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 | 499,315,432 | 5,541 | vzakaznikov | Hi @alexey-milovidov and @alex-zaitsev, please let me know if you think these changes meet the "unintrusive" requirement. Thanks. | 2019-06-06T01:34:44 | 2019-06-06T01:34:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,594,561 | 1,661 | Slind14 | I would like to reiterate this topic because there still isn't a solution.
To summarize, we are receiving data for the same object at different times, sometimes delayed by hours. We need the latest, summed or joined data on the spot in real-time including sampling support.
Suggested approaches and their result (u... | 2019-06-06T17:37:12 | 2019-06-06T17:37:12 | {} | NONE |
ClickHouse | ClickHouse | 499,922,851 | 1,342 | inv2004 | Can we reopen the issue for Avro format ? | 2019-06-07T15:10:07 | 2019-06-07T15:10:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,834,394 | 2,045 | jondeandres | Any update on this? It's in the roadmap for next months? | 2019-06-11T13:11:57 | 2019-06-11T13:11:57 | {} | NONE |
ClickHouse | ClickHouse | 500,845,410 | 5,582 | Felixoid | Thank you, Denis.
I'm also not sure what is meant by "`carbon-clickhouse` daemon that pushes the metrics into both of them". As far as I see, `carbon-clickhouse` doesn't push metrics to a few points. | 2019-06-11T13:38:55 | 2019-06-11T13:38:55 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 499,214,870 | 5,494 | alexey-milovidov | https://github.com/yandex/ClickHouse/pull/5539 | 2019-06-05T19:06:09 | 2019-06-05T19:06:09 | {} | MEMBER |
ClickHouse | ClickHouse | 500,325,961 | 5,552 | RudanovskiyM | According to CH documentation (https://clickhouse.yandex/docs/en/query_language/create/) LIMIT shoud work:
> .... A SELECT query can contain DISTINCT, GROUP BY, ORDER BY, LIMIT... Note that the corresponding conversions are performed independently on each block of inserted data. ....
The second MV work as expec... | 2019-06-10T07:52:24 | 2019-06-10T07:52:24 | {} | NONE |
ClickHouse | ClickHouse | 500,592,174 | 5,440 | SaltTan | The number of elements now gets multiplied by the number of attributes
```
Connected to ClickHouse server version 19.7.5 revision 54419.
SELECT
type,
length(attribute.names),
element_count
FROM system.dictionaries
WHERE name = 'network'
┌─type───┬─length(attribute.names)─┬─element_count─┐
│ Hash... | 2019-06-10T20:58:35 | 2019-06-10T20:58:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,839,009 | 5,582 | filimonov | @Felixoid can you answer? | 2019-06-11T13:24:03 | 2019-06-11T13:24:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,135,652 | 5,534 | alesapin | Fixes https://github.com/yandex/ClickHouse/issues/5287. | 2019-06-05T15:34:25 | 2019-06-05T15:34:25 | {} | MEMBER |
ClickHouse | ClickHouse | 499,852,026 | 4,631 | spyros87 | > I removed the kafka parameters from the config.xml and redefine the create table statement with kafka engine with these parameters;
> `kafka_num_consumers = 4, kafka_row_delimiter = '\n', kafka_skip_broken_messages = 1, kafka_max_block_size = 1000;`
>
> Now, It seems ok. I am monitoring.
I have applied the sam... | 2019-06-07T11:34:03 | 2019-06-07T11:34:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,861,889 | 5,377 | coraxster | I made a step-by-step "How to reproduce" repo.
https://github.com/coraxster/ch-case
Hope this will help to figure out the bug.
Feel free to ask some additional info. | 2019-06-07T12:14:02 | 2019-06-07T12:14:02 | {
"+1": 4,
"confused": 1,
"eyes": 1,
"heart": 1,
"hooray": 1,
"laugh": 1,
"rocket": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,480,603 | 2,718 | 4ertus2 | link #5421 | 2019-06-10T16:20:55 | 2019-06-10T16:20:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,824,271 | 5,582 | blind-oracle | Ok I've found out where the 64s come from, I get metrics every 30s like this:
```
┌────toDateTime(Time)─┬─Value─┐
│ 2019-06-11 13:13:59 │ 64 │
│ 2019-06-11 13:14:29 │ 1168 │
│ 2019-06-11 13:14:59 │ 64 │
│ 2019-06-11 13:15:29 │ 1336 │
│ 2019-06-11 13:15:59 │ 64 │
│ 2019-06-11 13:16:29 │ 1168 │
└────... | 2019-06-11T12:43:46 | 2019-06-11T12:43:46 | {} | NONE |
ClickHouse | ClickHouse | 500,829,493 | 5,585 | den-crane | ```
CREATE TABLE x23 (
field1 String,
field2 Int64,
field3 String
) Engine = Memory;
insert into x23 format CSV 0,,aaa
OK.
``` | 2019-06-11T12:58:38 | 2019-06-11T12:58:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,837,520 | 5,585 | den-crane | >but when I try to parse the file by using
attach your file then | 2019-06-11T13:20:11 | 2019-06-11T13:20:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,706,730 | 5,519 | Chocorean | Something went through my mind this morning: as I was inserting a lot of files at the same time, maybe it crashed because of unexpected behavior during multi-insertions and not because of the file itself (because I noticed even after inserting, the engine merges and removes parts, and I remember I experienced a lot of ... | 2019-06-11T06:33:56 | 2019-06-11T14:21:20 | {} | NONE |
ClickHouse | ClickHouse | 499,368,893 | 5,538 | audricbizouarn | Hello, thank you very much for your answer. So I made the following requests:
CREATE TABLE table2_join as table2 Engine = Join(ANY, INNER, id)
INSERT INTO table2_join SELECT * from table2
ALTER TABLE table1 UPDATE name= joinGet('table2_join', 'name', id)
Everything works well. Thanks for the great work!
| 2019-06-06T06:38:28 | 2019-06-11T14:30:27 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 500,892,343 | 5,585 | den-crane | @dgrr you have 37 columns in CSV
head -1 input.csv | awk -F "," '{print NF}'
37
and 35 columns in the table. | 2019-06-11T15:24:47 | 2019-06-11T15:24:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,715,204 | 5,450 | geldot | Could this be hooked into clickhouse-client's `--external` table feature as a first step?
That would add `INSERT SELECT` support transparently to the existing documented CLI interface reusing `--structure` and its parsers.
The new table function and AST to allow stream parsing can come as a second patch once synt... | 2019-06-07T00:18:28 | 2019-06-07T00:18:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,449,783 | 5,551 | den-crane | Then it's not the #3976 compile_expressions leak | 2019-06-10T15:02:02 | 2019-06-10T15:02:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,451,655 | 4,931 | vitlibar | > <max_memory_usage> - can that tag be changed to <memory_usage> - that's due to the fact that min AND max can be set in such XML block. Also - should we check for min values and if set for less than that the user will receive the error similar to "max" constraint?
`max_memory_usage` is a name of a setting. The sett... | 2019-06-10T15:07:10 | 2019-06-10T16:25:38 | {} | MEMBER |
ClickHouse | ClickHouse | 500,765,951 | 2,542 | topvisor-admin | `CREATE TABLE test (a UInt8) Engine = Set`
`SELECT 1 FROM `anyMergeTree` WHERE `id` IN test LIMIT 1`
>DB::Exception: Method read is not supported by storage Set
Временное решение:
`SELECT 1 FROM `anyMergeTree` WHERE `id` IN test LIMIT 1 SETTINGS use_index_for_in_with_subqueries` = 1`
| 2019-06-11T09:36:16 | 2019-06-11T09:36:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,832,525 | 5,585 | dgrr | If I take the rows that give me the error individually it doesn't show any other error, but when I try to parse the file by using `cat file.csv | clickhouse-client --query="INSERT INTO x23 (field1, field2, field3) FORMAT CSVWithNames"` it shows me the error above.
I am building the newest ClickHouse version. | 2019-06-11T13:07:03 | 2019-06-11T13:08:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,853,814 | 5,569 | filimonov | Distributed inserts by default are executed asynchronously, and till 19.8 asynchronous inserts were using default settings, not the settings passed with the query.
There are 3 possible solutions:
1) use synchronous inserts (`insert_distributed_sync` setting)
2) change the setting for default profile / system prof... | 2019-06-11T13:58:24 | 2019-06-11T13:58:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,136,297 | 3,882 | alesapin | https://github.com/yandex/ClickHouse/pull/5534 probably fix | 2019-06-05T15:35:57 | 2019-06-05T15:35:57 | {} | MEMBER |
ClickHouse | ClickHouse | 499,869,099 | 5,554 | alesapin | `Code: 1000, e.code() = 2, e.displayText() = File not found: /etc/clickhouse-server/config.d/part_log.xml `
Need to merge with master | 2019-06-07T12:39:33 | 2019-06-07T12:39:33 | {} | MEMBER |
ClickHouse | ClickHouse | 499,930,256 | 5,519 | alesapin | ```
CREATE TABLE CredentialPlus
( ... | 2019-06-07T15:28:23 | 2019-06-07T15:28:31 | {} | MEMBER |
ClickHouse | ClickHouse | 500,224,800 | 5,212 | andyyzh | Implementing functions similar to those in MySQL;
## week(date[,mode])
This function returns the week number for date or datetime. The two-argument form of week() enables you to specify whether the week starts on Sunday or Monday and whether the return value should be in the range from 0 to 53 or from 1 to 53. If the... | 2019-06-09T16:06:38 | 2019-06-09T16:06:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,458,666 | 4,931 | vitlibar | About violation of the lower boundary: for the constraints from my example the following query raises an exception too:
```
SET max_memory_usage=4999999999
``` | 2019-06-10T15:26:01 | 2019-06-10T15:26:01 | {} | MEMBER |
ClickHouse | ClickHouse | 500,477,326 | 1,148 | 4ertus2 | related #4420 | 2019-06-10T16:11:46 | 2019-06-10T16:11:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,579,651 | 3,987 | den-crane | >It says ok
ATTACH PARTITION just iterates detached folder and looking for accessible and suitable parts and returns OK if found nothing.
ALTER TABLE xp ATTACH PARTITION 9999
Ok.
In your case partitions had names like '2018-12-31' and CH converted 201812 toDate and got something like '1984-04-01' | 2019-06-10T20:22:29 | 2019-06-10T20:22:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,613,739 | 5,440 | azat | As a cherry-packable fix one of the following can be done:
- increment element_count only if any/all of the setAttributeValue() returns true (i.e. insertion has been happened)
- divide element_count by number of attributes
- leave it as-is, to underline that in reality there are element_count number of elements (i.e... | 2019-06-10T22:11:13 | 2019-06-10T22:11:13 | {} | MEMBER |
ClickHouse | ClickHouse | 500,861,520 | 1,952 | filimonov | I’m closing this issue because it has been inactive for a few months. This probably means that it is not reproducible or it has been fixed in a newer version.
Please reopen if you still encounter this issue with the latest stable version.
Thank you! | 2019-06-11T14:16:24 | 2019-06-11T14:16:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,884,225 | 2,502 | proller | In last versions clickhouse-copier can upload task description to zookeeper:
`clickhouse-copier --config ./zookeeper.xml --task-path /task3/1 --task-file task3.xml`
Also possible to use ` --task-upload-force` option to overwrite current job description in zookeeper | 2019-06-07T13:27:15 | 2019-06-07T13:27:15 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,949,571 | 5,519 | den-crane | 19.5.3.8
```
cat example.csv |clickhouse-client --query "INSERT INTO CredentialPlus FORMAT CSV"
OK
=> cat example.csv |curl 'http://localhost:8123/?query=INSERT%20INTO%20CredentialPlus%20FORMAT%20CSV' --data-binary @-
=> cat example.csv |curl 'http://localhost:8123/?query=INSERT%20INTO%20CredentialPlus%20FORMA... | 2019-06-07T16:22:33 | 2019-06-07T16:23:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,473,671 | 4,968 | 4ertus2 | Alternatives #5557 #5561 | 2019-06-10T16:02:18 | 2019-06-10T16:02:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,478,275 | 4,954 | 4ertus2 | Same issue with enums #3985 | 2019-06-10T16:14:21 | 2019-06-10T16:14:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,483,641 | 4,954 | 4ertus2 | More general request #1270 | 2019-06-10T16:29:42 | 2019-06-10T16:29:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,767,438 | 5,565 | amosbird | Yeah, this error message is useless. I borrowed it from
https://github.com/yandex/ClickHouse/blob/master/dbms/src/Interpreters/Join.cpp#L1331
It definitely deserves verbosity .
| 2019-06-11T09:40:23 | 2019-06-11T09:40:23 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 500,071,635 | 146 | geldot | > We have this feature request from other customers too.
>
> Implementation is not trivial, because ALTERs that changing columns are processed outside of usual replication queue, and adding rename without reworking of ALTERs will introduce race conditions in replicated tables.
Column rename on simple non-replicat... | 2019-06-07T23:44:10 | 2019-06-07T23:44:10 | {
"+1": 4
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,071,866 | 1,342 | filimonov | > Can we reopen the issue for Avro format ?
Please open a new issue for that. | 2019-06-07T23:45:46 | 2019-06-07T23:45:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,216,418 | 5,551 | alexey-milovidov | You should upgrade to version 19.7. | 2019-06-09T14:31:24 | 2019-06-09T14:31:24 | {} | MEMBER |
ClickHouse | ClickHouse | 500,418,925 | 5,565 | 4ertus2 | It's some misleading error message, saying your key type is not supported in StorageJoin yet. StorageJoin does not support all the types JOIN does.
FYI, @amosbird
| 2019-06-10T13:35:42 | 2019-06-10T13:35:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,423,776 | 5,240 | alex-zaitsev | Mutations should not be subject of max part size at all and other merge tree settings as well. In particular, we have faced an issue, when part could not be mutated in 3600 seconds, and therefore replicas started trying to fetch it (prefer_fetch_merged_part_time_threshold = 3600) and complain that part is missing.
E... | 2019-06-10T13:51:01 | 2019-06-10T13:53:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,857,703 | 2,199 | filimonov | I’m closing this issue because it has been inactive for a few months. This probably means that it is not reproducible or it has been fixed in a newer version.
Please reopen if you still encounter this issue with the latest stable version.
Thank you! | 2019-06-11T14:07:29 | 2019-06-11T14:07:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,894,120 | 5,585 | den-crane | 0.74874870 -- `field22` Int64,
0.00000000 -- `field23` UInt64
| 2019-06-11T15:28:35 | 2019-06-11T15:28:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,915,717 | 5,417 | akuzm | As an aside, I found an interesting benchmark of some hashing algorithms: https://github.com/vnmakarov/mum-hash#intel-i7-9700k
MUM looks promising for small strings, and Meow -- for long ones. Both are relatively new, maybe we should test them in the future. | 2019-06-07T14:51:57 | 2019-06-07T14:51:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,614,531 | 2,232 | mbarbierif | Hi! This was a big problem for me, but I used Docker and now it's working fine. Here are a few steps that might help:
1. Install [Docker Desktop](https://www.docker.com/products/docker-desktop) on your Mac.
2. Create a Clickhouse container, but remember to expose both ports:
`docker run -d --name clickhouse-local -p... | 2019-06-10T22:14:20 | 2019-06-10T22:14:20 | {
"+1": 10,
"heart": 1
} | NONE |
ClickHouse | ClickHouse | 500,827,870 | 5,585 | dgrr | This error doesn't make sense to me after seeing this. https://github.com/yandex/ClickHouse/blob/master/dbms/src/IO/ReadHelpers.cpp#L545 | 2019-06-11T12:54:12 | 2019-06-11T12:54:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,867,253 | 1,821 | Felixoid | Surprisingly, it affects our hypervisors during deploy of new nodes as well... Would like to see either speed or threads limitation | 2019-06-11T14:28:45 | 2019-06-11T14:28:45 | {} | MEMBER |
ClickHouse | ClickHouse | 500,896,024 | 5,585 | dgrr | Anyway, I changed all the `Int64` and `UInt64` to `Float64`, but the error stills there:
```
...
Column 24, name: field25, type: String, parsed text: "f1ya"
Column 25, name: field26, type: Float64, parsed text: "0041"
Column 26, name: field27, type: Float64, ERROR: text ",kaaisiqnq" is not like Float64
``` | 2019-06-11T15:32:39 | 2019-06-11T15:32:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,225,055 | 5,534 | alesapin | Session pool is used only for replication: https://github.com/yandex/ClickHouse/blob/master/dbms/src/Storages/MergeTree/DataPartsExchange.cpp#L190. It doesn't affect related https://github.com/yandex/ClickHouse/issues/5508. | 2019-06-05T19:36:34 | 2019-06-05T19:36:34 | {} | MEMBER |
ClickHouse | ClickHouse | 500,070,917 | 5,518 | filimonov | Most probably the reason is explained by the difference:
simple and fast linear disk read + simple & 'brute force' vector processing
**vs**
slow disk seeks + bit more complicated processing (leading to worse cpu & cache usage) + extra data processed on the borders of the blocks | 2019-06-07T23:40:16 | 2019-06-07T23:40:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,127,300 | 5,562 | den-crane | Competing proposal: Introduce another syntax #BUFFER, which explicitly points to a source table (table_name)
CREATE INSERT TRIGGER trigger_name ON table_name TO storage_name
AS SELECT ....... FROM ( SELECT ( ...... FROM **#BUFFER** )))
Or with ability to use custom name for #BUFFER
CREATE INSERT TRIGGER trigg... | 2019-06-08T14:13:16 | 2019-06-08T14:26:13 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,756,700 | 5,552 | RudanovskiyM | Thnx | 2019-06-11T09:10:26 | 2019-06-11T09:10:26 | {} | NONE |
ClickHouse | ClickHouse | 500,816,650 | 5,584 | den-crane | Could not reproduce
```
CREATE TABLE x22 (
field1 UInt64,
field2 Float64,
field3 String
) Engine = Memory;
insert into x22 format CSV 0,0.00000000,"aaa"
OK
insert into x22 format CSV 0,"0.00000000","aaa"
OK
```
19.9.1.623 | 2019-06-11T12:21:24 | 2019-06-11T12:21:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,833,762 | 5,583 | filimonov | It uses zookeeper leader election algorithm described here:
http://zookeeper.apache.org/doc/r3.4.5/recipes.html#sc_leaderElection
For details - please check the source: https://github.com/yandex/ClickHouse/blob/master/dbms/src/Common/ZooKeeper/LeaderElection.h | 2019-06-11T13:10:18 | 2019-06-11T13:10:18 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,843,569 | 5,582 | den-crane | it could be because a function any applied not avg.
I don't see required-columns Version
and column description
```
<path_column_name>Path</path_column_name>
<time_column_name>Time</time_column_name>
<value_column_name>Value</value_column_name>
<version_column_name>Version</version_column_... | 2019-06-11T13:34:47 | 2019-06-11T13:34:47 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,882,285 | 2,902 | filimonov | I’m closing this issue because it has been inactive for a few months. This probably means that it is not reproducible or it has been fixed in a newer version.
Kafka engine had a lot of changes/fixes in 19.7 and 19.8. Please try it.
Please reopen if you still encounter this issue with the latest stable version.
... | 2019-06-11T15:02:07 | 2019-06-11T15:02:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,883,917 | 5,585 | dgrr | The file:
I simulated the file content. The shown errors are the same.
The `query` file contains the query I am doing.
The `table.sql` contains the query to build the table.
The `input.csv` is the generated file.
https://files.catbox.moe/9wnd04.zip | 2019-06-11T15:05:41 | 2019-06-11T15:05:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,128,757 | 5,562 | inv2004 | @den-crane
If you use another db.table instead of #buffer - will it use full table in the case? It will recalculate something on full table in the case, not sure if it is easy to implement, and its bad from performance point of view, but probably could cover someone's cases. | 2019-06-08T14:33:54 | 2019-06-08T14:36:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,470,879 | 5,145 | 4ertus2 | fixed in master - checkAttributeType() in DictionaryStructure.h | 2019-06-10T15:54:56 | 2019-06-10T15:54:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,881,279 | 5,481 | filimonov | Related #979 | 2019-06-11T14:59:47 | 2019-06-11T14:59:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,069,947 | 5,377 | filimonov | It reproduces with your test case. | 2019-06-07T23:32:44 | 2019-06-07T23:32:44 | {
"+1": 3,
"heart": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,158,341 | 5,534 | abraithwaite | Thanks! I'm glad you found a way that doesn't get rid of the DNS cache.
I'm curious, does this affect or relate to #5508 that Alexey opened?
I noticed that this requires `getSession` be called to reset the host IP. Are we certain that clickhouse does that everywhere before using session pools?
If it's done e... | 2019-06-05T16:30:26 | 2019-06-05T16:30:26 | {} | NONE |
ClickHouse | ClickHouse | 499,227,629 | 3,271 | iameugenejo | @den-crane There is no difference | 2019-06-05T19:44:45 | 2019-06-05T19:44:45 | {} | NONE |
ClickHouse | ClickHouse | 500,225,128 | 5,212 | andyyzh | @alexey-milovidov
Does default mode value need to be added to the system settings?
FYI:
[https://dev.mysql.com/doc/refman/5.5/en/server-system-variables.html#sysvar_default_week_format](https://dev.mysql.com/doc/refman/5.5/en/server-system-variables.html#sysvar_default_week_format) | 2019-06-09T16:11:27 | 2019-06-09T16:11:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,384,277 | 5,518 | alexm93 |
> Most probably the reason is explained by the difference:
>
> simple and fast linear disk read + simple & 'brute force' vector processing
> **vs**
> slow disk seeks + bit more complicated processing (leading to worse cpu & cache usage) + extra data processed on the borders of the blocks
Сould you please gi... | 2019-06-10T11:29:20 | 2019-06-10T11:33:06 | {} | NONE |
ClickHouse | ClickHouse | 500,431,697 | 5,557 | 4ertus2 | Some interesting point. It's slight faster to use better compressed column as right one in JOIN then LZ4. And it's much slower to use it on the left side.
Original
```
SELECT
A.datetime,
B.datetime
FROM coin.trades AS A
ANY LEFT JOIN coin.trades AS B USING (datetime)
WHERE A.datetime != B.datetime
... | 2019-06-10T14:13:19 | 2019-06-10T14:13:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,857,462 | 1,621 | filimonov | I’m closing this issue because it has been inactive for a few months. This probably means that it is not reproducible or it has been fixed in a newer version.
Please reopen if you still encounter this issue with the latest stable version.
Thank you! | 2019-06-11T14:06:56 | 2019-06-11T14:06:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,102,781 | 3,271 | den-crane | @iameugenejo
Is there any difference in results of
select count() from sample.log_stats_20181212 where timestamp = '2019-02-11 00:00:00'
select count() from sample.log_stats_20181212 where identity(timestamp) = '2019-02-11 00:00:00'
(no index / no partition elimination) | 2019-06-05T14:18:12 | 2019-06-05T14:19:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,129,608 | 5,538 | alex-zaitsev | Use Join engine table. See https://stackoverflow.com/questions/55764093/updating-rows-with-values-from-another-table-in-clickhouse | 2019-06-05T15:20:33 | 2019-06-05T15:20:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,215,051 | 5,494 | alexey-milovidov | Fixed in master and will be available in next testing release within a day. | 2019-06-05T19:06:45 | 2019-06-05T19:06:45 | {
"+1": 1,
"hooray": 1
} | MEMBER |
ClickHouse | ClickHouse | 499,349,293 | 5,196 | sundy-li | LGTM, @zhang2014 I can't comment changes due to github bugs.
```
Request URL: https://github.com/yandex/ClickHouse/pull/5196/review_comment/create
Request Method: POST
Status Code: 422 Unprocessable Entity
Remote Address: 192.30.253.113:443
Referrer Policy: no-referrer-when-downgrade
```
Some improvements
... | 2019-06-06T05:03:24 | 2019-06-06T05:03:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,893,894 | 5,552 | den-crane | It just does not have any sense to use LIMIT inside MV.
MV it's an insert trigger which performs select over insert buffer and MV never reads real source table.
So when if you insert one row `VALUES('2019-05-27', 200, 121, 0.811);` MV query and limit will be evaluated over exactly this 1 ROW.
When you insert 10 mi... | 2019-06-07T13:54:33 | 2019-06-07T13:56:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,130,948 | 5,562 | den-crane | AS **_TTT** -- I meant a better usability to do not introduce a new keyword #BUFFER.
Sometimes I need this use-case (upsert/insert if not exists) for small tables -- dimensions:
CREATE INSERT TRIGGER trigger_name ON table_name AS _TTT TO storage_name
AS SELECT * FROM _TTT WHERE key not in (SELECT KEY FROM storag... | 2019-06-08T15:02:24 | 2019-06-08T15:03:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,132,011 | 5,562 | inv2004 | @den-crane
I suppose you can simulate WHERE key not in (SELECT FROM storage) by creating kv Join engine + mv_to_kv into it +joinGet in trigger_name mv . But it cannot help to solve initial problem mentioned in the issue. | 2019-06-08T15:17:33 | 2019-06-08T15:17:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,132,733 | 5,562 | den-crane | One more idea:
CREATE INSERT TRIGGER trigger_name ON table_name AS _TTT
**INTO storage_name(a,b,c,d)**
AS SELECT some_expression_without_alias, v, c,d FROM ( SELECT ( ...... FROM _TTT ))) | 2019-06-08T15:27:47 | 2019-06-08T15:34:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,828,951 | 5,585 | javisantana | could you change field2 to Nullable(Int64) and test again? | 2019-06-11T12:57:08 | 2019-06-11T12:57:08 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,870,460 | 5,582 | blind-oracle | @den-crane
> it could be because a function any applied not avg.
It is `avg`. I've cited the `graphite_rollup` part of `config.xml`
> I don't see required-columns Version and column description
The docs are wrong for a long time.
CH expects `Timestamp` instead of `Version` by default.
See https://githu... | 2019-06-11T14:35:57 | 2019-06-11T14:35:57 | {} | NONE |
ClickHouse | ClickHouse | 500,898,437 | 5,582 | Felixoid | Ok, I've seen in the code, it accepts a slice of upload configs.
Here's the only usage of time during merge, it is used to calculate precision https://github.com/yandex/ClickHouse/blob/master/dbms/src/DataStreams/GraphiteRollupSortedBlockInputStream.cpp#L110
Are you sure that `carbon-clickhouse` does manipulate a... | 2019-06-11T15:37:54 | 2019-06-11T15:37:54 | {} | MEMBER |
ClickHouse | ClickHouse | 499,136,143 | 4,970 | alesapin | https://github.com/yandex/ClickHouse/pull/5534 fix | 2019-06-05T15:35:36 | 2019-06-05T15:35:36 | {
"hooray": 1
} | MEMBER |
ClickHouse | ClickHouse | 499,766,204 | 4,931 | kagan770 | <max_memory_usage> - can that tag be changed to <memory_usage> - that's due to the fact that min AND max can be set in such XML block. Also - should we check for min values and if set for less than that the user will receive the error similar to "max" constraint? | 2019-06-07T05:54:26 | 2019-06-07T05:54:26 | {} | NONE |
ClickHouse | ClickHouse | 499,795,015 | 5,545 | spyros87 | Maybe is related to #4631 | 2019-06-07T08:01:07 | 2019-06-07T08:01:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,842,616 | 5,551 | filimonov | That warning is related to regular expression engine.
Do you have some complicated regexps in your DEFAULT expressions?
Can you check what exactly ClickHouse was doing by examing `/var/log/clickhouse-server/clickhouse-server.log` and provide the create statement of the table which was processed at that time? | 2019-06-07T10:53:35 | 2019-06-07T10:56:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,278,361 | 5,458 | yingfeng | This issue has been resolved by pr [5528](https://github.com/yandex/ClickHouse/pull/5528) | 2019-06-10T03:15:18 | 2019-06-10T03:15:18 | {} | NONE |
ClickHouse | ClickHouse | 500,332,759 | 5,551 | lfundaro | > Seems like CompiledExpressions leak.
> Check select * from system.asynchronous_metrics where metric = 'CompiledExpressionCacheCount'
> #3976
> And set` <compile_expressions>0</compile_expressions>` in user's profiles.
this one gives me nothing as output. I was checking the issue you mentioned and the graphs the... | 2019-06-10T08:18:26 | 2019-06-10T08:18:26 | {} | NONE |
ClickHouse | ClickHouse | 500,347,789 | 5,569 | maskshell | @alexey-milovidov
@BayoNet | 2019-06-10T09:10:58 | 2019-06-10T09:10:58 | {} | NONE |
ClickHouse | ClickHouse | 500,816,730 | 5,545 | spyros87 | @CurtizJ This issue occurred on version 19.7.3.9 too. | 2019-06-11T12:21:41 | 2019-06-11T12:21:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,294,640 | 5,470 | blinkov | @BayoNet there are some merge issues | 2019-06-05T23:38:34 | 2019-06-05T23:38:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,431,798 | 5,529 | blinkov | @BayoNet some merge issues | 2019-06-06T10:01:28 | 2019-06-06T10:01:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,800,086 | 5,182 | cce | re: the mention of OpenTracing and OpenCensus as alternatives in the above comments, the two projects are actually merging:
https://medium.com/opentracing/merging-opentracing-and-opencensus-f0fe9c7ca6f0
https://medium.com/opentracing/a-roadmap-to-convergence-b074e5815289
The two projects have similar tracing APIs ... | 2019-06-07T08:20:24 | 2019-06-07T08:20:24 | {
"heart": 2
} | NONE |
ClickHouse | ClickHouse | 499,890,527 | 5,551 | den-crane | Seems like CompiledExpressions leak.
Check select * from system.asynchronous_metrics where metric = 'CompiledExpressionCacheCount'
https://github.com/yandex/ClickHouse/issues/3976
And set` <compile_expressions>0</compile_expressions>` in user's profiles. | 2019-06-07T13:45:13 | 2019-06-07T13:45:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 499,920,882 | 5,417 | amosbird | I doubt any of those would beat CRC32 due to hardware acceleration. But it might be useful for `serialized` method when better hash functions are required. | 2019-06-07T15:05:04 | 2019-06-07T15:05:04 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 499,979,777 | 5,557 | 4ertus2 | Test
```
CREATE TABLE coin.trades_t64
(
collector_id UInt8 CODEC(T64,LZ4),
market Enum8('BINANCE' = 1, 'BITMEX' = 2, 'BITTREX' = 3, 'BITFINEX' = 4, 'HUOBI' = 5, 'ZB' = 6, 'OKEX' = 7, 'HITBTC' = 8, 'LBANK' = 9, 'BIBOX' = 10, 'BITZ' = 11, 'KRAKEN' = 12, 'BITSTAMP' = 13, 'POLONIEX' = 14, 'GATEIO' = 15, 'YOB... | 2019-06-07T17:54:15 | 2019-06-07T18:39:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,438,762 | 5,552 | den-crane | >Or when i make batch insert MV process record by record? If so why does the second MV work fine?
It's processed by blocks (max_insert_block_size/max_block_size/min_insert_block_size_rows/min_insert_block_size_bytes), and it depends of the nature of the source and partitioning. | 2019-06-10T14:32:31 | 2019-06-10T14:32:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,463,325 | 2,135 | 4ertus2 | related #1640 | 2019-06-10T15:38:02 | 2019-06-10T15:38:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 500,857,054 | 1,573 | filimonov | I’m closing this issue because it has been inactive for a few months. This probably means that it is not reproducible or it has been fixed in a newer version.
Please reopen if you still encounter this issue with the latest stable version.
Thank you! | 2019-06-11T14:05:58 | 2019-06-11T14:05:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,280,494 | 6,075 | abyss7 | Should fix #5846 - or at least the new logs will shred some light on the real cause of the problem. | 2019-07-19T15:51:25 | 2019-07-19T15:51:25 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.