14:57 | 2021-01-19T16:14:57 | {} | MEMBER |
ClickHouse | ClickHouse | 763,054,999 | 19,277 | alexey-milovidov | This is the source from master. The only difference is that it is not compiled on MacOS. It is cross-compiled for MacOS on Linux. | 2021-01-19T19:01:02 | 2021-01-19T19:01:02 | {} | MEMBER |
ClickHouse | ClickHouse | 763,369,267 | 19,295 | mathfool | 请提供些详细信息。please provide more details. | 2021-01-20T06:26:40 | 2021-01-20T06:26:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 763,479,866 | 19,273 | KochetovNicolai | #19303 | 2021-01-20T09:48:50 | 2021-01-20T09:48:50 | {} | MEMBER |
ClickHouse | ClickHouse | 762,596,622 | 19,255 | filimonov | > The packages in our repository and in Altinity repository are bitwise identical - the Altinity repo is just a mirror
That is NOT true currently - Altinity rebuild the binary from sources and uses a packaging which is out of sync with official.
So the code is exactly the same, but the package/binary is not byte... | 2021-01-19T04:39:35 | 2021-01-19T04:43:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,597,824 | 19,255 | filimonov | I think it happens on empty message.
UPD. no, it's checked by tests: https://github.com/ClickHouse/ClickHouse/blob/b8fcd75de75c908a422b221a6d10c30b53fe8aa4/tests/integration/test_storage_kafka/test.py#L193 | 2021-01-19T04:44:25 | 2021-01-19T04:49:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,607,873 | 15,165 | filimonov | Can you recheck it with `<secret>` in cluster config? See https://github.com/ClickHouse/ClickHouse/pull/13156 | 2021-01-19T05:19:57 | 2021-01-19T05:20:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,652,193 | 19,244 | alesapin | > Ok but I don't understand:
>
> * is epoll not instrumented by MSan?
>
> * why don't we have this issue for other usages of epoll?
>
> * or does `libuv` performing epoll in some unusual way?
Seems like `epoll` not instrumented by MSan but almost always we initialize events array before `epoll` ... | 2021-01-19T07:19:32 | 2021-01-19T07:19:32 | {} | MEMBER |
ClickHouse | ClickHouse | 762,793,547 | 19,205 | tavplubix | AST fuzzer (UBSan) - #19276 | 2021-01-19T11:52:15 | 2021-01-19T11:52:15 | {} | MEMBER |
ClickHouse | ClickHouse | 762,938,093 | 19,250 | alesapin | So everything works well? Ok, I'll merge. | 2021-01-19T15:58:22 | 2021-01-19T15:58:22 | {} | MEMBER |
ClickHouse | ClickHouse | 762,932,789 | 19,225 | matterumirai | @CurtizJ
Clickhouse works properly after around two hours while the error continues. (it downs again after a while)
i queried my table with just simple "select col12 from dtTable" and following error has occured:
`Cannot seek through file /var/lib/clickhouse/data/default/dtTable/20201127_24324_24324_0/col12.bin... | 2021-01-19T15:50:45 | 2021-01-20T07:05:44 | {} | NONE |
ClickHouse | ClickHouse | 762,914,625 | 19,280 | IVitamin-C | 
| 2021-01-19T15:25:02 | 2021-01-19T15:25:02 | {} | NONE |
ClickHouse | ClickHouse | 763,412,955 | 19,296 | mathfool | sounds more like a timeout issue. did you try to increase the query timeout? or maybe other timeout settings. | 2021-01-20T07:59:38 | 2021-01-20T07:59:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,869,002 | 19,282 | den-crane | 1. AggregatingMergeTree require some AggregateFunction or SimpleAggregateFunction. You need to use sumState / mergeState https://gist.github.com/den-crane/6eff375752a236a456e1b3dc2ca7db62 or https://gist.github.com/den-crane/e326eef3b6ee6ccd5ca0df1dc1849a05
in your case `total` is calculated by ANY()
2. Or you can... | 2021-01-19T14:19:44 | 2021-01-19T14:19:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,899,304 | 19,272 | slimtom95 | OK, let's take a closer look at this..
`set send_logs_level='trace'`
```
INSERT INTO a (id) VALUES
Query id: 3af8a964-e456-4d8e-b670-88c052b0b717
[C02CW714MD6T] 2021.01.19 22:58:51.504988 [ 1106729 ] {3af8a964-e456-4d8e-b670-88c052b0b717} <Debug> executeQuery: (from [::1]:49653, using production parser) inse... | 2021-01-19T15:03:10 | 2021-01-19T15:03:10 | {} | NONE |
ClickHouse | ClickHouse | 762,982,092 | 19,283 | filimonov | Reproduces | 2021-01-19T17:03:46 | 2021-01-19T17:03:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 763,473,042 | 19,260 | l1t1 | > > See `bitmapContains`
>
> the follwing query is not fast
>
> ```
> SELECT sum(sal)
> FROM men
> WHERE bitmapContains(
> (
> SELECT m_bit
> FROM men_bit
> WHERE name = '0'
> ), toUInt32(id)) = 1
>
> Query id: e4a161e3-1f4f-40e4-8fe3-5108c521f542
>
> ┌────sum(sal)─┐
> │ 27000000000 │
... | 2021-01-20T09:39:09 | 2021-01-20T09:39:09 | {} | NONE |
ClickHouse | ClickHouse | 762,711,613 | 19,255 | ivanprostran | > Thank you for the suggestions.
> Upgraded to newer version (for the moment without segmentation fault)
>
> Now we have problem with a read only mode (all tables)
>
> ```
> 2021.01.19 09:55:23.538606 [ 30176 ] {} <Error> void DB::StorageKafka::threadFunc(size_t): Code: 242, e.displayText() = DB::Exception: Tab... | 2021-01-19T09:19:37 | 2021-01-19T09:19:37 | {
"heart": 1
} | NONE |
ClickHouse | ClickHouse | 762,713,691 | 18,812 | vdimir | > @vdimir Yes, that's in order to allow more granular settings for different URLs. This is particularly important when user specifies path-style URLs.
Could you add exmple of usage for exmple in [test_storage_s3](https://github.com/ClickHouse/ClickHouse/tree/70d9c7ea0317edfdfaa0b6bf94c4028d328aa328/tests/integration... | 2021-01-19T09:23:02 | 2021-01-19T09:23:18 | {} | MEMBER |
ClickHouse | ClickHouse | 762,834,689 | 19,278 | den-crane | CH uses default database in this case by design.
https://github.com/ClickHouse/ClickHouse/issues/12489
https://github.com/ClickHouse/ClickHouse/issues/15898
https://github.com/ClickHouse/ClickHouse/issues/10471
you need to specify database: `dgz.test_query_local`
SELECT uniq(id) FROM test_query_all WHERE rep... | 2021-01-19T13:20:12 | 2021-01-19T13:20:12 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,835,690 | 19,274 | den-crane | related https://github.com/ClickHouse/ClickHouse/issues/18237 https://github.com/ClickHouse/ClickHouse/issues/16207 | 2021-01-19T13:22:04 | 2021-01-19T13:22:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,896,589 | 19,280 | IVitamin-C | So this is inconsistent with the actual situation described in the document. Is there any plan to optimize this | 2021-01-19T14:59:29 | 2021-01-19T14:59:29 | {} | NONE |
ClickHouse | ClickHouse | 762,900,747 | 19,229 | chiragguptadtu | @UnamedRus Any idea ? | 2021-01-19T15:05:17 | 2021-01-19T15:05:17 | {} | NONE |
ClickHouse | ClickHouse | 762,916,742 | 19,222 | amosbird | > @amosbird Will the pull request #18684 fix this alias issue?
It will. | 2021-01-19T15:28:14 | 2021-01-19T15:28:14 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 763,257,410 | 19,260 | l1t1 | if the men table is 1000 rows and men_bit is big, it also cost a large amout of memory
```
:) create table men_1k engine=MergeTree order by id as select * from men where id<1000;
0 rows in set. Elapsed: 0.030 sec. Processed 8.19 thousand rows, 65.54 KB (272.47 thousand rows/s., 2.18 MB/s.)
SELECT
name,
... | 2021-01-20T01:17:49 | 2021-01-20T01:17:49 | {} | NONE |
ClickHouse | ClickHouse | 762,701,621 | 19,255 | ivanprostran | Thank you for the suggestions.
Upgraded to newer version (for the moment without segmentation fault)
Now we have problem with a read only mode (all tables)
```
2021.01.19 09:55:23.538606 [ 30176 ] {} <Error> void DB::StorageKafka::threadFunc(size_t): Code: 242, e.displayText() = DB::Exception: Table is in read... | 2021-01-19T09:01:59 | 2021-01-19T09:01:59 | {} | NONE |
ClickHouse | ClickHouse | 762,757,870 | 14,196 | CurtizJ | Several cases were implemented in #17310. | 2021-01-19T10:41:04 | 2021-01-19T10:41:04 | {
"hooray": 1
} | MEMBER |
ClickHouse | ClickHouse | 762,877,155 | 19,277 | tisonkun | My rough idea is apply changes above guarding by `if OS_DARWIN` command, but I'm unfamiliar with submodule project and don't know the workflow in order to create a real pull request. | 2021-01-19T14:31:04 | 2021-01-19T14:31:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,919,547 | 19,262 | ralphM78 | @alexey-milovidov many thanks for putting me into the right direction. I had to adjust keyUsage in certificates. | 2021-01-19T15:31:53 | 2021-01-19T15:31:53 | {
"hooray": 1
} | NONE |
ClickHouse | ClickHouse | 762,934,965 | 19,229 | UnamedRus | There is no big difference in metrics as i see. (except of time, ofc)
Can you confirm that clickhouse is actually uses the all available threads (within limit of ```max_threads```) during query execution with top/htop?
What it you set ```max_threads=32``` ? | 2021-01-19T15:53:57 | 2021-01-19T15:53:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,957,084 | 19,251 | tavplubix | Probably related: #18063 | 2021-01-19T16:25:54 | 2021-01-19T16:25:54 | {} | MEMBER |
ClickHouse | ClickHouse | 763,045,052 | 13,194 | den-crane | @phamtai97
CH expects 1576225796.097 as Decimal because DateTime64 based on Decimal
INSERT INTO log VALUES (1, 1576225796.097); | 2021-01-19T18:43:43 | 2021-01-19T18:43:43 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 763,056,234 | 19,277 | alexey-milovidov | As a workaround you can simply disable Postgres in CMake:
```
cd build
cmake -D ENABLE_LIBPQXX=0 ..
``` | 2021-01-19T19:03:07 | 2021-01-19T19:03:07 | {} | MEMBER |
ClickHouse | ClickHouse | 763,236,279 | 19,273 | alexey-milovidov | @KochetovNicolai Please take a look on the fuzzer findings. | 2021-01-20T00:19:34 | 2021-01-20T00:19:34 | {} | MEMBER |
ClickHouse | ClickHouse | 762,612,137 | 19,260 | l1t1 | if i reduce the bitmap, it works
```
INSERT INTO men_bit SELECT
'0',
groupBitmapState(id)
FROM men
WHERE (age >= 0) AND (age <= 0)
0 rows in set. Elapsed: 0.140 sec. Processed 100.00 million rows, 499.92 MB (712.75 million rows/s., 3.56 GB/s.)
SELECT sum(sal)
FROM men
WHERE id IN bitmapToArray(
... | 2021-01-19T05:33:48 | 2021-01-19T05:33:48 | {} | NONE |
ClickHouse | ClickHouse | 762,653,663 | 19,237 | alesapin | `01442_merge_detach_attach` -- the table in the test even not replicated, so cannot be affected by these changes. Reported in CI fest. | 2021-01-19T07:22:47 | 2021-01-19T07:22:47 | {} | MEMBER |
ClickHouse | ClickHouse | 763,225,088 | 19,229 | alexey-milovidov | @filimonov
> you reach the maximum possible speed of memcpy (typically it's something around 20-30 Gbyte/sec).
This is wrong. | 2021-01-20T00:00:04 | 2021-01-20T00:00:04 | {} | MEMBER |
ClickHouse | ClickHouse | 763,228,703 | 19,229 | alexey-milovidov | https://presentations.clickhouse.tech/dataops_2019/# | 2021-01-20T00:04:44 | 2021-01-20T00:04:44 | {} | MEMBER |
ClickHouse | ClickHouse | 763,371,442 | 19,295 | SDUERLLei | 我用的是mycat的分库分表，然后想要测试同步mycat的数据到clickhouse，使用的命令是：insert into table select * from mysql('ip:8066','db','table','user','password'),其中8066为mycat端口，然后同一条命令有时可以查出结果，有时则报表不存在 | 2021-01-20T06:31:38 | 2021-01-20T06:31:38 | {} | NONE |
ClickHouse | ClickHouse | 742,640,009 | 17,944 | tavplubix | AST fuzzer - #15889 | 2020-12-10T16:42:28 | 2020-12-10T16:42:28 | {} | MEMBER |
ClickHouse | ClickHouse | 742,730,137 | 17,960 | den-crane | yes
```sql
set optimize_redundant_functions_in_order_by=0;
SELECT d, SUM(am) / dictGet('dw.currency', 'divider', toUInt64(1)) AS a
FROM (select toUInt32(1) d,'1' f ,'2' t,3e21 am union all select 1,'1','3',1e18) GROUP BY d,f,t ORDER BY d , a
┌─d─┬────a─┐
│ 1 │ 1 │
│ 1 │ 3000 │
└───┴──────┘
se... | 2020-12-10T19:04:24 | 2020-12-10T19:04:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 742,831,549 | 17,929 | alexey-milovidov | Bad test. | 2020-12-10T22:10:02 | 2020-12-10T22:10:02 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.