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 | 728,765,951 | 17,095 | KochetovNicolai | Yandex check is ok now.
Not waiting for the others. | 2020-11-17T08:17:52 | 2020-11-17T08:18:16 | {} | MEMBER |
ClickHouse | ClickHouse | 728,824,521 | 16,532 | atercattus | We have a similar problem with ReplicatedAggregatingMergeTree table with CH version v20.11.2.1-stable (and at earlier version too).
Neither CHECK TABLE nor DETACH/ATTACH helps. There are no frozen parts, active merges, replication queue.
The only solution: copy the whole partition to the temporary table, drop the... | 2020-11-17T10:04:41 | 2020-11-17T10:04:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,641,590 | 17,045 | l1t1 | I didn't modify the config and use ln -s | 2020-11-15T21:40:58 | 2020-11-15T21:40:58 | {} | NONE |
ClickHouse | ClickHouse | 727,721,054 | 17,050 | hexiaoting | Can the error be reproduced using Clickhouse client interactive mode? If we run the detail sql queries in 01281_join_with_prewhere_fix.sql multiply times? | 2020-11-16T04:00:02 | 2020-11-16T04:05:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,922,621 | 16,986 | nikitamikhaylov | Yandex synchronization and third-party checks are broken in master. Other seems unrelated. | 2020-11-16T11:33:25 | 2020-11-16T11:33:25 | {} | MEMBER |
ClickHouse | ClickHouse | 728,156,676 | 17,058 | den-crane | MV does not support UNION ALL. You can create 2 MV and write their select results into one destination table using TO.
| 2020-11-16T16:03:41 | 2020-11-16T16:03:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,509,416 | 17,054 | den-crane | ```
SET send_logs_level = 'debug'
1 insert : MemoryTracker: Peak memory usage (for query): 4.93 GiB.
2 insert : MemoryTracker: Peak memory usage (for query): 4.02 GiB.
set max_memory_usage='40G', max_bytes_before_external_group_by=0
1 insert : MemoryTracker: Peak memory usage (for query): 5.06 GiB.
2 insert... | 2020-11-17T00:29:44 | 2020-11-17T00:29:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,648,058 | 16,991 | den-crane | Maybe it's not major after all. Because Engine JOIN has supported only join on UInt64 keys before CH 20th something. | 2020-11-15T22:27:46 | 2020-11-15T22:28:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,727,662 | 17,053 | den-crane | for example:
```
create materialized view mymv Engine=MergeTree partition by .... order by ....
settings index_granularity = 8192,
storage_policy = 'default'
as select * from ....
```
https://den-crane.github.io/Everything_you_should_know_about_materialized_views_commented.pdf
https://youtu.be... | 2020-11-16T04:28:44 | 2020-11-16T04:28:44 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,748,715 | 16,184 | hczhcz | @Avogar Should we apply this setting to `groupArray*` functions? | 2020-11-16T05:43:23 | 2020-11-16T05:43:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,376,184 | 17,054 | l1t1 | thanks, and I wonder why those two inserts need different memory size? | 2020-11-16T22:47:14 | 2020-11-16T22:47:14 | {} | NONE |
ClickHouse | ClickHouse | 727,640,503 | 15,718 | alexey-milovidov | We are going to enable Query Fuzzer with ASan, MSan, TSan, but after all existing issues will be fixed.
https://github.com/ClickHouse/ClickHouse/issues?q=is%3Aissue+is%3Aopen+label%3Afuzz
Only a few remain... | 2020-11-15T21:32:48 | 2020-11-15T21:33:31 | {} | MEMBER |
ClickHouse | ClickHouse | 727,647,654 | 17,045 | den-crane | check stdout for
```
sudo -u clickhouse clickhouse-server -pid-file=/var/run/clickhouse-server/clickhouse-server.pid --config-file=/etc/clickhouse-server/config.xml
```
it could be SELINUX. | 2020-11-15T22:24:59 | 2020-11-15T22:24:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,720,169 | 17,049 | hexiaoting | Have you stopped the old clickhouse-server?
and try to check : the clickhouse server installed by clickhouse-server-20.11.3.3-2.noarch.rpm that really started with port 9000. | 2020-11-16T03:55:49 | 2020-11-16T03:56:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,265,176 | 17,006 | azat | @KochetovNicolai according to git blame you are the original author, can you please take a look? | 2020-11-16T19:09:26 | 2020-11-16T19:09:26 | {} | MEMBER |
ClickHouse | ClickHouse | 728,730,832 | 17,082 | bharatnc | @alesapin I think I can take a stab at this - can I have this assigned to me? | 2020-11-17T07:01:23 | 2020-11-17T07:01:23 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,794,657 | 17,088 | amosbird | > Are you sure it is 20.8.1.1
Well it's a homemade one. | 2020-11-17T09:12:51 | 2020-11-17T09:12:51 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 728,907,476 | 17,115 | filimonov | It's a part of https://github.com/ClickHouse/ClickHouse/issues/9336 /cc @BohuTANG
I would expect that it can be directly transformed to
```
set max_result_rows=123,result_overflow_mode='break';
```
But for some reason, it doesn't work properly...
```
select * from system.numbers;
--- 65505 rows in se... | 2020-11-17T12:52:36 | 2020-11-17T12:52:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,706,067 | 16,979 | lddlww | sorry for late reply；
[msg.bin.tar.gz](https://github.com/ClickHouse/ClickHouse/files/5543818/msg.bin.tar.gz)
| 2020-11-16T02:59:15 | 2020-11-16T02:59:15 | {} | NONE |
ClickHouse | ClickHouse | 727,719,182 | 17,050 | hexiaoting | Does the server crash? | 2020-11-16T03:51:44 | 2020-11-16T03:51:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,728,376 | 17,045 | den-crane | I've missed one dash -
```
sudo -u clickhouse clickhouse-server --pid-file=/var/run/clickhouse-server/clickhouse-server.pid --config-file=/etc/clickhouse-server/config.xml
``` | 2020-11-16T04:31:51 | 2020-11-16T04:31:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,758,357 | 17,055 | l1t1 | btw, the i5 has 12 cores, amd vm has 16 cores | 2020-11-16T06:10:18 | 2020-11-16T06:10:18 | {} | NONE |
ClickHouse | ClickHouse | 727,799,516 | 17,055 | filimonov | Maybe something is wrong with the storage subsystem on your VM.
Try to repeat your queries with
```
SET send_logs_level='trace'
```
and compare logs.
and also check the numbers in system.query_log (ProfileEvents columns) and compare numbers there. | 2020-11-16T07:46:05 | 2020-11-16T07:46:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,873,065 | 17,046 | alesapin | Suspicious failure https://gist.github.com/alesapin/accdb5e1ff470dcbd47f7fb0ae39efd4 | 2020-11-16T10:01:28 | 2020-11-16T10:01:28 | {} | MEMBER |
ClickHouse | ClickHouse | 728,407,147 | 17,054 | den-crane | Which 2 ? | 2020-11-16T23:32:26 | 2020-11-16T23:32:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,645,251 | 17,050 | hexiaoting | @mateng0915 clickhouse-test is a python file, you can debug clickhouse-test using 'ipdb'. | 2020-11-17T02:27:25 | 2020-11-17T02:27:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,755,378 | 17,052 | MrSsunlight | > Does it reproduce with non-replicated tables?
>
> Check this https://clickhouse.tech/docs/en/operations/settings/settings/#settings-deduplicate-blocks-in-dependent-materialized-views
Default value used for this configuration，The SQL I created acquaintance has not reproduced, but the above process still has prob... | 2020-11-17T08:00:22 | 2020-11-17T08:00:22 | {} | NONE |
ClickHouse | ClickHouse | 728,798,328 | 17,088 | thiningsun | > Well it's a homemade one.
You are too strong, the companies that can re-develop are very powerful。Have you tested other versions？ | 2020-11-17T09:19:15 | 2020-11-17T09:19:15 | {} | NONE |
ClickHouse | ClickHouse | 727,643,078 | 16,987 | filimonov | @azat it's very similar to the issue we discussed recently. | 2020-11-15T21:52:31 | 2020-11-15T21:52:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,719,689 | 17,050 | mateng0915 | > Does the server crash?
No. when facing this problem the server still running successfully. And maybe i executing this test case then i can pass, after receive the fail error.
| 2020-11-16T03:53:42 | 2020-11-16T03:53:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,837,057 | 17,050 | mateng0915 | > Can the error be reproduced using Clickhouse client interactive mode? If we run the detail sql queries in 01281_join_with_prewhere_fix.sql multiply times?
it will not reproduce this error if i using Clickhouse client interactive mode. I think maybe this problems comes from clickhouse-test. Dose any tool can help m... | 2020-11-16T08:58:42 | 2020-11-16T08:58:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,877,226 | 17,070 | fastio | @Akazz I create a new pr, and close the old one #15087
Thanks for your review. | 2020-11-16T10:08:55 | 2020-11-16T10:08:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,642,932 | 17,108 | alexey-milovidov | Arcadia Ok. | 2020-11-17T02:19:43 | 2020-11-17T02:19:43 | {} | MEMBER |
ClickHouse | ClickHouse | 728,819,805 | 17,081 | gmxyb | > my test is ok. can create table and insert data.
> btw, I am using wsl 2
O... I'm ws1, thank you ! | 2020-11-17T09:57:18 | 2020-11-17T09:57:18 | {} | NONE |
ClickHouse | ClickHouse | 728,882,752 | 17,128 | filimonov | Duplicate of https://github.com/ClickHouse/ClickHouse/issues/17088
It's a bug. We will fix it in next releases. Please avoid sorting the constants till the fix will pop up.
As another alternative workaround you can wrap your constant in the `materialize()`
I.e.
```
select materialize('1111') as name from n... | 2020-11-17T12:03:56 | 2020-11-17T12:03:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,887,418 | 17,128 | thiningsun | > materialize
ok，I test the method is PASS ！thanks for your help ，GOOD | 2020-11-17T12:13:15 | 2020-11-17T12:13:15 | {} | NONE |
ClickHouse | ClickHouse | 727,675,953 | 16,806 | TCeason | Oh I will add other case and I forget to modify it.
---Original---
From: "tavplubix"<notifications@github.com>
Date: Mon, Nov 16, 2020 04:27 AM
To: "ClickHouse/ClickHouse"<ClickHouse@noreply.github.com>;
Cc: "Author"<author@noreply.github.com>;"TCeason"<tai_chong@foxmail.com>;
Subject: Re: [Clic... | 2020-11-16T01:05:13 | 2020-11-16T01:05:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,725,592 | 11,643 | ghost | @amosbird
Sorry for the delay.
It is also possible to use subqueries.
```SQL
SELECT *
FROM
(
SELECT
d,
dt,
id,
groupArray(v) AS g,
anyLast(v) AS l
FROM wl1
GROUP BY
d,
dt,
id
ORDER BY
d ASC,
dt ... | 2020-11-16T04:19:54 | 2020-11-16T07:10:28 | {} | NONE |
ClickHouse | ClickHouse | 728,081,627 | 17,060 | den-crane | It's how mutations work now.
will be addressed here https://github.com/ClickHouse/ClickHouse/pull/13403 | 2020-11-16T14:10:21 | 2020-11-16T14:10:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,088,835 | 17,079 | den-crane | hm, but Gregorian calendar itself does support leap seconds? AFAIK there is no leap seconds in human calendars?
Leap seconds just last longer than general seconds ?
Or you asking about attribute (function) to check is it leap second or not? | 2020-11-16T14:15:14 | 2020-11-16T14:17:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,168,222 | 17,088 | filimonov | ```
select '1111' as name from system.numbers_mt order by name limit 10000;
```
| 2020-11-16T16:22:08 | 2020-11-16T16:22:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,732,162 | 16,588 | KochetovNicolai | Have reverted for release branches. | 2020-11-17T07:04:36 | 2020-11-17T07:04:36 | {} | MEMBER |
ClickHouse | ClickHouse | 728,737,831 | 17,113 | amosbird | Perhaps we can also implement some SQL action to invoke customized http handlers. It'll look like naive user defined SQL functions. | 2020-11-17T07:18:39 | 2020-11-17T07:18:39 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 728,770,732 | 17,081 | l1t1 | my test is ok. can create table and insert data.
btw, I am using wsl 2 | 2020-11-17T08:27:56 | 2020-11-17T08:27:56 | {} | NONE |
ClickHouse | ClickHouse | 728,835,903 | 16,184 | Avogar | > @Avogar Should we apply this setting to `groupArray*` functions?
No, we will apply this setting only for new functions. | 2020-11-17T10:25:32 | 2020-11-17T10:25:32 | {} | MEMBER |
ClickHouse | ClickHouse | 728,250,398 | 17,054 | den-crane | ```
set max_memory_usage='10G', max_bytes_before_external_group_by='3G';
CREATE TABLE sfz15y
ENGINE = MergeTree()
ORDER BY id AS
SELECT (a.number * 10000) + b.number AS id
FROM numbers(50000) AS a
, numbers(10000) AS b
0 rows in set. Elapsed: 10.107 sec.
INSERT INTO sfz15y SELECT (a.number * 10000) +... | 2020-11-16T18:41:10 | 2020-11-16T18:41:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,621,731 | 17,045 | qinglok | [root@localhost ~]# sudo -u clickhouse clickhouse-server --pid-file=/var/run/clickhouse-se rver/clickhouse-server.pid --config-file=/etc/clickhouse-server/config.xml
Processing configuration file '/etc/clickhouse-server/config.xml'.
Including configuration file '/etc/clickhouse-server/metrika.xml'.
... | 2020-11-17T01:27:18 | 2020-11-17T01:28:03 | {} | NONE |
ClickHouse | ClickHouse | 727,791,077 | 15,179 | chang0391 | The two issues have nothing to do with each other | 2020-11-16T07:27:53 | 2020-11-16T07:27:53 | {} | NONE |
ClickHouse | ClickHouse | 727,880,188 | 16,849 | alesapin | Reopen this issue if the error will reproduce with newer version. | 2020-11-16T10:14:01 | 2020-11-16T10:14:01 | {} | MEMBER |
ClickHouse | ClickHouse | 727,891,883 | 15,087 | Akazz | Work to be continued in #17070 | 2020-11-16T10:34:18 | 2020-11-16T10:34:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,109,418 | 16,424 | filimonov | rebased on newer master. | 2020-11-16T14:48:47 | 2020-11-16T14:48:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,646,710 | 16,944 | hexiaoting | @tavplubix Why does this need't an Engine setting?
```
CREATE TABLE table2 AS numbers(5);
CREATE TABLE table3 AS table2;
``` | 2020-11-17T02:32:19 | 2020-11-17T02:32:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,802,970 | 16,991 | filimonov | > Maybe it's not major after all. Because Engine JOIN has supported only join on UInt64 keys before CH 20th something.
What do you mean? Your test perfectly works on 20.3, 19.11, 18.14 (on 18.14 I needed to list the columns explicitly).
| 2020-11-16T07:52:19 | 2020-11-16T07:53:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,144,862 | 17,073 | filimonov | That sounds like client issue. You need to record the trafic beween problematic client and server, and check which headers does it send. (use tcpdump / wireshark or similar tools). | 2020-11-16T15:45:29 | 2020-11-16T15:45:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,725,769 | 17,052 | den-crane | Does it reproduce with non-replicated tables?
Check this https://clickhouse.tech/docs/en/operations/settings/settings/#settings-deduplicate-blocks-in-dependent-materialized-views | 2020-11-16T04:20:40 | 2020-11-16T04:20:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,736,095 | 17,049 | l1t1 | rpm -e clickhouse-server
systemctl daemon-reload
systemctl stop clickhouse-server
rpm -ivh clickhouse-server-20.11.3.3-2.noarch.rpm
systemctl start clickhouse-server | 2020-11-16T05:04:10 | 2020-11-16T05:04:10 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 727,933,649 | 15,419 | myrrc | Hi @akuzm, is there any news about the review? | 2020-11-16T11:57:46 | 2020-11-16T11:57:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,155,074 | 17,087 | filimonov | There is no way for that (currently).
As a workaround you can create 2 views: one with final, another without. | 2020-11-16T16:01:14 | 2020-11-16T16:01:14 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,158,036 | 17,088 | filimonov | Are you using standard builds? | 2020-11-16T16:05:55 | 2020-11-16T16:05:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,165,795 | 17,088 | filimonov | 20.8.3 - Reproduced easily... | 2020-11-16T16:18:21 | 2020-11-16T16:18:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,650,183 | 17,045 | lhcg | chown -R clickhouse:clickhouse /var/run/clickhouse-server
... | 2020-11-17T02:44:09 | 2020-11-17T02:44:09 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 728,762,398 | 17,006 | KochetovNicolai | `Yandex third-party checks` was broken in master.
Ignore timeouts in ci. | 2020-11-17T08:11:23 | 2020-11-17T08:11:23 | {} | MEMBER |
ClickHouse | ClickHouse | 728,764,087 | 17,045 | qinglok | chown: cannot access ‘/var/run/clickhouse-server’: No such file or directory
@lhcg | 2020-11-17T08:13:49 | 2020-11-17T08:13:49 | {} | NONE |
ClickHouse | ClickHouse | 728,788,556 | 17,081 | l1t1 | ```
lt@DESKTOP-RS3EG9A:~$ clickhouse-client -m
ClickHouse client version 20.11.3.3 (official build).
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 20.11.3 revision 54442.
DESKTOP-RS3EG9A.localdomain :) create database mydb;
CREATE DATABASE mydb
Query id: df069962-718d... | 2020-11-17T09:02:31 | 2020-11-17T09:02:31 | {} | NONE |
ClickHouse | ClickHouse | 728,812,202 | 17,073 | RonanMorgan | Hello,
thanks for your answer,
I am doing hundred's of queries per seconds and it seems that the error can happen at anytime during the day, so it will pretty hard to find it with wireshark.
I have just created an issue on the [client side](https://github.com/ClickHouse/clickhouse-jdbc/issues/518) to see if th... | 2020-11-17T09:44:00 | 2020-11-17T09:44:00 | {} | NONE |
ClickHouse | ClickHouse | 727,642,772 | 16,991 | filimonov | > alexey-milovidov removed the prio-major
@alexey-milovidov actually such non-deterministic behaviour doesn't sounds great, use case sounds valid, silent incorrect result is very bad behaviour, and older version worked. So all that sounds like the issue is important and require attention.
So why not major? | 2020-11-15T21:50:04 | 2020-11-15T21:50:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,743,285 | 16,907 | l1t1 | normal CTE works, but recursive CTE not ready yet.
```
ClickHouse client version 20.11.3.3 (official build).
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 20.11.3 revision 54442.
DESKTOP-RS3EG9A.localdomain :) WITH test1 AS (SELECT number n FROM numbers(10)) SELECT * FROM t... | 2020-11-16T05:28:11 | 2020-11-16T05:28:11 | {} | NONE |
ClickHouse | ClickHouse | 727,745,744 | 16,702 | baibaichen | @alexey-milovidov
here are relies from [slack](https://app.slack.com/client/TUEGQCHJT/CU170QE9H/thread/CU170QE9H-1604975270.033500)
| 2020-11-16T05:34:51 | 2020-11-16T06:17:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,060,849 | 17,091 | zhang2014 | CC @tavplubix | 2020-11-16T13:57:35 | 2020-11-16T13:57:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,470,602 | 16,991 | den-crane | hm, maybe I am confused.
one more issue
https://github.com/ClickHouse/ClickHouse/blob/master/CHANGELOG.md
Add new ANY JOIN logic for StorageJoin consistent with JOIN operation. To upgrade without changes in behaviour you need add SETTINGS any_join_distinct_right_table_keys = 1 to Engine Join tables metadata or... | 2020-11-17T00:11:08 | 2020-11-17T00:13:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,747,226 | 15,235 | kaka11chen | I have the same issue too. | 2020-11-17T07:41:42 | 2020-11-17T07:41:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,753,309 | 17,088 | thiningsun | > ```
> select '1111' as name from system.numbers_mt order by name limit 10000;
> ```
What do you mean ？Can you reproduce my problem
As long as the version is greater than 20.03, and the amount of data in the table is tens of millions, this error will appear. I tested the version as version: 20.3.15.133, 19.17.10... | 2020-11-17T07:55:51 | 2020-11-17T07:56:55 | {} | NONE |
ClickHouse | ClickHouse | 728,891,810 | 17,100 | akuzm | About the changes in https://github.com/ClickHouse/ClickHouse/pull/17100/commits/b0e660651ab2d5573c07cab772c96f021d09d565 .
The version with lambda is wasteful in debug mode, and just looks weird. I am surprised that `const a = []() {return b;};` might give better code than `a; a = b;`, but this is C++...
I look... | 2020-11-17T12:22:12 | 2020-11-17T12:25:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,693,683 | 17,045 | qinglok | This command cannot be run.
`sudo -u clickhouse clickhouse-server -pid-file=/var/run/clickhouse-server/clickhouse-server.pid --config-file=/etc/clickhouse-server/config.xml`
`Unknown option specified: pid-file=/var/run/clickhouse-server/clickhouse-server.pid` | 2020-11-16T02:20:12 | 2020-11-16T02:20:12 | {} | NONE |
ClickHouse | ClickHouse | 727,721,712 | 16,944 | hexiaoting | It runs ok with version: 20.12.1.1.
```
ClickHouse client version 20.12.1.1.
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 20.12.1 revision 54442.
dell123 :) CREATE TABLE table2 engine=Log AS select number from numbers(5)
CREATE TABLE table2
ENGINE = Log AS
SELECT numb... | 2020-11-16T04:02:56 | 2020-11-16T04:04:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,939,112 | 15,419 | akuzm | > Hi @akuzm, is there any news about the review?
Didn't have a chance too look at this last week because I was on support duty. Trying to reproduce the perf test failure now, without success. | 2020-11-16T12:09:24 | 2020-11-16T12:09:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,139,128 | 17,079 | filimonov | > Gregorian calendar itself does support leap seconds? human calendars?
They have.
> Leap seconds just last longer than general seconds ?
That's how it works with Unix timestamps.
We use unix epoch internally. Leap seconds are not supported there.
So '2005-12-31 23:59:60 UTC' (which can be stored as ... | 2020-11-16T15:36:31 | 2020-11-16T15:37:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,728,443 | 17,095 | alesapin | Stress failure is unrelated https://gist.github.com/alesapin/fa9ae8ac1f07a9d6073ca64d6e6fb0c5. But seems like we have to fix something in "arcadia"? | 2020-11-17T06:56:09 | 2020-11-17T06:56:09 | {} | MEMBER |
ClickHouse | ClickHouse | 728,774,604 | 17,088 | thiningsun | > Sounds very dangerous but I cannot reproduce it on 20.8.1.1
Are you sure it is 20.8.1.1? ? I can't find it in ：https://repo.yandex.ru/clickhouse/rpm/stable/x86_64/
or https://repo.yandex.ru/clickhouse/rpm/lts/x86_64/ | 2020-11-17T08:35:50 | 2020-11-17T08:35:50 | {} | NONE |
ClickHouse | ClickHouse | 728,782,196 | 16,937 | satanson | @alexey-milovidov
the pr can be merged? | 2020-11-17T08:50:16 | 2020-11-17T08:50:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,859,168 | 16,680 | Avogar | After this changes we don't need an additional layer from my previous fix, so I removed it. My tests show that quantileTDigest is much better now, thank you for your work. After the CI checks are completed and there are no errors related to this PR we will merge it. | 2020-11-17T11:12:22 | 2020-11-17T11:12:22 | {} | MEMBER |
ClickHouse | ClickHouse | 727,651,209 | 16,987 | azat | >2020.11.13 22:11:00.370979 [ 12414 ] {} app_ps_metrics_w.Distributed.DirectoryMonitor: Code: 210, e.displayText() = DB::NetException: Connection refused (xj01-www-01.xj01:9001), Stack trace
This is expected
>but when i delete the corresponds directory to the shard /home/disk3/clickhouse/data/aaa/bbb/default:OEg... | 2020-11-15T22:50:10 | 2020-11-15T22:50:10 | {} | MEMBER |
ClickHouse | ClickHouse | 727,732,899 | 17,050 | den-crane | 20.4.4.18 is out of support. Supported versions 20.9 20.10 20.11 and LTS.
Please provide excerpts around the error from a logfile | 2020-11-16T04:52:05 | 2020-11-16T04:52:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,798,107 | 16,907 | filimonov | > recursive CTE
https://www.sqlservertutorial.net/sql-server-basics/sql-server-recursive-cte/
Most probably they will not be supported in near future.
| 2020-11-16T07:43:17 | 2020-11-16T07:43:17 | {
"eyes": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,839,031 | 16,048 | tavplubix | Yandex synchronization check - broken in master | 2020-11-16T09:02:17 | 2020-11-16T09:02:17 | {} | MEMBER |
ClickHouse | ClickHouse | 728,264,112 | 16,987 | azat | >by the way，Is there a more elegant way when I modify the replica config next time?
If you need to change the host name you can use `use_compact_format_in_distributed_parts_names=1`, this setting changes the format of the directories that will be used for async distributed insert, and will use `shardX_replicaX` inst... | 2020-11-16T19:07:21 | 2020-11-16T19:07:21 | {} | MEMBER |
ClickHouse | ClickHouse | 728,840,340 | 16,424 | alesapin | Again rocks DB failure https://gist.github.com/alesapin/c9a8a57e7b57a81edc1d63aa70e469ce | 2020-11-17T10:34:19 | 2020-11-17T10:34:19 | {} | MEMBER |
ClickHouse | ClickHouse | 727,641,416 | 17,045 | l1t1 | try to mv all of the content under old path to the new path | 2020-11-15T21:39:36 | 2020-11-15T21:39:36 | {} | NONE |
ClickHouse | ClickHouse | 727,907,268 | 15,977 | TCeason | There are still a few mistakes, but it's probably none of my business. @tavplubix | 2020-11-16T11:03:13 | 2020-11-16T11:03:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,141,882 | 17,081 | filimonov | We don't support WSL, sorry. Most probably it's some of POSIX features missing in the filesystem / or on WSL layer.
Use VM. | 2020-11-16T15:40:47 | 2020-11-16T15:40:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 728,145,445 | 17,055 | filimonov | BTW we don't support WSL, and clickhouse may misbehave there. | 2020-11-16T15:46:23 | 2020-11-16T15:46:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 727,909,215 | 15,977 | tavplubix | > There are still a few mistakes, but it's probably none of my business. @tavplubix
Yes, failed tests are unrelated. | 2020-11-16T11:06:31 | 2020-11-16T11:06:31 | {} | MEMBER |
ClickHouse | ClickHouse | 727,988,294 | 16,946 | alesapin | > Log fragment attached
> [log_fragment.txt.gz](https://github.com/ClickHouse/ClickHouse/files/5541343/log_fragment.txt.gz)
Unfortunately log level is too silent, no query which leads to crash, but there are a lot of errors:
```
2020.11.12 12:40:29.860175 [ 24881 ] {} <Error> default.view_tizer: auto DB::StorageR... | 2020-11-16T13:12:43 | 2020-11-16T13:12:43 | {} | MEMBER |
ClickHouse | ClickHouse | 728,453,749 | 17,054 | l1t1 | 1.`insert into sfzcm select id,count(*)c from sfz15y where id<100000000 group by id having count(*)>1;`
2.`insert into sfzcm select id,count(*)c from sfz15y where id>=100000000 and id<200000000 group by id having count(*)>1;` | 2020-11-17T00:00:54 | 2020-11-17T00:00:54 | {} | NONE |
ClickHouse | ClickHouse | 728,573,188 | 17,054 | l1t1 | thanks, one more question
how to check the current value of `max_memory_usage, max_bytes_before_external_group_by` etc | 2020-11-17T00:57:49 | 2020-11-17T00:57:49 | {} | NONE |
ClickHouse | ClickHouse | 728,772,579 | 17,088 | amosbird | Sounds very dangerous but I cannot reproduce it on 20.8.1.1 | 2020-11-17T08:31:43 | 2020-11-17T08:31:43 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 728,901,196 | 17,088 | filimonov | > Sounds very dangerous but I cannot reproduce it on 20.8.1.1
I guess somewhere (PartialSortingTransform?) we don't expect to get a constant column.
There were a lot of bugs like that several years ago.
Try build with asan or just repeat the query lot of time with clickhouse-benchmark.
```
2020.11.17 13:37:... | 2020-11-17T12:39:52 | 2020-11-17T12:39:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,565,735 | 15,977 | tavplubix | > "Integration tests flaky check (asan)" failed but "Integration tests (.*)" is succeeded. Which one should I believe？ @tavplubix
Both ones. `Integration tests` check runs all integration tests. `Integration tests flaky check` runs new or modified integration tests multiple times to check if some test is flacky (i.e... | 2020-10-29T09:53:59 | 2020-10-29T09:53:59 | {} | MEMBER |
ClickHouse | ClickHouse | 718,575,531 | 15,977 | TCeason | > > "Integration tests flaky check (asan)" failed but "Integration tests (.*)" is succeeded. Which one should I believe？ @tavplubix
>
> Both ones. `Integration tests` check runs all integration tests. `Integration tests flaky check` runs new or modified integration tests multiple times to check if some test is flack... | 2020-10-29T10:03:52 | 2020-10-29T10:03:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 718,890,294 | 14,731 | azat | @alexey-milovidov looks like `Yandex third-party checks` is broken in master, maybe it's time to break it additionally with this changes? :) | 2020-10-29T17:01:58 | 2020-10-29T17:01:58 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.