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 | 450,297,550 | 3,942 | zhang2014 | I've been tracking the feedback on this setting (there doesn't seem to be a lot of problems right now : ) #3885), this is a reply from @alexey-milovidov : https://github.com/yandex/ClickHouse/pull/2929#issuecomment-415650398 | 2018-12-28T06:13:50 | 2018-12-28T06:13:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,655,791 | 3,971 | Slind14 | They are boolean (numbers which can only be 0 and 1).
It is not possible that one row has visible and engaged set. It is always
impressionA visible 1 engaged 0
impressionA visible 0 engaged 1
`18.16.1 revision 54412` | 2018-12-31T15:06:24 | 2018-12-31T15:06:24 | {} | NONE |
ClickHouse | ClickHouse | 450,670,769 | 3,974 | den-crane | * you need **readonly 2**
2 — Read data and change settings queries are allowed.
https://clickhouse.yandex/docs/en/operations/settings/permissions_for_queries/#settings_readonly
but
* When using the GET method in the HTTP interface, readonly = 1 is set automatically.
To modify data, use the **POST** metho... | 2018-12-31T17:17:53 | 2018-12-31T17:17:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,454,154 | 3,987 | filimonov | 19840401 doesn't sound like 201812. What partition key do you use (on both tables)? | 2019-01-04T14:13:17 | 2019-01-04T14:13:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,492,126 | 3,983 | hodgesrm | @filimonov, thanks. It does not seem to be the openssl bug referenced as `openssl s_client -quiet -connect raw.githubusercontent.com:443` works without issue. But I had the same idea that it's somehow not finding the CA certs directory, which is the usual /etc/ssl/certs. I'll play around with the settings and report... | 2019-01-04T16:23:34 | 2019-01-04T18:36:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,963,394 | 3,918 | alexey-milovidov | > It looks like farmHash64 gives the best (or almost the best) performance in most of the checked cases.
This is quite interesting, because I was not sure, that we have enabled CPU dispatching in FarmHash in our build. BTW, FarmHash is intended to be not stable across different CPUs (the user should not save the res... | 2019-01-02T19:41:12 | 2019-01-02T19:41:37 | {} | MEMBER |
ClickHouse | ClickHouse | 451,357,550 | 2,534 | tungdam | Hi
> Do you see these errors in log? check ip address from the messages, it maybe some load-balancer is checking tcp port availability.
Yes, i do. And you're absolutely right, the IPs are our tcp LB hosts only, no other clients causes the same error.
But we do the healthcheck every 2 seconds, with 3seconds timeou... | 2019-01-04T05:39:11 | 2019-01-04T05:39:11 | {} | NONE |
ClickHouse | ClickHouse | 451,737,907 | 3,986 | alexey-milovidov | Some checks have failed due to GitHub API rate limit. | 2019-01-06T12:29:10 | 2019-01-06T12:29:10 | {} | MEMBER |
ClickHouse | ClickHouse | 451,870,925 | 3,830 | TH-HA | Thanks , i'll try asap and share some feedback on this.
Regards
| 2019-01-07T09:20:38 | 2019-01-07T09:20:38 | {} | NONE |
ClickHouse | ClickHouse | 450,721,570 | 3,968 | amosbird | optimized based on https://github.com/amdn/itoa . The performance gain varies from 10% to 50%
```
Testing SQL : SELECT count() FROM test.hits WHERE NOT ignore(toString(<column>)) SETTINGS max_threads = 1
Column Before : fallback hybrid After : fallback hybrid ... | 2019-01-01T10:44:12 | 2019-01-01T10:47:43 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 451,042,912 | 2,755 | iameugenejo | I'm seeing the same issue with 18.5.1. Has anyone looked into the issue yet? | 2019-01-03T02:33:28 | 2019-01-03T02:33:28 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 450,357,052 | 3,573 | healiseu | Hi Ivan (@abyss7) I would like to propose to your team associative, semiotic, hypergraph methodology. You solve efficiently the problem of secondary indexes, no need to define composite indexes, and it is extremely flexible as you can index fast all the columns of a table. It is also a good compressed size storage tech... | 2018-12-28T13:10:55 | 2018-12-28T14:43:37 | {} | NONE |
ClickHouse | ClickHouse | 450,423,526 | 3,783 | alexey-milovidov | @simPod Could you please share table structure and some amount of example data?
PS. You can use `clickhouse-obfuscator` for data anonymization. | 2018-12-28T20:39:21 | 2018-12-28T20:39:21 | {} | MEMBER |
ClickHouse | ClickHouse | 450,489,510 | 3,964 | filimonov | 1. Some types of queries can't be killed (for example if it contains scalar subquery).
2. Some of queries check the 'killed' flag only on certain moments of execution and you need to wait till query will come to that moment.
3. In versions before 18.16.0 where was no possibility to kill the query which was not starte... | 2018-12-29T12:27:24 | 2018-12-29T12:27:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,648,471 | 3,971 | den-crane | So probably meta has duplicates for some reason
SELECT impressionUUID, count() cnt
FROM `meta` WHERE date = yesterday() group by impressionUUID having cnt > 1
What table engine is used by meta ? Replacing ? | 2018-12-31T13:55:53 | 2018-12-31T13:55:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,809,012 | 3,973 | amosbird | fixed along with some comments. The style errors don't look accurate to me. | 2019-01-02T08:36:52 | 2019-01-02T09:14:13 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 451,378,294 | 3,939 | ghost | @den-crane @TH-HA
Thanks for your reply and advice. I tried to modify my sort key, and I really works for me.
Best regards and happy new year. | 2019-01-04T08:04:37 | 2019-01-04T08:04:37 | {} | NONE |
ClickHouse | ClickHouse | 451,381,236 | 3,984 | submergerock | os:centos7.5 kernel:3110
create table bpmtp.bpm_tp_ind_result_local(
tran_seq UInt64,
tran_date Date,
data_time DateTime,
ind_no String,
ind_name String,
ind_value Float64,
ccy String,
dim_one St... | 2019-01-04T08:20:29 | 2019-01-04T08:20:29 | {} | NONE |
ClickHouse | ClickHouse | 450,423,353 | 2,197 | alexey-milovidov | Fixed long time ago. | 2018-12-28T20:37:53 | 2018-12-28T20:37:53 | {} | MEMBER |
ClickHouse | ClickHouse | 450,623,919 | 3,918 | filimonov | > Numbers in MB/sec looks like MB of source data per second, where source data is something like system.numbers table. It means, that this is not MB/sec, but MHashes/sec divided by 8.
Yep. Looks like that is the reason of those weird results.
> Testing with set of strings with constant length is non-representativ... | 2018-12-31T09:20:41 | 2018-12-31T09:49:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,655,529 | 3,971 | den-crane | >Yes, these are not really duplicates, visible and engaged are treated separately.
so are they flags?
visible 0 engaged 1
visible 1 engaged 1
or numbers?
What CH version do you use?
| 2018-12-31T15:03:51 | 2018-12-31T15:03:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,961,635 | 3,918 | alexey-milovidov | > I can make benchmarks or can rewrite that tests to use that tables. Should I?
This is worth doing. Tests with
MobilePhoneModel, PageCharset, Params, URLDomain, UTMSource, Referer, URL, Title
columns should be representable.
This dataset will soon became our official dataset for benchmarks.
BTW, we also hav... | 2019-01-02T19:35:02 | 2019-01-02T19:43:00 | {} | MEMBER |
ClickHouse | ClickHouse | 451,391,377 | 2,534 | filimonov | > That means very often our clickhouse takes more than 3 seconds to finish 1 simple TCP check, which is weird.
If ClickHouse would take more than 3 sec to answer your check will go red. As I understand the only problem is the errors in CH error log, checks are green.
Generally, that error means that connection b... | 2019-01-04T09:14:43 | 2019-01-04T09:14:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,447,116 | 3,830 | yurifal | As a workaround (in Tableau), the Dimensions used on a view
(and in the LOD expressions as well) could be written as
(an example of the STRING data type):
`IFNULL( [String_Dimension], '' )`
[Link to the relevant thread on the Tableau Forums](https://community.tableau.com/message/856536) | 2019-01-04T13:44:39 | 2019-01-04T13:44:39 | {} | NONE |
ClickHouse | ClickHouse | 450,489,156 | 3,966 | filimonov | related #3828 | 2018-12-29T12:21:02 | 2018-12-29T12:21:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,581,893 | 3,968 | alexey-milovidov | Build has failed. | 2018-12-30T19:36:41 | 2018-12-30T19:36:41 | {} | MEMBER |
ClickHouse | ClickHouse | 450,590,844 | 820 | zyclonite | @proller i see a similar problem using the latest docker container (9 days old build, v18.16.1)
```ExternalDictionaries: Failed reloading 'foobar' external dictionary: Poco::Exception. Code: 1000, e.code() = 0, e.displayText() = SSL Exception: error:14004438:SSL routines:CONNECT_CR_SRVR_HELLO:tlsv1 alert internal er... | 2018-12-30T22:24:35 | 2018-12-30T22:24:35 | {} | NONE |
ClickHouse | ClickHouse | 450,663,162 | 3,968 | amosbird | Performance improves ranging from 1% - 10% . However for tiny strings (length <= 2) there are regressions.
```
Testing SQL : SELECT count() FROM test.hits WHERE NOT ignore(toString(<column>)) SETTINGS max_threads = 1
Column Before : fallback hybrid After : fallback hy... | 2018-12-31T16:11:19 | 2018-12-31T16:11:19 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 450,806,255 | 2,534 | tungdam | We have such kind of error quite a lot, every 2-3 mins. Clickhouse 18.14.18 ( Actually this error's been appearing since long time ago )
Our network are quite stable.
Could you please show me some hints to investigate this ?
Thanks in advance | 2019-01-02T08:17:37 | 2019-01-02T08:17:37 | {} | NONE |
ClickHouse | ClickHouse | 451,172,407 | 3,977 | filimonov | Theoretically it should work. You need to provide openssl parameters in client config file ([like here](https://github.com/yandex/ClickHouse/blob/0eed697623afd251c1908073c4a2850a9bfc9cdc/dbms/tests/client-test.xml)) and set up your client certificate parameters like described [here](https://pocoproject.org/docs/Poco.Ne... | 2019-01-03T15:14:31 | 2019-01-03T15:42:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,229,524 | 2,534 | den-crane | > We have such kind of error quite a lot, every 2-3 mins. Clickhouse 18.14.18 ( Actually this error's been appearing since long time ago )
> Our network are quite stable.
> Could you please show me some hints to investigate this ?
> Thanks in advance
Do you see these errors in log? check ip address from the messa... | 2019-01-03T18:14:32 | 2019-01-03T18:14:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,400,181 | 2,534 | tungdam | > If ClickHouse would take more than 3 sec to answer your check will go red. As I understand the only problem is the errors in CH error log, checks are green.
Yes, checks are green because we retry it 3 times :D before deciding it's red. This basically just makes me feel more noisy than worry ( Even though i did fee... | 2019-01-04T09:54:48 | 2019-01-04T09:54:48 | {} | NONE |
ClickHouse | ClickHouse | 450,421,783 | 1,148 | alexey-milovidov | FixedStrings as dictionary keys are still not supported.
But you can load FixedString content as a String.
Simply specify `String` as a type in dictionary structure and it will load FixedStrings from the target table.
| 2018-12-28T20:25:32 | 2018-12-28T20:25:32 | {} | MEMBER |
ClickHouse | ClickHouse | 450,447,706 | 3,954 | alexey-milovidov | Note: there is also `emptyArrayToSingle` function. | 2018-12-29T00:33:42 | 2018-12-29T00:33:42 | {} | MEMBER |
ClickHouse | ClickHouse | 450,493,721 | 3,966 | alesapin | https://github.com/yandex/ClickHouse/pull/3967/files | 2018-12-29T13:35:33 | 2018-12-29T13:35:33 | {} | MEMBER |
ClickHouse | ClickHouse | 450,576,684 | 3,966 | alesapin | It's hard to give any advice without information about amount of data in tables and disk load (`dstat` output, for example). | 2018-12-30T17:58:48 | 2018-12-30T17:58:48 | {} | MEMBER |
ClickHouse | ClickHouse | 450,582,985 | 3,971 | alexey-milovidov | If I understand your question correctly, this is how ALL JOIN is expected to behave.
(And ALL JOIN is the default behaviour of JOIN in other relational DBMSs)
If there are multiple rows in the right table with corresponding CounterID, it will create multiple rows in the result. | 2018-12-30T19:58:38 | 2018-12-30T19:58:38 | {} | MEMBER |
ClickHouse | ClickHouse | 450,585,876 | 3,971 | den-crane | >Usually this inflation is just 1-4% but in some cases it is 10-100 times.
CH always gives exact numbers. If you experience some inflation, even 0.00001% it means something wrong with design or it's some unknown CH's bug.
`GROUP BY CounterID ) USING CounterID`
means that the right table has only one row with ea... | 2018-12-30T20:52:52 | 2018-12-30T20:53:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,669,950 | 3,971 | den-crane | It's in your data. Just check what is going on for one of these (of 13) impressionUUID.
There is no mystery or magic or inflation.
>Btw. is there any way to tell ANY to pick the latest record? (go through it staring with the latest)
No.
It will be something like that
```
select mx.1 impressionUUID, sum(mx.2... | 2018-12-31T17:10:54 | 2018-12-31T17:11:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,386,702 | 3,984 | submergerock | SELECT table, create_time, command, is_done
FROM system.mutations
ORDER BY create_time DESC
table |create_time |command |is_done |
------------------|--------------------|---------------------------------... | 2019-01-04T08:51:50 | 2019-01-04T08:51:50 | {} | NONE |
ClickHouse | ClickHouse | 450,423,880 | 3,585 | alexey-milovidov | You can help if you share results of `addr2line` for addresses from the stack trace. Example:
`addr2line -iCe /usr/bin/clickhouse 0x5041235`
(the stack trace should correspond to the version installed) | 2018-12-28T20:42:05 | 2018-12-28T20:42:05 | {} | MEMBER |
ClickHouse | ClickHouse | 450,387,218 | 3,953 | healiseu | ### Without 20msec sleeping time between updates
```sql
SELECT
table,
create_time,
command,
block_numbers.partition_id AS blk_id,
block_numbers.number AS blk_num,
parts_to_do,
is_done
FROM system.mutations
ORDER BY create_time DESC
LIMIT 2
┌─table─────┬─────────creat... | 2018-12-28T16:31:43 | 2018-12-29T07:20:26 | {} | NONE |
ClickHouse | ClickHouse | 450,708,569 | 3,918 | alexey-milovidov | https://gist.githubusercontent.com/alexey-milovidov/811ce0a62cc142227e4910e525c06116/raw/dfa39ccdf9f475ccc8abd6395de959f978cbb7b2/datasets_example.txt | 2019-01-01T04:05:42 | 2019-01-01T04:05:42 | {} | MEMBER |
ClickHouse | ClickHouse | 451,113,051 | 3,978 | filimonov | Use `max_execution_time` setting or increase timeouts for http connection on your client.
( Related: #1403 ) | 2019-01-03T11:04:38 | 2019-01-03T11:05:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,748,150 | 3,783 | 4ertus2 | Reproduced on 18.16.1
https://pastebin.com/Am4VwE4E | 2019-01-06T15:01:38 | 2019-01-06T15:01:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,762,408 | 3,990 | gajus | @diafour perhaps this is something that you could contribute given your work at
https://github.com/flant/loghouse/tree/master/charts/loghouse
? | 2019-01-06T18:17:02 | 2019-01-06T18:17:02 | {} | NONE |
ClickHouse | ClickHouse | 450,405,470 | 3,959 | KochetovNicolai | #3379 | 2018-12-28T18:28:30 | 2018-12-28T18:28:30 | {} | MEMBER |
ClickHouse | ClickHouse | 450,424,515 | 3,478 | alexey-milovidov | Fixed in recent releases. | 2018-12-28T20:47:00 | 2018-12-28T20:47:00 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 450,427,988 | 2,555 | alexey-milovidov | Does it reproduce every time or sporadically?
Does it reproduce on the latest version?
I'm trying to generate some data for `agmt_table`... | 2018-12-28T21:10:12 | 2018-12-28T21:10:12 | {} | MEMBER |
ClickHouse | ClickHouse | 450,650,339 | 3,971 | den-crane | OK. Is it valid to have duplicates?
The observed difference is because these duplicates multiply select's result rows with ALL and does not with ANY. | 2018-12-31T14:15:15 | 2018-12-31T14:15:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,670,255 | 3,971 | Slind14 | I see. Thank you. | 2018-12-31T17:13:58 | 2018-12-31T17:13:58 | {} | NONE |
ClickHouse | ClickHouse | 451,128,843 | 3,978 | AlexanderJLiu | @filimonov Yes, it works, thanks. 👍
I use this setting via passing HTTP CGI parameters: `URL?max_execution_time=1`, not in config file. By this way, querying from the clickhouse console client with the same user can avoid timeout limit.
---
**Another question:** If querying a distributed table with `max_executio... | 2019-01-03T12:23:46 | 2019-01-03T12:23:46 | {} | NONE |
ClickHouse | ClickHouse | 451,381,996 | 3,984 | submergerock | clickhouse version:18.14.10 | 2019-01-04T08:24:41 | 2019-01-04T08:24:41 | {} | NONE |
ClickHouse | ClickHouse | 451,554,022 | 3,976 | den-crane | rss 25.4GB
metric | value | formatReadableSize(value)
-- | -- | --
CompiledExpressionCacheBytes | 92513984 | 88.23 MiB
CompiledExpressionCacheCount | 45757 | 44.68 KiB
jemalloc.active | 1673678848 | 1.56 GiB
jemalloc.allocated | 1653140456 | 1.54 GiB
jemalloc.mapped | 2982797312 | 2.78 GiB
jemalloc.metadata... | 2019-01-04T20:05:59 | 2019-01-04T20:05:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,442,925 | 3,783 | simPod | Would this do?
```sql
CREATE TABLE default.sensor_value ( received_at DateTime, device_id UUID, sensor_id UUID, value Nullable(Decimal(18, 4)), threshold_low_warning Nullable(Decimal(18, 4)), threshold_low_critical Nullable(Decimal(18, 4)), threshold_high_warning Nullable(Decimal(18, 4)), threshold_high_cr... | 2018-12-28T23:29:39 | 2018-12-28T23:29:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,578,935 | 3,966 | k1dav | Thx @alesapin
I will checkout about the system and setting | 2018-12-30T18:38:35 | 2018-12-30T18:38:35 | {} | NONE |
ClickHouse | ClickHouse | 450,651,090 | 3,971 | Slind14 | Yes, these are not really duplicates, visible and engaged are treated separately. So there is one row for engaged and one for visible.
Why would this inflate them so much. I'm don't think this ever happened with mysql. | 2018-12-31T14:21:57 | 2018-12-31T14:21:57 | {} | NONE |
ClickHouse | ClickHouse | 450,657,192 | 3,971 | den-crane | OK. So is any difference in result
```
SELECT
date,
count() AS impressions,
uniq(sessionUUID) as sessions,
sum(value) as value,
sum(visible) AS visible,
sum(engaged) AS engaged
FROM `impressions` ANY LEFT JOIN
(
SELECT
impressionUUID,
sum(visible) visible,
... | 2018-12-31T15:18:32 | 2018-12-31T15:18:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,657,731 | 3,971 | den-crane | >18.16.1 revision 54412
and check any difference if you execute
set compile_expressions = 0;
before query | 2018-12-31T15:22:41 | 2018-12-31T15:22:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,675,109 | 3,962 | den-crane | duplicate for #3710 | 2018-12-31T17:59:41 | 2018-12-31T17:59:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,870,359 | 3,852 | bgranvea | I am working on a patch of SummingMergeTree to solve our use case described above. The idea is to specify a custom aggregation function (here it would be 'anyLast') which will be used instead of 'sumWithOverflow'.
I have a working prototype but there are some issues:
- this custom function needs to be applied not o... | 2019-01-07T09:18:41 | 2019-01-07T09:18:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,395,056 | 1,594 | vicdashkov | Hey, could you please post your solution? | 2018-12-28T17:19:50 | 2018-12-28T17:19:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,608,940 | 3,971 | Slind14 | Here is a real query and the result:
Both tables are a simple MergeTree.
```
SELECT
date,
count() AS impressions,
uniq(sessionUUID) as sessions,
sum(value) as value,
sum(visible) AS visible,
sum(engaged) AS engaged
FROM `impressions` ANY LEFT JOIN
(
SELECT
impression... | 2018-12-31T05:08:58 | 2018-12-31T05:16:28 | {} | NONE |
ClickHouse | ClickHouse | 450,663,260 | 3,968 | amosbird | Also fixed a potential stack underflow (int128 might have 40 digits) | 2018-12-31T16:12:18 | 2018-12-31T16:12:18 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 450,790,604 | 3,973 | alexey-milovidov | Some tests have failed. | 2019-01-02T05:48:00 | 2019-01-02T05:48:00 | {} | MEMBER |
ClickHouse | ClickHouse | 451,453,478 | 3,983 | filimonov | > Note: The certificate is actually OK--Chrome shows that the URL is properly protected. I can fetch the content from the same host using curl without problems, so the host overall is fine.
Sounds like openssl client doesn't use default CA file.
You can try to set up proper path to your CA files directory (usually ... | 2019-01-04T14:10:27 | 2019-01-04T14:28:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,745,627 | 3,980 | alexey-milovidov | Functional stateless tests (unbundled): failed test looks like it contains race condition. | 2019-01-06T14:26:41 | 2019-01-06T14:26:41 | {} | MEMBER |
ClickHouse | ClickHouse | 450,778,520 | 3,963 | jasper-zhang | - I know the MV is only received the rows of the current replica, but the result between MV and T of current replica should match, also.
btw, For ReplicatedMergeTree, How can l create MV ?
- sorry, I just input error, CREATE MATERIALIZED VIEW FROM **cdn_access_log**, Actually. | 2019-01-02T02:42:43 | 2019-01-02T02:42:43 | {} | NONE |
ClickHouse | ClickHouse | 451,397,616 | 1,851 | PhantomPhreak | Still have this problem with ClickHouse v18.14.18 | 2019-01-04T09:43:00 | 2019-01-04T09:43:00 | {} | NONE |
ClickHouse | ClickHouse | 450,422,221 | 1,467 | alexey-milovidov | This place in code was completely rewritten long time ago.
Nevertheless, I will download the data and try to load into the `Log` table. | 2018-12-28T20:28:40 | 2018-12-28T20:28:40 | {} | MEMBER |
ClickHouse | ClickHouse | 450,428,312 | 2,555 | alexey-milovidov | `CREATE TABLE agmt_table ( field1 String, group_id UInt64, aggstate_field AggregateFunction(count)) ENGINE = AggregatingMergeTree() ORDER BY (group_id, field1);`
`INSERT INTO agmt_table SELECT toString(number % 11), number % 524287, countState() FROM numbers(10000000) GROUP BY number`
```
CREATE TABLE table2 E... | 2018-12-28T21:12:45 | 2018-12-28T21:12:45 | {} | MEMBER |
ClickHouse | ClickHouse | 450,474,923 | 3,742 | alesapin | Closed in favor of https://github.com/yandex/ClickHouse/pull/3598. | 2018-12-29T08:13:16 | 2018-12-29T08:13:16 | {} | MEMBER |
ClickHouse | ClickHouse | 450,493,710 | 3,828 | alesapin | https://github.com/yandex/ClickHouse/pull/3967/files | 2018-12-29T13:35:28 | 2018-12-29T13:35:28 | {} | MEMBER |
ClickHouse | ClickHouse | 450,502,108 | 3,960 | alexey-milovidov | Build in Jenkins: `Directory renamed before its status could be extracted` - unrelated failure. | 2018-12-29T15:57:39 | 2018-12-29T15:57:39 | {} | MEMBER |
ClickHouse | ClickHouse | 450,582,732 | 3,972 | alexey-milovidov | That's technically ok: implementation specific behaviour on garbage data. | 2018-12-30T19:53:29 | 2018-12-30T19:53:29 | {} | MEMBER |
ClickHouse | ClickHouse | 450,661,482 | 3,971 | Slind14 | You are right, it works with the group by. My mistake, sorry.
What I still don't get is why this usually results in an inflation of 1-4% and only on this one database of 10 times. The type of data is the same and the duplicates are not really different in those other databases.
Btw. is there any way to tell `ANY... | 2018-12-31T15:56:17 | 2018-12-31T16:41:05 | {} | NONE |
ClickHouse | ClickHouse | 450,884,933 | 3,918 | filimonov | > https://gist.githubusercontent.com/alexey-milovidov/811ce0a62cc142227e4910e525c06116/raw/dfa39ccdf9f475ccc8abd6395de959f978cbb7b2/datasets_example.txt
I can make benchmarks or can rewrite that tests to use that tables. Should I?
BTW: also some synthetic data can be used, that can be useful to create data with ... | 2019-01-02T14:59:21 | 2019-01-02T15:42:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,072,536 | 3,964 | xiashuijun | > 1. Some types of queries can't be killed (for example if it contains scalar subquery).
> 2. Some of queries check the 'killed' flag only on certain moments of execution and you need to wait till query will come to that moment.
> 3. In versions before 18.16.0 where was no possibility to kill the query which was not ... | 2019-01-03T07:39:51 | 2019-01-03T07:39:51 | {} | NONE |
ClickHouse | ClickHouse | 451,464,582 | 3,987 | yurigazinskiy | partition key is *insertdate Date DEFAULT toDate(now())* | 2019-01-04T14:51:12 | 2019-01-04T14:51:12 | {} | NONE |
ClickHouse | ClickHouse | 450,421,386 | 1,147 | alexey-milovidov | Fixed long time ago.
Now it is not necessarily that the structure of dictionary (with souce "clickhouse") completely corresponds to the structure of the target table. ClickHouse will do required type conversions on the fly. | 2018-12-28T20:22:32 | 2018-12-28T20:22:32 | {} | MEMBER |
ClickHouse | ClickHouse | 450,459,734 | 3,963 | jasper-zhang | Add: ClickHouse server version is 18.14.15 | 2018-12-29T02:49:57 | 2018-12-29T02:49:57 | {} | NONE |
ClickHouse | ClickHouse | 450,492,382 | 3,820 | alesapin | Maybe linkage became clearer, but code become messy. I don't think it's a good practice. | 2018-12-29T13:12:52 | 2018-12-29T13:12:52 | {} | MEMBER |
ClickHouse | ClickHouse | 450,536,176 | 3,286 | alexey-milovidov | Fixed in master. | 2018-12-30T03:10:56 | 2018-12-30T03:10:56 | {} | MEMBER |
ClickHouse | ClickHouse | 450,569,848 | 3,966 | k1dav | Thanks Guys
But I have another problems now...
Load Data slowly
maybe I need to edit the compression setting (is default now)?
| 2018-12-30T16:01:03 | 2018-12-30T16:01:03 | {} | NONE |
ClickHouse | ClickHouse | 450,586,629 | 3,972 | andrea-s | Thank you! What would be the recommended way to check wheter a field can be parsed as a datetime though? Since random strings of 19 characters seem to result in "valid" values... | 2018-12-30T21:06:19 | 2018-12-30T21:06:19 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 451,057,559 | 3,979 | SunguckLee | ```
SELECT version()
┌─version()─┐
│ 18.14.8 │
└───────────┘
``` | 2019-01-03T05:15:53 | 2019-01-03T05:15:53 | {} | NONE |
ClickHouse | ClickHouse | 450,304,533 | 3,943 | ghost | I'm sorry I should have looked in the error logs first, the problem is solved.
There was a problem with a certificate. | 2018-12-28T07:10:21 | 2018-12-28T07:10:21 | {} | NONE |
ClickHouse | ClickHouse | 450,342,323 | 3,891 | alesapin | Fixed here https://github.com/yandex/ClickHouse/pull/3932 | 2018-12-28T11:07:24 | 2018-12-28T11:07:24 | {} | MEMBER |
ClickHouse | ClickHouse | 450,371,222 | 3,953 | ztlpn | So what is the contents of `block_numbers.partition_id` and `block_numbers.number` columns in the case when mutations fail to complete? | 2018-12-28T14:53:11 | 2018-12-28T14:53:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,671,270 | 3,974 | edonin | Super, it is working. Thanks a lot | 2018-12-31T17:22:32 | 2018-12-31T17:22:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,084,698 | 2,521 | nanohayder | Hi,
is it possible to histogram return fixed size ranges within the number of buckets specified for example 0-5,5-10,10-15,... that would be very helpful
Thanks | 2019-01-03T08:55:27 | 2019-01-03T08:55:27 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 451,170,171 | 3,979 | den-crane | fixed in 18.16.0 https://github.com/yandex/ClickHouse/pull/3588
both queries return
lara |
-- | --
18.16.0
Workaround: set enable_optimize_predicate_expression = 0
| 2019-01-03T15:07:21 | 2019-01-03T15:07:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,747,989 | 3,783 | kbespalov | +1 the same issue | 2019-01-06T14:59:19 | 2019-01-06T14:59:19 | {} | NONE |
ClickHouse | ClickHouse | 450,503,450 | 3,953 | ztlpn | Are there any errors in ClickHouse logs? | 2018-12-29T16:18:25 | 2018-12-29T16:18:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,648,780 | 3,971 | Slind14 | > Both tables are a simple MergeTree.
Your query returns: `13 rows in set. Elapsed: 0.485 sec.` all with a cnt of 2
| 2018-12-31T13:59:42 | 2018-12-31T13:59:42 | {} | NONE |
ClickHouse | ClickHouse | 450,673,426 | 3,963 | den-crane | * T: ENGINE = ReplicatedMergeTree
MV: ENGINE = AggregatingMergeTree()
MV is a pre-insert trigger so your MV sees only rows inserted into the current replica.
* you actually have 2 different tables
CREATE TABLE **cdn_access_log**
CREATE MATERIALIZED FROM **log**
| 2018-12-31T17:43:19 | 2018-12-31T17:49:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,125,015 | 3,964 | filimonov | Related #1576 | 2019-01-03T12:04:27 | 2019-01-03T12:04:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 451,135,053 | 3,963 | jasper-zhang | My fault, I understand it wrong. | 2019-01-03T12:54:30 | 2019-01-03T12:54:30 | {} | NONE |
ClickHouse | ClickHouse | 451,538,842 | 2,534 | den-crane | >Yes, i do. And you're absolutely right, the IPs are our tcp
> LB hosts only, no other clients causes the same error.
>But we do the healthcheck every 2 seconds, with 3seconds timeout for each request.
>That means very often our clickhouse takes more than 3
> seconds to finish 1 simple TCP check, which is weird.
... | 2019-01-04T19:10:19 | 2019-01-04T19:10:19 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 450,529,906 | 3,953 | healiseu | Hi Alex, I discovered that **my error log was several GBs in size** due to processes that were writing to the (log, err.log) very frequently the same errors over and over again from other tables that were not in my default TriaDB database.
* I dropped the offending tables, cleaned databases, deleted the logs and res... | 2018-12-30T00:07:13 | 2019-01-06T16:22:37 | {} | NONE |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.