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 | 762,376,565 | 19,236 | akuzm | Fuzzer https://github.com/ClickHouse/ClickHouse/issues/19251 | 2021-01-18T17:13:10 | 2021-01-18T17:13:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,510,892 | 16,274 | alexey-milovidov | > Additional context
In 20.8 MOVE PARTITION working correctly.
It means that the issue is already fixed in recent versions?
But we are not going to search for the fix and backport it to version 20.3, because it happens only in specific scenario that is only possible with manual intervention. | 2021-01-18T23:21:52 | 2021-01-18T23:22:04 | {} | MEMBER |
ClickHouse | ClickHouse | 762,512,344 | 18,013 | alexey-milovidov | @simPod I cannot reproduce this issue:
```
$ clickhouse-client
ClickHouse client version 21.1.1.5646.
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 21.1.1 revision 54443.
milovidov-desktop :) CREATE TABLE t_a
:-] (
:-] `fetched_at` D... | 2021-01-18T23:28:09 | 2021-01-18T23:28:09 | {} | MEMBER |
ClickHouse | ClickHouse | 762,082,286 | 13,405 | akuzm | Fuzzer https://github.com/ClickHouse/ClickHouse/issues/15889 | 2021-01-18T08:37:00 | 2021-01-18T08:37:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,175,727 | 6,462 | alesapin | ```
:) CREATE TABLE example_table
( ... | 2021-01-18T11:06:52 | 2021-01-18T11:08:27 | {} | MEMBER |
ClickHouse | ClickHouse | 762,214,153 | 19,201 | alexey-milovidov | Something like
`awk '{ print $1 "\t" $1 + 1 }'`
This script represents a dictionary that have value `x + 1` for every key `x`, in TSV format. | 2021-01-18T12:17:48 | 2021-01-18T12:18:03 | {} | MEMBER |
ClickHouse | ClickHouse | 762,476,151 | 18,778 | alexey-milovidov | No answer... | 2021-01-18T21:22:39 | 2021-01-18T21:22:39 | {} | MEMBER |
ClickHouse | ClickHouse | 762,056,322 | 19,168 | alesapin | Not required. | 2021-01-18T07:51:56 | 2021-01-18T07:51:56 | {} | MEMBER |
ClickHouse | ClickHouse | 762,093,304 | 19,215 | alesapin | Let me run integration tests with msan manually. | 2021-01-18T08:54:15 | 2021-01-18T08:54:15 | {} | MEMBER |
ClickHouse | ClickHouse | 762,122,536 | 18,788 | hexiaoting | @nikitamikhaylov ready to review. | 2021-01-18T09:41:52 | 2021-01-18T09:41:52 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,203,349 | 6,462 | NikolayPraydkin | > select * from example_table
не сбрасывается даже на дефолтное значение. | 2021-01-18T11:57:41 | 2021-01-18T11:57:41 | {} | NONE |
ClickHouse | ClickHouse | 762,550,346 | 16,925 | vvhungy | Any update on this issue?
We got similar error for some of our partitions. Before the data is in .mrk2 (format) files, after an FULL optimmize (OPTIMIZE ... PARTITION... FINAL), data files in these partitions became .mrk3 format and then those partitions couldn't be read again and produced error "Bad size of marks f... | 2021-01-19T01:52:37 | 2021-01-19T01:55:23 | {} | NONE |
ClickHouse | ClickHouse | 762,056,205 | 19,166 | alesapin | Not required. | 2021-01-18T07:51:44 | 2021-01-18T07:51:44 | {} | MEMBER |
ClickHouse | ClickHouse | 762,158,692 | 6,462 | NikolayPraydkin | Создал таблицу как в примере, только интервал удаления через минуту
CREATE TABLE example_table
(
d DateTime,
a Int TTL d + INTERVAL 1 minute,
b Int TTL d + INTERVAL 1 minute,
c String
)
ENGINE = MergeTree
PARTITION BY toYYYYMM(d)
ORDER BY d;
вставил пару строк
insert into example_table val... | 2021-01-18T10:40:28 | 2021-01-18T10:40:28 | {} | NONE |
ClickHouse | ClickHouse | 762,204,776 | 19,201 | MaxJPowers | On your website it uses the command cat however as you said I realise this reads from stdout. What script should I run for stdout? I've tried using `read ...` but this lead an error. Note the dictionary is loaded.
`
Code: 302, e.displayText() = DB::Exception: Child process was exited with return code 1 (version 20.8.... | 2021-01-18T12:00:25 | 2021-01-18T12:06:43 | {} | NONE |
ClickHouse | ClickHouse | 762,214,581 | 19,229 | chiragguptadtu | > Can you execute your query via clickhouse client with set setting:
And show logs of execution?
[ch_query_execution_trace.log](https://github.com/ClickHouse/ClickHouse/files/5830112/ch_query_execution_trace.log)
| 2021-01-18T12:18:35 | 2021-01-18T12:18:35 | {} | NONE |
ClickHouse | ClickHouse | 762,428,242 | 19,249 | den-crane | https://clickhouse.tech/docs/en/engines/table-engines/mergetree-family/replication/
Data blocks are deduplicated. For multiple writes of the same data block (data blocks of the same size containing the same rows in the same order), the block is only written once. The reason for this is in case of network failures wh... | 2021-01-18T19:14:32 | 2021-01-18T19:14:32 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,457,105 | 18,927 | alexey-milovidov | @filimonov In your example, I see that version 20.10 has /etc/clickhouse-server owned by root and 20.11 has /etc/clickhouse-server owned by clickhouse. But version 20.10 has `-rw-r--r--` (everyone can read config) and 20.11 has `-r--------` (only `clickhouse` can read its own configs).
Allowing everyone to read conf... | 2021-01-18T20:28:25 | 2021-01-18T20:30:09 | {} | MEMBER |
ClickHouse | ClickHouse | 762,513,029 | 19,255 | alexey-milovidov | Does it reproduce on version 20.8.12 or 21.1?
These versions contain bugfixes. | 2021-01-18T23:31:05 | 2021-01-18T23:31:05 | {} | MEMBER |
ClickHouse | ClickHouse | 762,026,913 | 19,223 | Goolen | > /etc/clickhouse-serv
thank you first
After changing the permissions to 755 , start is OK.
There is no permission to change this directory after installation，I don't know why it became 660.
| 2021-01-18T06:53:17 | 2021-01-18T06:53:17 | {} | NONE |
ClickHouse | ClickHouse | 762,177,816 | 19,131 | alesapin | Actually false-positive. Msan doesn't consider data returned from `epoll_wait` initialized. | 2021-01-18T11:10:28 | 2021-01-18T11:10:28 | {} | MEMBER |
ClickHouse | ClickHouse | 762,184,431 | 19,215 | alesapin | Tons of failures https://clickhouse-test-reports.s3.yandex.net/19215/ce6dc40888de09919013809c34af3c59c05e87a5/integration_tests_(memory).html and kerberized kafka tests even didn't start. | 2021-01-18T11:22:16 | 2021-01-18T11:22:16 | {} | MEMBER |
ClickHouse | ClickHouse | 762,351,690 | 19,229 | UnamedRus | What if we disable two level aggregation.
```
set group_by_two_level_threshold_bytes=0;
set group_by_two_level_threshold=0;
``` | 2021-01-18T16:24:41 | 2021-01-18T16:24:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,492,716 | 19,063 | CurtizJ | `test_ttl_replicated/test.py::test_ttl_empty_parts` fixed in master. | 2021-01-18T22:14:52 | 2021-01-18T22:14:52 | {} | MEMBER |
ClickHouse | ClickHouse | 762,278,656 | 19,215 | alesapin | Ok, let's merge. | 2021-01-18T14:17:48 | 2021-01-18T14:17:48 | {} | MEMBER |
ClickHouse | ClickHouse | 762,327,218 | 19,153 | KochetovNicolai | LGTM too | 2021-01-18T15:41:52 | 2021-01-18T15:41:52 | {} | MEMBER |
ClickHouse | ClickHouse | 762,380,611 | 15,162 | PHaroZ | Hi,
This PR breaks a lots of queries in my app : queries like
```sql
SELECT
t.value,
i18n.title
FROM
aTable t
LEFT ANY JOIN aTable_i18n i18n
ON i18n.ref_id = t.id AND i18n.lang = 'en'
```
Is it really what you wanted or it's an unexpected side-effect ? | 2021-01-18T17:21:09 | 2021-01-18T17:21:09 | {
"+1": 4
} | NONE |
ClickHouse | ClickHouse | 762,537,200 | 19,258 | den-crane | What do you mean by `awk -F:',' '{print $0}' ` ?
You probably should use `awk -F ','`
And for cached dictionaries executable script should be much more complicated.
You need to read STDIN. Because CH passes the requested keys using STDIN.
| 2021-01-19T01:01:41 | 2021-01-19T01:01:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,474,829 | 18,617 | alexey-milovidov | @vzakaznikov Wrong exception code.
The LOGICAL_ERROR code acts as debug assertion. | 2021-01-18T21:19:17 | 2021-01-18T21:19:17 | {} | MEMBER |
ClickHouse | ClickHouse | 762,475,973 | 18,813 | alexey-milovidov | @tkionshao Did you check the newer version? | 2021-01-18T21:22:11 | 2021-01-18T21:22:11 | {} | MEMBER |
ClickHouse | ClickHouse | 762,583,078 | 16,925 | den-crane | @vvhungy your issue is solved https://github.com/ClickHouse/ClickHouse/pull/18381 | 2021-01-19T03:44:31 | 2021-01-19T03:44:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,135,275 | 14,100 | KochetovNicolai | Are there recent fails? Attached logs are already removed from s3.
I suppose it may be already fixed | 2021-01-18T10:02:21 | 2021-01-18T10:02:21 | {} | MEMBER |
ClickHouse | ClickHouse | 762,157,688 | 19,030 | tavplubix | It's because `SET` query changes setting value in current session. Distributed DDL queries are executed in separate session and `SET` query does not affect it. You should enable `allow_experimental_database_materialize_mysql` setting globally in server config to make it work with `ON CLUSTER` queries. Example:
```
... | 2021-01-18T10:38:52 | 2021-01-18T10:38:52 | {} | MEMBER |
ClickHouse | ClickHouse | 762,169,635 | 19,216 | alexey-milovidov | > Performance — 9 faster, 1 slower, 87 unstable
Makes sense.
FYI @BohuTANG @sundy-li | 2021-01-18T10:57:11 | 2021-01-18T10:57:11 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 762,172,063 | 19,186 | alexey-milovidov | Fuzzer, ASan:
`SELECT max(id) FROM bloom_filter WHERE hasToken(s, 'zzzzzzzzzzzzzzzzzz\0zzzzz')` | 2021-01-18T11:01:24 | 2021-01-18T11:01:24 | {} | MEMBER |
ClickHouse | ClickHouse | 762,198,186 | 19,229 | chiragguptadtu | Thanks for the reply.
> How long that query takes time? And how many rows Clickhouse is reading?
With 14 threads it takes 27.205 seconds
With 28 threads it takes 15.511 seconds
With 56 threads it takes 16.886 seconds
Clickhouse is reading 13.77 billion rows
> If it short, it's possible that overhead on creat... | 2021-01-18T11:47:52 | 2021-01-18T11:47:52 | {} | NONE |
ClickHouse | ClickHouse | 762,206,584 | 19,229 | UnamedRus | BTW:
```
SELECT
CAST(type as Enum('NA' = 0, 'First Price' = 1, 'Second Price' = 2, 'Unknown' = 3)) AS optimization_auction_type,
SUM(unique_wins) AS result_wins,
SUM(unique_impressions) AS result_impressions,
SUM(unique_clicks) AS result_clicks,
SUM(unique_installs) AS result_mmp_click_inst... | 2021-01-18T12:03:30 | 2021-01-18T12:04:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,258,408 | 18,944 | alexey-milovidov | FYI this bug was not present in any ClickHouse release. | 2021-01-18T13:41:58 | 2021-01-18T13:41:58 | {} | MEMBER |
ClickHouse | ClickHouse | 762,368,377 | 19,229 | UnamedRus | Memory usage is down almost 2 times, but no impact on speed.
If you have enabled query_log table, can you share log records for those queries (you can filter by query_id) from it? | 2021-01-18T16:56:17 | 2021-01-18T16:56:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,409,244 | 17,951 | alexey-milovidov | 21.1 is released to stable. | 2021-01-18T18:26:10 | 2021-01-18T18:26:10 | {} | MEMBER |
ClickHouse | ClickHouse | 762,417,541 | 19,229 | alexey-milovidov | @UnamedRus A few remarks:
1. Thread creation is expected to be not longer than 10 microseconds on Linux, so it's not more than 1 millisecond to create 100 threads. Usually neglible. Even more - ClickHouse does not create threads for every query as thread pool is used (but the performance of assigning tasks to thread... | 2021-01-18T18:47:06 | 2021-01-18T18:49:17 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 762,423,045 | 19,244 | alexey-milovidov | 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? | 2021-01-18T19:01:33 | 2021-01-18T19:01:33 | {} | MEMBER |
ClickHouse | ClickHouse | 762,499,811 | 15,449 | UnamedRus | Master
Second approach works.
```
CREATE TABLE test_index
(
`key_string` String,
`key_uint32` ALIAS toUInt32(key_string),
INDEX idx toUInt32(key_string) TYPE set(0) GRANULARITY 1
)
ENGINE = MergeTree
PARTITION BY tuple()
PRIMARY KEY tuple()
ORDER BY key_string;
INSERT INTO test_index SELECT *... | 2021-01-18T22:38:57 | 2021-01-18T22:38:57 | {
"rocket": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,506,019 | 18,197 | alexey-milovidov | > Always the first element of an enum (would not recommend)
Actually it is expected to be implemented this way - get the default value of the data type, which is the first element for Enum. | 2021-01-18T23:03:06 | 2021-01-18T23:03:06 | {} | MEMBER |
ClickHouse | ClickHouse | 762,132,195 | 19,028 | KochetovNicolai | > What feature will be losed without ssl support? I'll try it next to build on lastest version.
https://clickhouse.tech/docs/en/operations/server-configuration-parameters/settings/#server_configuration_parameters-openssl
In other words, `https_port` and `tcp_port_secure` will not work | 2021-01-18T09:57:14 | 2021-01-18T09:57:14 | {} | MEMBER |
ClickHouse | ClickHouse | 762,171,366 | 19,210 | alexey-milovidov | > AST fuzzer (debug) — Task failed: $?=1
There were changes in CI. | 2021-01-18T11:00:14 | 2021-01-18T11:00:14 | {} | MEMBER |
ClickHouse | ClickHouse | 762,223,077 | 15,174 | alexey-milovidov | We don't release packages (deb, rpm) for AArch64, only binaries.
Adding packages requires more efforts. | 2021-01-18T12:35:16 | 2021-01-18T12:35:33 | {} | MEMBER |
ClickHouse | ClickHouse | 762,231,159 | 19,201 | MaxJPowers | Ah okay thanks, for a csv file I used `awk -F:',' '{print $1 "," $1 + 1}' /var/lib/clickhouse/user_files/testCache.csv`. However now when I run queries such as `SELECT dictGet('cacheDictionary', 'random1', toUInt64(392)) LIMIT 1` take longer than 60s, is this normal? | 2021-01-18T12:51:07 | 2021-01-18T12:51:23 | {} | NONE |
ClickHouse | ClickHouse | 762,376,492 | 19,251 | akuzm | https://clickhouse-test-reports.s3.yandex.net/19236/b2547e99a0fa7b2fe6ba6261e95af49f089458eb/fuzzer_ubsan/report.html#fail1 | 2021-01-18T17:13:01 | 2021-01-18T17:13:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,473,350 | 19,234 | alexey-milovidov | Yes, it's duplicate.
Also it's not a "bug" because no observable behaviour of ClickHouse is incorrect. | 2021-01-18T21:15:10 | 2021-01-18T21:15:10 | {} | MEMBER |
ClickHouse | ClickHouse | 762,484,672 | 18,315 | alexey-milovidov | Workaround: add replace: `<max_memory_usage replace="replace" from_env="MAX_MEMORY_USAGE" />` | 2021-01-18T21:48:12 | 2021-01-18T21:48:12 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 762,355,066 | 19,229 | chiragguptadtu | > What if we disable two level aggregation.
Similar results after disabling two level aggregation.
[ch_query_execution_trace_2.log](https://github.com/ClickHouse/ClickHouse/files/5831403/ch_query_execution_trace_2.log)
| 2021-01-18T16:31:14 | 2021-01-18T16:31:14 | {} | NONE |
ClickHouse | ClickHouse | 762,479,882 | 18,242 | alexey-milovidov | What is the percentage of completion when it starts to consume a lot of memory?
How the graph of memory usage looks - is it linear growth or a spike at some point of time? | 2021-01-18T21:32:51 | 2021-01-18T21:32:51 | {} | MEMBER |
ClickHouse | ClickHouse | 762,027,944 | 19,223 | Goolen | > > > check `/etc` and `/etc/clickhouse-server` directories's permission(mode and owner)
> >
> >
> > # ll -d /etc/
> > drwxr-xr-x. 110 root root 8192 Jan 15 15:13 /etc/
> > # ll -d /etc/clickhouse-server
> > drw-rw---- 4 clickhouse clickhouse 183 Jan 18 09:06 /etc/clickhouse-server
>
> chown /etc to owner cl... | 2021-01-18T06:55:22 | 2021-01-18T06:55:22 | {} | NONE |
ClickHouse | ClickHouse | 762,182,383 | 19,229 | UnamedRus | > Someone please explain to me, why this might be happening. Am is missing something ?
How long that query takes time? And how many rows Clickhouse is reading?
If it short, it's possible that overhead on creation of additional threads much bigger than performance gain from them. | 2021-01-18T11:18:28 | 2021-01-18T11:18:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,327,821 | 19,247 | alexey-milovidov | https://github.com/apla/node-clickhouse/issues | 2021-01-18T15:42:56 | 2021-01-18T15:42:56 | {} | MEMBER |
ClickHouse | ClickHouse | 762,480,357 | 18,242 | alexey-milovidov | ClickHouse has builtin memory profiler that can show the actual places where memory is allocated.
I can tell the details so we can debug this issue... | 2021-01-18T21:34:18 | 2021-01-18T21:34:18 | {} | MEMBER |
ClickHouse | ClickHouse | 762,513,525 | 19,255 | ivanprostran | > Does it reproduce on version 20.8.12 or 21.1?
> These versions contain bugfixes.
Thank you for the quick reply.
I must check it because I use Altinity repo (obviously not so recent?)
| 2021-01-18T23:33:08 | 2021-01-18T23:33:08 | {} | NONE |
ClickHouse | ClickHouse | 762,537,902 | 19,258 | MaxJPowers | What do you mean should be much more complicated, as in I should be filtering by the keys inputed when launching a query? There aren't any examples online that I've found so I'm unsure. | 2021-01-19T01:04:31 | 2021-01-19T01:04:54 | {} | NONE |
ClickHouse | ClickHouse | 762,589,440 | 18,927 | filimonov | The bad thing here: since clickhouse is owner the process can do chmod and do wharever it need after that in config folder => insecure.
Also now even root can not edit files w/o making them writable => not handy, easy to forget to switch back to r/o
I guess we can make it this way:
files in etc/clickhouse-serve... | 2021-01-19T04:10:34 | 2021-01-19T04:10:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,592,653 | 18,671 | filimonov | Ok. It seems now Kafka tests are green.
Several other failed:
Test name | Test status | Test time, sec.
-- | -- | --
test_adaptive_granularity/test.py::test_version_single_node_update[n1-tables1] | FAIL | 51.88
test_adaptive_granularity/test.py::test_mixed_granularity_single_node[node0] | FAIL | 40.93
test_... | 2021-01-19T04:23:35 | 2021-01-19T04:23:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,477,034 | 19,131 | alexey-milovidov | If false positive, it cannot be a "bug" even theoretically. | 2021-01-18T21:25:03 | 2021-01-18T21:25:03 | {} | MEMBER |
ClickHouse | ClickHouse | 762,479,098 | 18,242 | alexey-milovidov | > I would not say that String fields are large. For example bid_id is no more than 16 characters. Maybe the user_agent field contains a lot of characters, but it is not all filled. There are many fields of type String themselves in this form: tracker LowCardinality (String) and city String
What are the actual `max(l... | 2021-01-18T21:30:29 | 2021-01-18T21:30:29 | {} | MEMBER |
ClickHouse | ClickHouse | 762,511,870 | 17,759 | alexey-milovidov | Does it reproduce in version 21.1?
We have made some important fixes for joins recently. | 2021-01-18T23:26:02 | 2021-01-18T23:26:02 | {} | MEMBER |
ClickHouse | ClickHouse | 762,190,138 | 19,234 | alesapin | Possibly fixed in https://github.com/ClickHouse/ClickHouse/pull/19214 | 2021-01-18T11:32:44 | 2021-01-18T11:32:44 | {} | MEMBER |
ClickHouse | ClickHouse | 762,408,309 | 19,229 | chiragguptadtu | > If you have enabled query_log table, can you share log records for those queries (you can filter by query_id) from it?
console.kayzen.io :) select * from system.query_log where query_id in ('0dccb324-3153-4b5b-853c-0f07f9900988', 'af4f355e-de89-4059-bf04-980df65a2723') \G
SELECT *
FROM system.query_log
WHERE ... | 2021-01-18T18:24:01 | 2021-01-18T18:24:01 | {} | NONE |
ClickHouse | ClickHouse | 762,517,461 | 19,255 | alexey-milovidov | @ivanprostran You can use ClickHouse official repository:
https://clickhouse.tech/
\- go to Quick Start;
\- select "CentOS or RedHat".
The packages in our repository and in Altinity repository are bitwise identical - the Altinity repo is just a mirror (that may lag behind but usually should not). | 2021-01-18T23:47:39 | 2021-01-18T23:47:39 | {} | MEMBER |
ClickHouse | ClickHouse | 762,549,937 | 19,028 | ElapsedSoul | @alexey-milovidov
When I build with the newest version(master),I got this :
```
[root@10-20-103-5 build]# ninja clickhouse
[0/2] Re-checking globbed directories...
[11/5109] Building C object contrib/croaring-cmake/CMakeFiles/roaring.dir/__/croaring/src/roaring.c.o
FAILED: contrib/croaring-cmake/CMakeFiles/roari... | 2021-01-19T01:51:05 | 2021-01-19T03:36:03 | {} | NONE |
ClickHouse | ClickHouse | 762,171,052 | 19,214 | alexey-milovidov | Fuzzer, ASan: https://github.com/ClickHouse/ClickHouse/issues/19219 | 2021-01-18T10:59:42 | 2021-01-18T10:59:42 | {} | MEMBER |
ClickHouse | ClickHouse | 762,173,296 | 19,186 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/issues/19233 | 2021-01-18T11:03:12 | 2021-01-18T11:03:12 | {} | MEMBER |
ClickHouse | ClickHouse | 762,185,380 | 19,215 | alexey-milovidov | So everything is alright :) | 2021-01-18T11:24:00 | 2021-01-18T11:24:00 | {} | MEMBER |
ClickHouse | ClickHouse | 762,198,862 | 15,174 | alexey-milovidov | Tests under `qemu-user` trigger segfault in QEMU, although they work perfectly fine on real AArch64 server (e.g. Graviton).
Maybe we have to use full-featured `qemu-system` (that is significantly harder) or to disable some functionality (but it's difficult to figure out the culprit) or to use real server (but we don... | 2021-01-18T11:49:11 | 2021-01-18T11:49:11 | {} | MEMBER |
ClickHouse | ClickHouse | 762,392,914 | 19,224 | Avogar | Failed tests are not related. | 2021-01-18T17:48:24 | 2021-01-18T17:48:24 | {} | MEMBER |
ClickHouse | ClickHouse | 762,574,188 | 11,193 | BigRantLing | @filimonov Hi ! Is `background_pool` used by all MergeTree() engine on single node? | 2021-01-19T03:13:45 | 2021-01-19T03:13:45 | {} | NONE |
ClickHouse | ClickHouse | 762,157,746 | 19,114 | alesapin | ```
2021.01.15 16:37:04.061493 [ 247 ] {} <Fatal> BaseDaemon: (version 20.12.5.5688, build id: 06E771BF3A63C600) (from thread 562) Terminate called for uncaught exception:
2021.01.15 16:37:04.062882 [ 371136 ] {} <Fatal> BaseDaemon: ########################################
2021.01.15 16:37:04.063891 [ 371136 ] {} <F... | 2021-01-18T10:38:58 | 2021-01-18T10:38:58 | {} | MEMBER |
ClickHouse | ClickHouse | 762,219,432 | 15,174 | alexey-milovidov | I have run functional tests on AArch64 server and get the following results:
- ClickHouse is perfectly stable;
- the messages `libunwind: Unsupported .eh_frame_hdr version` are printed in stderr sometimes;
- there are minor differences in results of floating point functions. | 2021-01-18T12:28:11 | 2021-01-18T12:28:11 | {} | MEMBER |
ClickHouse | ClickHouse | 762,221,633 | 15,174 | l1t1 | > I have run functional tests on AArch64 server and get the following results:
>
> * ClickHouse is perfectly stable;
> * the messages `libunwind: Unsupported .eh_frame_hdr version` are printed in stderr sometimes;
> * there are minor differences in results of floating point functions.
could you also release an ... | 2021-01-18T12:32:29 | 2021-01-18T12:32:29 | {} | NONE |
ClickHouse | ClickHouse | 762,226,074 | 15,174 | l1t1 | > We don't release packages (deb, rpm) for AArch64, only binaries.
> Adding packages requires more efforts.
gzip binary file without debug info is enough. thanks | 2021-01-18T12:40:56 | 2021-01-18T12:40:56 | {} | NONE |
ClickHouse | ClickHouse | 762,236,379 | 19,201 | alexey-milovidov | Cache dictionaries are not for static files. They are for the cases when data is prepared on the fly, dynamically.
If data is static, simply use flat/hashed dictionaries. | 2021-01-18T13:01:41 | 2021-01-18T13:01:41 | {} | MEMBER |
ClickHouse | ClickHouse | 762,334,013 | 6,462 | NikolayPraydkin | > > select * from example_table
>
> не сбрасывается даже на дефолтное значение.
вопрос снимается! создание и запрос optimize делал через DBeaver(вы я так понимаю делали через clickhouse-client.)
Ошибки не было, но как только указал перед названием таблицы название базы, запрос отработал как ожидалось и значени... | 2021-01-18T15:53:30 | 2021-01-18T15:53:30 | {} | NONE |
ClickHouse | ClickHouse | 762,360,572 | 19,201 | MaxJPowers | Okay, I thought it was an issue with speed testing on smaller files I get a line feed error when running the following:
`
awk -F:',' '{print $1 "," $1 + 1}' /var/lib/clickhouse/user_files/testCache.csv
`
I changed it to
`awk -v RS='\r\n' -F:',' '{print $1 "," $1 + 1 "\r\n"}' /var/lib/clickhouse/user_files/testCac... | 2021-01-18T16:41:11 | 2021-01-18T16:41:11 | {} | NONE |
ClickHouse | ClickHouse | 762,445,593 | 19,227 | den-crane | It's unclear what are you going to count? number of ?
You can different tools, for example https://clickhouse.tech/docs/en/engines/table-engines/mergetree-family/summingmergetree/#nested-structures
Or do pivot and store rows instead of columns.
| 2021-01-18T19:58:44 | 2021-01-18T19:58:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,480,812 | 19,212 | sevirov | @den-crane Please, review this PR. | 2021-01-18T21:35:51 | 2021-01-18T21:35:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,501,365 | 18,617 | vzakaznikov | I'll look into fixing this.
| 2021-01-18T22:45:18 | 2021-01-18T22:45:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,508,579 | 18,063 | alexey-milovidov | ```
==1826225==ERROR: AddressSanitizer: heap-buffer-overflow on address 0x7fa38bc95800 at pc 0x00000ae19e86 bp 0x7fa2a86f5e20 sp 0x7fa2a86f5e18
#0 0xae19e85 in DB::Decimal<__int128>::operator __int128() const /home/milovidov/work/ClickHouse/build/../src/Core/Types.h:146:34
#1 0xae19e85 in void DB::AggregateF... | 2021-01-18T23:12:52 | 2021-01-18T23:12:52 | {} | MEMBER |
ClickHouse | ClickHouse | 762,509,998 | 13,348 | alexey-milovidov | There are no new reports since August. | 2021-01-18T23:18:10 | 2021-01-18T23:18:10 | {} | MEMBER |
ClickHouse | ClickHouse | 762,513,096 | 19,255 | ivanprostran | Additional info about environment and package:
- rpm: clickhouse-server-20.8.3.18-1.el7.x86_64 (altinity)
- Linux data.server.com 3.10.0-1160.6.1.el7.x86_64 #1 SMP Tue Nov 17 13:59:11 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux
- CentOS Linux release 7.9.2009 (Core)
```
Program received signal SIGSEGV, Segmentat... | 2021-01-18T23:31:20 | 2021-01-18T23:31:20 | {} | NONE |
ClickHouse | ClickHouse | 762,514,370 | 19,255 | ivanprostran | Do you suggest building from the source (latest stable) and avoid obsoleted yum repos like the one I currently use?
Thank you for your time!
Regards,
Ivan
| 2021-01-18T23:36:40 | 2021-01-18T23:36:40 | {} | NONE |
ClickHouse | ClickHouse | 762,038,051 | 19,222 | zhangjmruc | @amosbird Will the pull request #18684 fix this alias issue? | 2021-01-18T07:15:33 | 2021-01-18T07:15:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,169,220 | 14,524 | tavplubix | AFAIK MySQL 5.6 will reach its end of life in Feb 2021, it will not be longer officially supported.
Maybe we can close this issue too? | 2021-01-18T10:56:31 | 2021-01-18T10:56:31 | {} | MEMBER |
ClickHouse | ClickHouse | 762,184,690 | 19,218 | alexey-milovidov | Performance Ok, tests Ok.
Also I have validated manually that the issue is fixed. | 2021-01-18T11:22:44 | 2021-01-18T11:22:44 | {} | MEMBER |
ClickHouse | ClickHouse | 762,327,160 | 19,247 | alexey-milovidov | You should create an issue in GitHub of the Node.js driver. | 2021-01-18T15:41:46 | 2021-01-18T15:41:46 | {} | MEMBER |
ClickHouse | ClickHouse | 762,404,883 | 19,249 | UnamedRus | It's because of [insert-deduplicate](https://clickhouse.tech/docs/en/operations/settings/settings/#settings-insert-deduplicate)
```
Data blocks are deduplicated. For multiple writes of the same data block (data blocks of the same size containing the same rows in the same order), the block is only written once. The ... | 2021-01-18T18:16:30 | 2021-01-18T18:16:30 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 762,458,320 | 18,927 | alexey-milovidov | > and that prevents using clickhouse in non-root containers
If the issue is only relevant to Docker, should you just fix the Docker image? | 2021-01-18T20:31:42 | 2021-01-18T20:31:42 | {} | MEMBER |
ClickHouse | ClickHouse | 762,507,030 | 18,197 | alexey-milovidov | HashJoin.cpp, line 705
```
columns[j]->insertManyDefaults(lazy_defaults_count);
```
should use the method of IDataType, not IColumn. | 2021-01-18T23:06:30 | 2021-01-18T23:06:30 | {} | MEMBER |
ClickHouse | ClickHouse | 762,514,361 | 19,185 | alexey-milovidov | All fuzzers are Ok. | 2021-01-18T23:36:36 | 2021-01-18T23:36:36 | {} | MEMBER |
ClickHouse | ClickHouse | 762,552,536 | 19,028 | ElapsedSoul | @KochetovNicolai
It seems not easy to unenble this ,for what I have got below:
```
-- Using exception handler: unwind
-- Using libcxx: cxx
-- Using libcxxfs:
-- Using libcxxabi: cxxabi
-- Unit tests are enabled
-- Building for: Linux-4.19.90-17.ky10.aarch64 aarch64 ;
USE_STATIC_LIBRARIES=ON
MAKE_... | 2021-01-19T02:00:59 | 2021-01-19T02:00:59 | {} | NONE |
ClickHouse | ClickHouse | 760,253,605 | 18,689 | traceon | > Another question - how elapsed times are merged from multiple servers in distributed query execution?
@alexey-milovidov if they send multiple Progress packages, which are processed in same place, the elapsed time increments will be aggregated together in the common Progress snapshot. That's just a wild guess... | 2021-01-14T15:03:43 | 2021-01-14T15:03:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 760,289,798 | 19,040 | alesapin | Performance tests failure not related to changes. | 2021-01-14T15:59:56 | 2021-01-14T15:59:56 | {} | MEMBER |
ClickHouse | ClickHouse | 760,446,814 | 18,671 | alexey-milovidov | Sorry, it will be closed if it does not pass tests. | 2021-01-14T20:08:12 | 2021-01-14T20:08:12 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.