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 | 484,432,804 | 5,030 | alexey-milovidov | @YiuRULE You should also add dependency to Dockerfiles in `docker` directory, whenever appropriate.
(there are Dockerfiles that are used by our CI) | 2019-04-18T09:53:14 | 2019-04-18T09:53:56 | {} | MEMBER |
ClickHouse | ClickHouse | 484,881,057 | 2,402 | abyss7 | It's possible to have multiple views on the same kafka table - just tested it on 19.5. The cascading mat. views is not a kafka-specific feature, so it's better to create a new separate issue for this feature. | 2019-04-19T12:33:34 | 2019-04-19T12:33:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,120,756 | 5,002 | ipoese | Only non consistent news so far.
We ran the dev version you supplied for about 24 hours without any crashes and then had to revert back to the production version as the performance is really... low. The estimate is that we're running about 1/100th of the "normal" version
After the switch back clickhouse (19.4.6)... | 2019-04-20T13:36:30 | 2019-04-20T13:36:40 | {} | NONE |
ClickHouse | ClickHouse | 485,201,358 | 5,067 | kmatt | Attempting service restart with init.d script results in multiple concurrent server daemons consuming small amounts of CPU, clickhouse-server must be manually killed before starting new:
```
$ ps aux | grep clickhouse-server
clickho+ 7811 1.2 0.1 4131708 184396 ? Ssl 21:30 1:54 clickhouse-server --daemo... | 2019-04-21T00:01:58 | 2019-04-21T03:49:46 | {} | NONE |
ClickHouse | ClickHouse | 484,074,573 | 3,579 | bicubic | I am having the same issue and haven't found any combination of `kafka_row_delimiter` and data that works. | 2019-04-17T13:01:40 | 2019-04-17T13:01:40 | {} | NONE |
ClickHouse | ClickHouse | 484,124,564 | 5,016 | alexey-milovidov | @alesapin TTL test: debt. | 2019-04-17T14:53:23 | 2019-04-17T14:53:23 | {} | MEMBER |
ClickHouse | ClickHouse | 484,175,658 | 4,846 | alexey-milovidov | 00933_ttl_simple in tsan - TODO @alesapin | 2019-04-17T17:00:03 | 2019-04-17T17:00:03 | {} | MEMBER |
ClickHouse | ClickHouse | 484,278,034 | 5,042 | alexey-milovidov | > when mergetable is created with order by desc? IRC his is allowed.
No, it's not allowed. You can write `ORDER BY -x` instead (but it doesn't make sense). | 2019-04-17T22:02:34 | 2019-04-17T22:02:34 | {} | MEMBER |
ClickHouse | ClickHouse | 484,458,959 | 4,190 | maddhuapplicate | Does Clickhouse dictionary support Mongo cluster as source? Most production systems will be on clusters with SSL enabled. When we tried the cluster URI as host, Clickhouse did not connect to the cluster. | 2019-04-18T11:13:05 | 2019-04-18T11:15:13 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 484,867,465 | 5,048 | ogorbacheva | Изначально я планировала только добавить описание для arraySort/arrayReverseSort. Но в итоге решила пересортировать все методы (вернее группы методов), так как:
- Сортировки по условным критериям усложняют восприятие и навигацию. Чтобы найти нужный метод, приходится глазами просматривать весь список. Когда методов н... | 2019-04-19T11:41:04 | 2019-04-19T11:41:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,260,830 | 1,226 | nvartolomei | @alexey-milovidov do you have a preference for alter materialized view query syntax? I might give it a try. | 2019-04-21T15:33:04 | 2019-04-21T15:33:04 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,375,610 | 5,002 | alesapin | Ok, thank you. | 2019-04-22T09:30:23 | 2019-04-22T14:19:50 | {} | MEMBER |
ClickHouse | ClickHouse | 484,053,606 | 5,029 | den-crane | Just use select ... format JSONEachRow to see how CH expects data for input
```
select * from (
select 1 row, 2 int_sample, [1,2] int_array_sample
union all
select 2 row, 0 int_sample, [4,5] int_array_sample)
format JSONEachRow
{"row":1,"int_sample":2,"int_array_sample":[1,2]}
{"row":2,"int_sample":0,"int_a... | 2019-04-17T12:00:50 | 2019-04-17T12:00:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,169,671 | 5,013 | alexey-milovidov | Nevertheless, LZ4 should be upgraded to latest. | 2019-04-17T16:44:00 | 2019-04-17T16:44:00 | {} | MEMBER |
ClickHouse | ClickHouse | 484,197,944 | 5,013 | alexey-milovidov | https://github.com/lz4/lz4/commit/219abab74bf8914d3f8761db5b398103cbbd77d7#commitcomment-27520069 | 2019-04-17T17:59:48 | 2019-04-17T17:59:48 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 475,826,284 | 975 | yu45020 | Thanks all for previous discussions, and here is an alternative way to perform running difference per group
Suppose we have the following data
```sql
create view nid_date as
select nid, date
from data
group by nid, date
order by nid, date
limit 20;
```
| nid | date |
| ------------- | ------------- ... | 2019-03-23T01:07:36 | 2019-04-20T11:38:36 | {} | NONE |
ClickHouse | ClickHouse | 485,100,697 | 975 | yu45020 | @alexz0000
Do you mean running difference within group ? One way to do it is to turn each group into array, then use running difference, then use array join to flat them back into a column. It will be easier if you provide some sample data. | 2019-04-20T11:44:28 | 2019-04-20T11:44:28 | {} | NONE |
ClickHouse | ClickHouse | 485,447,543 | 4,211 | jelharou | Is there a workaround for this? | 2019-04-22T15:20:14 | 2019-04-22T15:20:14 | {} | NONE |
ClickHouse | ClickHouse | 484,069,358 | 5,029 | Gillani0 | Thanks Denis.
So there is no way to process Json arrays of objects in CH. | 2019-04-17T12:47:35 | 2019-04-17T12:47:35 | {} | NONE |
ClickHouse | ClickHouse | 484,126,477 | 5,030 | alexey-milovidov | > Also I wasnt' really sure for to things, with two domain from the test, one who have co.yandex and org.рф, I wasn't able to find this domain both on the list and also doing a quick research from Google, so let me know if I should add them on the list too.
These domains fit to the older logic of this function, but ... | 2019-04-17T14:57:42 | 2019-04-17T14:57:42 | {} | MEMBER |
ClickHouse | ClickHouse | 484,516,127 | 5,028 | alexey-milovidov | Also cherry-picked to #4884 | 2019-04-18T13:46:30 | 2019-04-18T13:46:30 | {} | MEMBER |
ClickHouse | ClickHouse | 484,550,370 | 5,032 | filimonov | > So for this, any solution or any option is there to handle this?
I see the following solutions:
1) check if you use clickhouse in a proper use case (using OLAP database for certain scenarios can be ineffective)
2) optimize your schema / usage patterns if possible (follow @den-crane questions)
3) scale cluster ... | 2019-04-18T15:11:56 | 2019-04-18T15:11:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,157,127 | 5,066 | den-crane | The same with 19.7.1.360
```
./crash.sh
1
2
3
4
5
6
7
Error on processing query: SELECT DISTINCT arrayJoin(extractAll(name, '[\w_]{2,}')) AS res FROM (SELECT name FROM system.functions UNION ALL SELECT name FROM system.table_engines UNION ALL SELECT name FROM system.formats UNION ALL SELECT name FROM system... | 2019-04-20T19:38:34 | 2019-04-20T20:58:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,226,475 | 2,797 | evead | +1 | 2019-04-21T05:39:30 | 2019-04-21T05:39:30 | {} | NONE |
ClickHouse | ClickHouse | 485,336,714 | 5,068 | sundy-li | Great, it's much better now. | 2019-04-22T06:15:10 | 2019-04-22T06:15:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,152,410 | 5,022 | 4ertus2 | You need to batch inserts. ClickHouse expects one INSERT statement per second or less. Each of them should contain lots rows of data. If you're inserting your data correctly and face with full hardware utilisation you probably need a better hardware. | 2019-04-17T15:59:52 | 2019-04-17T15:59:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,794,274 | 5,003 | Enmk | > Ok.
> BTW, functions with base32 output should not be the only available variants.
> UInt64 variant will be more ClickHouse friendly and base32 variants can be provided for convenience.
Ok, could you please specify the function name and signature?
Also, I am wondering about use cases for putting that into the d... | 2019-04-19T07:41:46 | 2019-04-19T07:41:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,845,008 | 5,003 | alex-zaitsev | > Ok.
> BTW, functions with base32 output should not be the only available variants.
> UInt64 variant will be more ClickHouse friendly and base32 variants can be provided for convenience.
Geohash is kind of standard (see https://en.wikipedia.org/wiki/Geohash). Only base32 and base36 encodings are used, the latter ... | 2019-04-19T10:31:44 | 2019-04-19T10:31:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,879,729 | 5,013 | danlark1 | We tested a bit new LZ4 and it turned out to be slightly faster than our own implementation. We will investigate more and maybe update our LZ4 decompression with referenced implementation | 2019-04-19T12:28:59 | 2019-04-19T12:30:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,002,326 | 4,872 | nvartolomei | Workaround, if you don’t want to wait for this feature/fix: generate query ids, then sleep for few seconds flush logs and get timings from query log table. | 2019-04-19T20:10:31 | 2019-04-19T20:10:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,003,607 | 4,518 | alexey-milovidov | That's a race condition.
`SYSTEM FLUSH LOGS` should flush both queues to ensure that all previous queries are written to logs. | 2019-04-19T20:16:28 | 2019-04-19T20:16:28 | {} | MEMBER |
ClickHouse | ClickHouse | 485,269,825 | 5,013 | danlark1 | https://github.com/yandex/ClickHouse/pull/5070 | 2019-04-21T17:44:52 | 2019-04-21T17:44:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,402,906 | 5,074 | alexey-milovidov | You must download all the submodules.
```
git submodule init
git submodule update
```
We don't support linking with the OS "mysqlclient" library.
Linking with libraries from the OS is always discouraged. | 2019-04-22T12:08:32 | 2019-04-22T12:08:32 | {} | MEMBER |
ClickHouse | ClickHouse | 484,278,374 | 5,042 | alexey-milovidov | @anrodigina It's not merged with master, that's why all builds have failed. | 2019-04-17T22:03:50 | 2019-04-17T22:03:50 | {} | MEMBER |
ClickHouse | ClickHouse | 484,453,442 | 4,890 | abyss7 | Confirmed the problem. Trying to fix. | 2019-04-18T10:57:23 | 2019-04-18T10:57:23 | {
"hooray": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,873,849 | 5,048 | blinkov | Имхо основной плюс сортировки по условному убыванию популярности в том, что когда ничего не знаешь по теме статью можно читать по порядку и к тому моменту как тебе это надоест ты будешь знать самое важное, а не какое-то случайное подмножество по алфавиту. А чтобы когда люди уже в теме ищут что-то конкретное, то надо не... | 2019-04-19T12:06:35 | 2019-04-19T12:06:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,875,517 | 4,890 | abyss7 | Looks like the duplicate of #4736 | 2019-04-19T12:13:09 | 2019-04-19T12:13:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,014,067 | 4,970 | andrei-dascalu | @filimonov Unfortunately in the meantime the cluster where this persistent problem happened was removed as we had to replace it to get things going. However, I will try to reconstruct it on a smaller scale as the issue happened virtually every time when running in Kubernetes and I will investigate a bit further. It's b... | 2019-04-19T21:05:28 | 2019-04-19T21:05:28 | {} | NONE |
ClickHouse | ClickHouse | 485,135,894 | 4,646 | AlexPereverzyev | Some observations -
Case 1: no messages/load
Every second _BackgroundSchedulePoolTaskInfo_ logs trace message - "Execution took ~2500 ms" (6 active consumers).
Prometheus metric _clickhouse_memory_tracking_in_background_schedule_pool_ grows constantly.
Overall memory leak depends on number of active Kafka... | 2019-04-20T15:27:23 | 2019-04-20T15:41:45 | {} | NONE |
ClickHouse | ClickHouse | 485,283,417 | 5,066 | den-crane | The minimal
cat crash.sql
```
SELECT name FROM system.tables
UNION ALL
SELECT name FROM system.columns
format Null;
``` | 2019-04-21T21:25:54 | 2019-04-21T21:25:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,159,295 | 4,860 | aadant | @filimonov : MySQL and Postgres natively support datetime(6) and timestamp(6) that have microsecond precision. If you want to replicate them, you need to support this data type. I think MariaDB column store just added it. https://mariadb.com/kb/en/library/columnstore-data-types/ | 2019-04-17T16:17:15 | 2019-04-17T16:17:15 | {} | NONE |
ClickHouse | ClickHouse | 484,169,264 | 5,013 | 4ertus2 | #5040 | 2019-04-17T16:42:58 | 2019-04-17T16:42:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,193,736 | 5,013 | llchan | Very interesting, didn't know clickhouse had a custom decompressor. What does Yann Collett think? Is there a chance lz4 could eventually include an adaptive decompressor? | 2019-04-17T17:48:42 | 2019-04-17T17:48:42 | {} | NONE |
ClickHouse | ClickHouse | 484,866,947 | 4,248 | nvartolomei | @Junus most likely this fixes it. https://github.com/edenhill/librdkafka/commit/cb4be1fc5158dd7502a46d7e1d43560a3c291700
This is present in v19.5+ for lower version you'll have to build it yourself. If you're using v19.3+ then you also should be aware of https://github.com/yandex/ClickHouse/issues/4736. | 2019-04-19T11:38:31 | 2019-04-19T11:38:31 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,157,505 | 5,066 | den-crane | select name,value from system.settings where changed;
```
┌─name───────────────────┬─value───────┐
│ use_uncompressed_cache │ 0 │
│ load_balancing │ random │
│ max_memory_usage │ 10000000000 │
└────────────────────────┴─────────────┘
``` | 2019-04-20T19:40:46 | 2019-04-20T20:58:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,246,319 | 5,068 | blinkov | cc @zhang2014 @sundy-li | 2019-04-21T12:05:20 | 2019-04-21T12:05:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,075,499 | 5,029 | bicubic | Between this and #3579 there seems to be no way to pass a collection of records per kafka message at all.
Edit: actually it is possible.
Like implied above, the JSON in each message must be not valid json. I have had success sending multiple JSON rows with messages such as this:
```
b'{"timestamp": 1555504526... | 2019-04-17T13:04:12 | 2019-04-17T15:21:43 | {} | NONE |
ClickHouse | ClickHouse | 484,171,581 | 5,013 | alexey-milovidov | Video (in russian): https://www.youtube.com/watch?v=V2CqQBICt7M
Presentation (in russian): https://yandex.github.io/clickhouse-presentations/highload_siberia_2018/# | 2019-04-17T16:49:19 | 2019-04-17T16:49:19 | {} | MEMBER |
ClickHouse | ClickHouse | 484,305,301 | 5,043 | champtar | for now `-DGLIBC_COMPATIBILITY=OFF` | 2019-04-18T00:01:47 | 2019-04-18T00:01:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,346,329 | 5,027 | Giant1 | No , I didn't . I want to save the logs they are repeatable when using ReplicatedMergeTree engine . " Deduplication parameters are controlled by merge_tree server settings " , Do you mean I should reduce the " replicated_deduplication_window " parameter ? | 2019-04-18T03:53:51 | 2019-04-18T03:53:51 | {} | NONE |
ClickHouse | ClickHouse | 484,461,464 | 5,003 | Enmk | > How this implementation compares to https://github.com/yinqiwen/geohash-int/ in terms of performance?
Good question, lets find it out! One thing though... that repo lacks support of base32 encoding and decoding. And for geohashing non-standard alphabet is used and we can't just reuse existing base32-coding from el... | 2019-04-18T11:20:39 | 2019-04-18T11:21:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,078,591 | 4,518 | jangorecki | Thanks for explanations.
I agree with Alexey, `SYSTEM FLUSH LOGS` should flush all queues used to log queries. I will proceed my development assuming such behaviour. Hopefully it will be addressed sooner than later, so I can include clickhouse in https://h2oai.github.io/db-benchmark/ once it is resolved. | 2019-04-20T09:32:22 | 2019-04-20T12:06:25 | {} | NONE |
ClickHouse | ClickHouse | 484,050,391 | 5,024 | den-crane | Just use **request_time** in a partitioning expression and primary key expression.
create table (
....)
partition by **toYYYYMM(request_time)**
order by **field_name_1, request_time**
CH able to eliminate partitions and uses a table order (index organized table) for locating rows. | 2019-04-17T11:49:23 | 2019-04-17T11:50:10 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,095,107 | 5,032 | den-crane | 1. How many columns in your table?
2. How many rows you insert per day?
3. How many rows you insert per batch?
4. Do you use buffer tables?
5. How many disks (HDD or SSD) in your server?
6. How many CPUs in your server?
7. Can you show your table DDL (at least PARTITION BY , ORDER BY sections) | 2019-04-17T13:42:40 | 2019-04-17T13:43:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,875,356 | 4,890 | abyss7 | > @abyss7 thanks! While you're dealing with that, do you mind explaining or defining CH behaviour if it can't process kafka events as quickly as they are being published?
>
> At high rate mismatch (continuous 10m/s publish, CH only processing about 1m/s), I find that 90% of the data seems to get blackholed never to ... | 2019-04-19T12:12:29 | 2019-04-19T12:12:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,382,079 | 975 | alexz0000 | @yu45020 @filimonov Thank you. My table looks like below
| ts | env | group | instance | domain | url | val |
| --------------- | --------- | ----------- | ------------ | ----------- | ------------ | ----- |
| 1555926155 | public | blu... | 2019-04-22T10:03:44 | 2019-04-22T11:22:04 | {} | NONE |
ClickHouse | ClickHouse | 485,399,247 | 5,002 | nvartolomei | Can this be caused by bad memory? Are you using ECC memory? | 2019-04-22T11:45:49 | 2019-04-22T11:45:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,087,193 | 5,027 | den-crane | Did you check *Replicated data deduplication ?
Try the same with simple MergeTree
https://clickhouse.yandex/docs/en/operations/table_engines/replication/?highlight=deduplicated
Data blocks are deduplicated. For multiple writes of the same data block (data blocks of the same size containing the same rows in the s... | 2019-04-17T13:21:38 | 2019-04-17T13:21:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,169,359 | 5,013 | alexey-milovidov | We are using our own decompression loop that is significantly faster than in LZ4:
https://github.com/yandex/ClickHouse/blob/master/dbms/src/Compression/LZ4_decompress_faster.h
https://github.com/yandex/ClickHouse/blob/master/dbms/src/Compression/LZ4_decompress_faster.cpp
Yann Collett is aware of our improvements... | 2019-04-17T16:43:12 | 2019-04-17T16:43:12 | {} | MEMBER |
ClickHouse | ClickHouse | 484,521,108 | 5,046 | den-crane | The problem that one may need to use external columns in lambda function
SELECT sum(c1) AS v
FROM
(
SELECT
1 AS c1,
['v'] AS c2,
['d'] **as d**
)
WHERE arrayExists(i -> (**d** = ['d']), c2) | 2019-04-18T13:57:51 | 2019-04-18T13:59:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,781,637 | 5,030 | YiuRULE | Thanks, I add `gperf` dependency both on `binary` and `deb` folder.
Otherwise it seem that we still have a problem with the build, it seem that CI fetch the docker image from the one in the [docker hub](https://hub.docker.com/r/yandex/clickhouse-deb-builder) right?
Otherwise, I tried running locally the script `p... | 2019-04-19T06:34:32 | 2019-04-19T07:33:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,836,418 | 5,048 | blinkov | @ogorbacheva в чем задумка? на вид там было вручную отсортировано условно по убыванию популярности, а ты хочешь по алфавиту сделать? | 2019-04-19T10:01:43 | 2019-04-19T10:01:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,920,328 | 5,048 | ogorbacheva | > когда ничего не знаешь по теме статью можно читать по порядку
Это удобно, когда в статье перечислено 5-10 методов, но не 30. Когда пользователь открывает статью, он видит в дереве (справа) 100500 методов, следовательно, сразу фокусируется на дереве. Читать последовательно он ничего не будет. Когда он попытается со... | 2019-04-19T14:53:47 | 2019-04-19T14:54:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,091,126 | 5,002 | alesapin | Any updates? | 2019-04-20T10:52:07 | 2019-04-20T10:52:07 | {} | MEMBER |
ClickHouse | ClickHouse | 484,198,469 | 5,003 | alexey-milovidov | How this implementation compares to https://github.com/yinqiwen/geohash-int/ in terms of performance? | 2019-04-17T18:01:13 | 2019-04-17T18:01:13 | {} | MEMBER |
ClickHouse | ClickHouse | 484,441,240 | 5,030 | YiuRULE | @alexey-milovidov Ok thanks for your help, I added the dependency in the `builder` folder, I think it should be ok now | 2019-04-18T10:22:39 | 2019-04-18T10:22:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,995,240 | 4,872 | jangorecki | any chance to have this feature incorporated? seems to be pretty straightforward. `query_log` doesn't seems to be reliable, as explained in linked issue. | 2019-04-19T19:38:04 | 2019-04-19T19:38:04 | {} | NONE |
ClickHouse | ClickHouse | 485,003,206 | 4,872 | alexey-milovidov | Also you can use `grep -P '^\d'` to preprocess your file. | 2019-04-19T20:14:33 | 2019-04-19T20:14:33 | {} | MEMBER |
ClickHouse | ClickHouse | 485,255,561 | 5,043 | champtar | Hi @proller
It's not fixed, it fails for each binary it seems
```
[ 50%] Linking CXX executable zstd_test
[ 50%] Built target Crypto
[ 52%] Built target Net
c++: error: libs/libglibc-compatibility/libglibc-compatibility.a: Aucun fichier ou dossier de ce type
make[2]: *** [utils/compressor/zstd_test] Erreur 1
m... | 2019-04-21T14:21:15 | 2019-04-21T14:21:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,275,763 | 5,042 | nvartolomei | On top of that, what happens when mergetree is created with order by desc? IRC it is allowed. | 2019-04-17T21:54:15 | 2019-04-17T21:57:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,455,965 | 4,890 | bicubic | @abyss7 thanks! While you're dealing with that, do you mind explaining or defining CH behaviour if it can't process kafka events as quickly as they are being published?
At high rate mismatch (continuous 10m/s publish, CH only processing about 1m/s), I find that 90% of the data seems to get blackholed never to be se... | 2019-04-18T11:04:31 | 2019-04-18T11:04:31 | {} | NONE |
ClickHouse | ClickHouse | 484,502,502 | 5,045 | alexey-milovidov | Let's wait for performance tests to run (we have `performance/url/url_hits.xml` for these functions). | 2019-04-18T13:14:30 | 2019-04-18T13:14:30 | {} | MEMBER |
ClickHouse | ClickHouse | 484,573,421 | 5,027 | den-crane | I meant need to understand the root cause. Try to reproduce the issue against MergeTree. | 2019-04-18T16:01:20 | 2019-04-18T16:01:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,057,767 | 975 | alexz0000 | Hi, Could anyone help me here ?
Let's say I have a table and it has columns ts, env, instance, value. the env only has two values 'env1' and 'env2', and instance has 10 values 'instance1', 'instance2' ...., How could I runningDifference for the group of env and instance?
| 2019-04-20T04:33:52 | 2019-04-20T04:34:05 | {} | NONE |
ClickHouse | ClickHouse | 485,437,508 | 5,069 | alexey-milovidov | There are new PVS warnings:
```
General Analysis | FillingBlockInputStream.cpp:71 | High | V730 | Not all members of a class are initialized inside the constructor. Consider inspecting: last_row_ref.
General Analysis | LimitBlockInputStream.cpp:10 | High | V730 | Not all members of a class are initialized inside t... | 2019-04-22T14:43:42 | 2019-04-22T14:43:42 | {} | MEMBER |
ClickHouse | ClickHouse | 485,449,855 | 4,935 | zhang2014 | > Using of same queries for Replicated and Distributed tables is confusing. It would be better to add another query, for example SYSTEM SYNC DISTRIBUTED, SYSTEM STOP DISTRIBUTED SENDS, etc.
Done
> Also I'm not sure that this type of queries for Distributed tables are useful. It only triggers additional findFile a... | 2019-04-22T15:28:12 | 2019-04-22T15:28:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,175,513 | 4,846 | alexey-milovidov | > I don't see any differences in the PVS report, that you mentioned earlier in Telegram Chat. Can you point out what new errors/warnings were introduced in this PR?
There is no difference in PVS report for now. | 2019-04-17T16:59:41 | 2019-04-17T16:59:41 | {} | MEMBER |
ClickHouse | ClickHouse | 484,664,530 | 5,003 | alexey-milovidov | Ok.
BTW, functions with base32 output should not be the only available variants.
UInt64 variant will be more ClickHouse friendly and base32 variants can be provided for convenience. | 2019-04-18T19:52:44 | 2019-04-18T19:52:44 | {} | MEMBER |
ClickHouse | ClickHouse | 484,771,436 | 5,055 | filimonov | Duplicate https://github.com/yandex/ClickHouse/issues/1821 | 2019-04-19T05:25:45 | 2019-04-19T05:25:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,774,162 | 5,051 | filimonov | Does the select itself (i.e. without insert) finishes successfully? | 2019-04-19T05:45:59 | 2019-04-19T05:45:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,055,655 | 5,029 | Gillani0 | Hi den,
It expects it to start with `{` and it's complaining when it starts with `[` for a Json array.
Syd | 2019-04-17T12:07:34 | 2019-04-17T12:07:34 | {} | NONE |
ClickHouse | ClickHouse | 484,517,405 | 4,015 | sjoerdmulder | @alexey-milovidov @ztlpn This bug is blocking us from upgrading to newer versions; Is it possible to fix or revert the changes done by #3673 since it clearly causing this regression | 2019-04-18T13:49:15 | 2019-04-18T13:49:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,994,101 | 4,518 | jangorecki | It looks like there is a bug somewhere there. Simplified reproducible example below:
```sh
clickhouse-client --query="CREATE TABLE IF NOT EXISTS G1 (id1 Int32, id2 Int32) ENGINE = Memory()"
echo -e "SET log_queries = 1;\nSELECT id1, sum(id2) AS id2 FROM G1 GROUP BY id1;\nSET log_queries = 0;\nSYSTEM FLUSH LOGS;\nSEL... | 2019-04-19T19:32:39 | 2019-04-19T19:32:39 | {} | NONE |
ClickHouse | ClickHouse | 485,111,098 | 975 | filimonov | Please check that https://stackoverflow.com/questions/51856397/clickhouse-running-diff-with-grouping | 2019-04-20T12:40:02 | 2019-04-20T12:40:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,433,629 | 5,071 | alexey-milovidov | Missing test. | 2019-04-22T14:30:04 | 2019-04-22T14:30:04 | {} | MEMBER |
ClickHouse | ClickHouse | 484,057,539 | 5,029 | den-crane | https://clickhouse.yandex/docs/en/interfaces/formats/ Check input format column
https://clickhouse.yandex/docs/en/interfaces/formats/#jsoneachrow
>It expects it to start with { and it's complaining when it starts with [ for a Json array.
Yes CH expects the data as a whole is not valid JSON
| 2019-04-17T12:14:09 | 2019-04-17T12:14:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,066,249 | 3,797 | mkabilov | if you're using postgresql >= 10, you might look into [pg2ch](https://github.com/mkabilov/pg2ch) project | 2019-04-17T12:43:07 | 2019-04-17T12:43:07 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,374,629 | 4,970 | filimonov | > @filimonov thanks for looking into my question. Unfortunately the `SYSTEM DROP DNS CACHE` query didn't seem to do much. Are you referring to this query?
Yes.
So once again:
1) pod was recreated
2) ip changed
3) replication stuck, and SYSTEM DROP DNS CACHE doesn't help
Did I understand that correctly?
... | 2019-04-18T06:34:19 | 2019-04-18T06:34:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,484,323 | 5,043 | proller | Fixed: https://github.com/yandex/ClickHouse/commit/ad42179bece7788007d308ae7fff8f4f6e1d9341 | 2019-04-18T12:27:19 | 2019-04-18T12:27:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,512,436 | 5,030 | alexey-milovidov | It's not enough:
https://clickhouse-builds.s3.yandex.net/5030/5c67ecd46d803f97908563b27fc9cf9a95f7b393/build_log_415594408_1555584319.txt
As far as I remember, you should add to `packager` Dockerfiles. | 2019-04-18T13:38:09 | 2019-04-18T13:38:09 | {} | MEMBER |
ClickHouse | ClickHouse | 484,703,193 | 5,013 | alexey-milovidov | @danlark1 has found additional ASan report (in addition to https://github.com/lz4/lz4/issues/676) in internal Yandex CI (tests across all Yandex codebase). It will be reported.
(`LZ4_wildCopy32` read 16 bytes after end of buffer) | 2019-04-18T21:56:48 | 2019-04-18T21:56:48 | {} | MEMBER |
ClickHouse | ClickHouse | 484,803,703 | 5,014 | 243f6a8885a308d313198a2e037 | I am not a ClickHouse expert (one of a recent user), but I have seen a scenario like this before.
At that time, "XID overflow" was the direct cause of this issue.
(Restarting the server will resolve the error (and this is what I did at that time), but some auto-recovery mechanism should be prepared against this sit... | 2019-04-19T08:15:16 | 2019-04-19T08:15:16 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 484,832,834 | 5,051 | alexm93 | > Does the select itself (i.e. without insert) finishes successfully?
Yes, this query falls with Memory limit exception.
`SELECT
sign AS sign,
argMax(emission_type, version) AS emission_type,
argMax(batch, version) AS batch
FROM test
GROUP BY sign
HAVING sum(ch_sign) > 0
LIMIT 1` | 2019-04-19T09:49:59 | 2019-04-19T09:49:59 | {} | NONE |
ClickHouse | ClickHouse | 484,884,466 | 2,675 | abyss7 | > @abyss7 thanks for that attempt.
>
> Now with clickhouse:19.1.6 kafka engine _silently_ drops messages both with kafka_skip_broken_messages = 1 and kafka_skip_broken_messages = 0.
>
> Is there any change to log errors for skipped messages?
Can't confirm your problem at least for 19.5. The `kafka_skip_broken_... | 2019-04-19T12:45:30 | 2019-04-19T12:45:30 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,001,233 | 4,518 | nvartolomei | There are 2 stages. Need to sleep a bit before calling flush logs.
Logs are inserted into a queue first. Then in background moved to another queue. Second queue is flushed periodically to the table or on flush logs. | 2019-04-19T20:05:21 | 2019-04-19T20:05:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,003,081 | 4,872 | alexey-milovidov | > any chance to have this feature incorporated?
It's quite easy to implement, the chance is high. But this feature request is not assigned to anyone in my team right now. | 2019-04-19T20:13:58 | 2019-04-19T20:13:58 | {} | MEMBER |
ClickHouse | ClickHouse | 485,004,236 | 5,013 | alexey-milovidov | Happy enough, there are two totally different ways to improve performance. There is a chance that we can combine both. | 2019-04-19T20:19:18 | 2019-04-19T20:19:18 | {} | MEMBER |
ClickHouse | ClickHouse | 484,129,123 | 4,890 | bicubic | I observe similar issues.
Pushing 1m events burst is lossless.
Pushing 10m events burst results in partial loss to the materialized view with missing rows
`kafka_max_block_size = 1` Significantly slows down the population rate but still loses tens of thousands of rows.
I am detecting missing rows by row cou... | 2019-04-17T15:03:45 | 2019-04-17T15:09:08 | {} | NONE |
ClickHouse | ClickHouse | 484,877,434 | 2,251 | abyss7 | Can't reproduce on recent CH versions. | 2019-04-19T12:20:32 | 2019-04-19T12:20:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 485,452,448 | 5,074 | tumurzakov | Cmake wont start build process without thouse two steps.
I installed poco from contrib folder
I symlink some include dirs to system one because didnt work out to add target_include directories in corresponding CMakeLists
I built clickhouse without mysql support successfully
But this problem has arisen and i... | 2019-04-22T15:37:20 | 2019-04-22T15:37:20 | {} | NONE |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.