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 | 517,253,731 | 6,246 | filimonov | Please use build from our CI (you can find links to builds next to each commit in github).
For example - current master:
https://clickhouse-builds.s3.yandex.net/0/eeeab85bbe01db807f97a7a673e561f6b308cdf5/report.html
Pick debs from clang-7 / relwithdebuginfo / address.
Look into stderr or run server in terminal... | 2019-08-01T11:56:37 | 2019-08-01T11:56:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,275,505 | 6,249 | 4ertus2 | There's some UB. "DB::Exception: Wrong columns nullability." in gcc builds only. | 2019-08-01T12:56:17 | 2019-08-01T12:56:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,343,001 | 6,246 | filimonov | > The schema is part of a project the sponsors intends to open source in the future, but unfortunately we cannot currently share.
That asan report is useful.
Schema sample would also be helpful - we don't need real columns names, something like should be enough:
```
CREATE TABLE t1 (c1 UInt8, c2 String, c3 ..... | 2019-08-01T15:42:37 | 2019-08-01T15:43:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,818,090 | 6,248 | den-crane | >Comparisons to string work so I re-interpreted your question as:
Please don't guess and check exactly my query!!!
select id, toFloat64(id) = 439385272.0 as compare_indicator
from myTable
where identity(id) = '439385272' | 2019-08-02T19:28:16 | 2019-08-02T19:28:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,087,773 | 6,209 | alexey-milovidov | Continued in #6255 | 2019-08-01T02:00:16 | 2019-08-01T02:00:16 | {} | MEMBER |
ClickHouse | ClickHouse | 517,252,778 | 5,846 | josepowera | 19.11.4.24 is out on https://hub.docker.com/r/yandex/clickhouse-server/tags | 2019-08-01T11:54:00 | 2019-08-01T11:54:00 | {} | NONE |
ClickHouse | ClickHouse | 517,289,242 | 5,042 | MahmoudGoda0 | Yes, its start_time only, but why its expected ?
Will do statistics and share the results with you.
| 2019-08-01T13:31:37 | 2019-08-01T13:31:37 | {} | NONE |
ClickHouse | ClickHouse | 517,312,948 | 6,246 | banburybill | I tried the current master build as you linked.
The build originally worked, but failed when I started the incoming data feed. I then realised I needed llvm-symbolizer installed, so after finding the LLVM APT repo installed it and restarted. clickhouse-server fails immediately with the following.
```
Aug 1 15:1... | 2019-08-01T14:31:08 | 2019-08-01T14:31:08 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 517,577,988 | 6,273 | pachico | Thanks for the quick reply.
It is quite unfortunate but I take comfort in the word "yet" :)
In the meanwhile, I have found a workaround moving the sub query logic to a materialized column in the source table so the materialized view does not require it.
Just in case it might be useful to someone else.
Needless to... | 2019-08-02T06:51:18 | 2019-08-02T06:51:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,679,977 | 6,278 | alesapin | Failed tests are not connected with changes. | 2019-08-02T12:23:10 | 2019-08-02T12:23:10 | {} | MEMBER |
ClickHouse | ClickHouse | 517,715,511 | 6,294 | ghost | It is worth to mention that it also might be possible to report progress in [HTTP chunked encoding chunk extensions](https://tools.ietf.org/html/rfc7230#section-4.1.1), which allows to set the progress after some data was sent, unlike updating of the headers that is implemented by `send_progress_in_http_headers` option... | 2019-08-02T14:11:06 | 2019-08-02T14:17:22 | {} | NONE |
ClickHouse | ClickHouse | 517,427,337 | 6,271 | alexey-milovidov | Sorry for misguidance, but switch/case doesn't work at all in this case. | 2019-08-01T19:35:21 | 2019-08-01T19:35:21 | {} | MEMBER |
ClickHouse | ClickHouse | 517,502,977 | 6,256 | alexey-milovidov | It looks like `path` became 19% slower. Let's check it carefully with `clickhouse-benchmark` and think what we can do... | 2019-08-02T00:25:44 | 2019-08-02T00:25:44 | {} | MEMBER |
ClickHouse | ClickHouse | 517,644,805 | 6,293 | den-crane | workaround
```
DESCRIBE TABLE
(
SELECT CAST(a, 'String') = 'foo'
FROM test
)
┌─name─────────────────────────────┬─type──┬─default_type─┬─default_expression─┬─comment─┬─codec_expression─┬─ttl_expression─┐
│ equals(CAST(a, 'String'), 'foo') │ UInt8 │ │ │ │ ... | 2019-08-02T10:07:34 | 2019-08-02T10:07:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,727,352 | 6,297 | 4ertus2 | Fixes #6116 | 2019-08-02T14:44:09 | 2019-08-02T14:44:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,890,173 | 6,298 | alexey-milovidov | Checked perf tests manually. | 2019-08-03T03:23:00 | 2019-08-03T03:23:00 | {} | MEMBER |
ClickHouse | ClickHouse | 517,150,210 | 6,253 | hczhcz | It breaks the default cmake settings on my computer (linuxmint 19, with g++ & libstdc++) since they have not moved `<experimental/filesystem>` to `<filesystem>` yet.
For now, `-DUSE_LIBCXX=1 -DCMAKE_C_COMPILER=/usr/bin/clang -DCMAKE_CXX_COMPILER=/usr/bin/clang++` works.
| 2019-08-01T07:05:37 | 2019-08-01T07:05:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,237,898 | 6,237 | proller | Its a bug!
https://github.com/yandex/ClickHouse/pull/6264 | 2019-08-01T11:09:04 | 2019-08-01T11:09:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,634,496 | 6,274 | CurtizJ | Seems that failed tests: `00834_kill_mutation, 00834_kill_mutation_replicated_zookeeper` are broken in master.
| 2019-08-02T09:36:17 | 2019-08-02T09:36:17 | {} | MEMBER |
ClickHouse | ClickHouse | 517,648,673 | 6,216 | BolshakovNO | Thank you | 2019-08-02T10:20:03 | 2019-08-02T10:20:03 | {} | NONE |
ClickHouse | ClickHouse | 517,701,027 | 6,272 | ghost | @filimonov In this issue I'm mostly concerned about JSON formatting of the errors, but it can be combined with multiplexing of the data.
I like the idea of HTTP headers from #4971. Issue #902 is mostly the same as this one.
Currently I'm trying to add formatting options to [getCurrentExceptionMessage](https://git... | 2019-08-02T13:29:16 | 2019-08-02T13:29:35 | {} | NONE |
ClickHouse | ClickHouse | 517,889,859 | 6,285 | alexey-milovidov | > Runtime dispatching on x64 processors
Does it require to compile all translation units with `-mavx2`?
```
// Determining which implementation is the more suitable
// Should be done at runtime. Does not make any sense on preprocessor.
#ifdef __AVX2__
instruction_set best_implementation = instruction_se... | 2019-08-03T03:16:47 | 2019-08-03T03:16:58 | {} | MEMBER |
ClickHouse | ClickHouse | 517,033,872 | 6,251 | alexey-milovidov | TODO: add to documentation (after we fix all the remaining issues: #5815). | 2019-07-31T21:45:17 | 2019-07-31T21:45:17 | {} | MEMBER |
ClickHouse | ClickHouse | 517,085,964 | 6,238 | quyaxing | Thank you very much for your reply！
I would like to ask what the specific function of this configuration item uncompressed_cache_size is. Can't it be used to control the query cache?
| |
屈亚星
|
|
quyaxing23@126.com
|
签名由网易邮箱大师定制
On 08/1/2019 00:59，Denis Zhuravlev<notifications@github.com> wrote：
There is no query ... | 2019-08-01T01:50:48 | 2019-08-01T01:50:48 | {} | NONE |
ClickHouse | ClickHouse | 517,295,832 | 5,042 | den-crane | >Yes, its start_time only, but why its expected ?
Because full scan and order by of millions rows is faster than millions sequential index jumps (in all databases, it will be the same in mysql and PG). | 2019-08-01T13:48:29 | 2019-08-01T13:49:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,671,730 | 5,846 | Rostish | We have updated our clickhouse too.
We are experiencing next issiue:
i have about 150 records in my kafka topic, but clickhouse is consuming it infinite times, i have already about 500 000 rows im my table. | 2019-08-02T11:50:27 | 2019-08-02T11:50:27 | {} | NONE |
ClickHouse | ClickHouse | 517,776,579 | 6,272 | ghost | @filimonov I've just noticed that implementations of server-sent events in browsers [don't support HTTP methods other than `GET`](https://stackoverflow.com/a/34285526). So for `POST` requests, such as `INSERT` queries, some other mechanism of reporting errors still might be useful. | 2019-08-02T17:06:38 | 2019-08-02T17:06:38 | {} | NONE |
ClickHouse | ClickHouse | 517,263,938 | 5,846 | simpl1g | we have updated to this version half an hour ago, and currently we are experiencing issues with consuming messages, everything stuck | 2019-08-01T12:25:07 | 2019-08-01T12:25:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,274,938 | 6,259 | filimonov | Duplicate
https://github.com/yandex/ClickHouse/issues/5125
https://github.com/yandex/ClickHouse/issues/5409
https://github.com/yandex/ClickHouse/issues/4159
https://github.com/yandex/ClickHouse/issues/1178
https://github.com/yandex/ClickHouse/issues/3312
and others... | 2019-08-01T12:54:46 | 2019-08-01T12:54:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,674,053 | 6,272 | filimonov | See also #4971 #902 | 2019-08-02T12:00:10 | 2019-08-02T12:01:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,919,898 | 6,166 | kreuzerkrieg | hm... I've added the `-Wshadow` to the CMakeList and got a bunch of errors like
>> /home/Development/ClickHouse/dbms/src/Common/RWLock.h:71:35: error: declaration of ‘type’ shadows a member of ‘DB::RWLockImpl::Group’ [-Werror=shadow]
explicit Group(Type type) : type{type} {}
or
>> /home/Development/Clic... | 2019-08-03T12:10:18 | 2019-08-03T12:10:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,301,637 | 6,271 | alexey-milovidov | But this error may happen due to incorrect data on server side, that is independent of the user request. | 2019-08-01T14:02:58 | 2019-08-01T14:02:58 | {} | MEMBER |
ClickHouse | ClickHouse | 517,306,138 | 6,271 | ghost | The idea was to use HTTP status codes to let the clients distinguish between retriable and non-retriable errors, so that if a client received 4xx code it would not retry the same request and if the client received 5xx code it would retry the same request again. | 2019-08-01T14:14:32 | 2019-08-01T14:14:32 | {} | NONE |
ClickHouse | ClickHouse | 517,777,075 | 6,252 | reflection | Thanks @kaijianding, we're fairly sure that was the issue now. We have another database that is exhibiting inode exhaustion:
```
df -h /var/lib/clickhouse/
Filesystem Size Used Avail Use% Mounted on
/dev/sdk 976M 214M 747M 23% /var/lib/clickhouse
df -i /var/lib/clickhouse/
Filesystem Inod... | 2019-08-02T17:08:10 | 2019-08-02T17:08:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,823,536 | 6,248 | AprilSun007 | Yes. The query you suggested works. Referring back to my first code example (shown below), even without the identity function, the query works when it is comparing to a string.
```
select id, toFloat64(id) = 439385272.0 as compare_indicator
from myTable
where id = '439385272'
-- query returns one matching row
```... | 2019-08-02T19:48:05 | 2019-08-02T19:48:05 | {} | NONE |
ClickHouse | ClickHouse | 517,823,565 | 6,297 | alexey-milovidov | Don't understand - why is was crashing and whether the root cause was fixed. | 2019-08-02T19:48:10 | 2019-08-02T19:48:10 | {} | MEMBER |
ClickHouse | ClickHouse | 517,907,539 | 6,308 | ivanprostran | I confirm this behavior, I had to drop 2 partitions ('2019-08-03', '2019-08-02'). 1.5 Billion records.
Now downgraded to : clickhouse-server-19.11.3.11-1.el7.x86_64
We use clickhouse in production. Kafka engine drives us crazy.
I Don't know what to do (perhaps bulk insert through clickhosue-client as temporary s... | 2019-08-03T08:52:13 | 2019-08-03T08:52:13 | {} | NONE |
ClickHouse | ClickHouse | 517,245,535 | 6,266 | jbwl | Please disregard and close this. It was added by mistake. | 2019-08-01T11:34:11 | 2019-08-01T11:34:11 | {} | NONE |
ClickHouse | ClickHouse | 517,312,383 | 6,271 | alexey-milovidov | Ok, got it. | 2019-08-01T14:29:48 | 2019-08-01T14:29:48 | {} | MEMBER |
ClickHouse | ClickHouse | 517,733,370 | 6,290 | filimonov | Actually ClickHouse supports timezones from tzdata package, i.e. known in OS level.
https://unix.stackexchange.com/questions/504043/how-to-list-timezones-known-to-the-system
There is no timezone named `Z` in tzdata, but if you need you can easely add symlink in appropriate folder, and such a timezone will appear.... | 2019-08-02T15:00:06 | 2019-08-02T15:00:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,782,939 | 6,252 | alexey-milovidov | OS and libc doesn't give us better message. But we may craft our own with additional `fstatvfs` call. This is quite easy to implement. | 2019-08-02T17:27:12 | 2019-08-02T17:27:12 | {} | MEMBER |
ClickHouse | ClickHouse | 517,845,388 | 2,743 | adhamelia | We are running clickhouse on Butter FS and performance got increased almost
3 times than ext4 and ZFS.
ZFS was giving us much better compression but no performance gain.
On Friday, August 2, 2019, Taras Shevchenko <notifications@github.com>
wrote:
> @adhamelia <https://github.com/adhamelia> Do you have any results, ... | 2019-08-02T21:10:27 | 2019-08-02T21:10:27 | {} | NONE |
ClickHouse | ClickHouse | 517,925,861 | 6,308 | luck143 | > I confirm this behavior, I had to drop 2 partitions ('2019-08-03', '2019-08-02'). 1.5 Billion records.
> Now downgraded to : clickhouse-server-19.11.3.11-1.el7.x86_64
> We use clickhouse in production. Kafka engine drives us crazy.
> I Don't know what to do (perhaps bulk insert through clickhosue-client as tempora... | 2019-08-03T13:45:51 | 2019-08-03T13:45:51 | {} | NONE |
ClickHouse | ClickHouse | 517,215,872 | 6,184 | filimonov | Can you provide some concrete samples of datasets in sas format?
Formal format description
https://cran.r-project.org/web/packages/sas7bdat/vignettes/sas7bdat.pdf
Java Parso libraby mentioned:
https://github.com/epam/parso
Pure Python implementation (also include a tool to convert sas7bdat to csv)
https://... | 2019-08-01T09:57:27 | 2019-08-01T09:58:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,274,261 | 5,274 | filimonov | #2765 | 2019-08-01T12:53:07 | 2019-08-01T12:53:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,300,135 | 5,042 | MahmoudGoda0 | Actually I cant get your point, but the create of my table as below :
create table tsnew(
ext_rec_num Nullable(UInt64),
xdr_id Nullable(UInt64),
xdr_grp Nullable(UInt64),
xdr_type Nullable(UInt64),
xdr_subtype Nullable(Int16),
xdr_direction Nullable(Int16),
xdr_location Nullable(Int16),
time UInt64,
stop_time... | 2019-08-01T13:59:23 | 2019-08-01T13:59:23 | {} | NONE |
ClickHouse | ClickHouse | 517,477,383 | 6,273 | den-crane | CH does not have full support of subselect in MV yet. https://github.com/yandex/ClickHouse/issues/2878 | 2019-08-01T22:14:22 | 2019-08-01T22:14:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,625,304 | 6,291 | blinkov | You can always see what's changed via `git log` or GitHub web interface like https://github.com/yandex/ClickHouse/compare/19.7...19.11 , while manually reviewed [CHANGELOG.md](https://github.com/yandex/ClickHouse/blob/master/CHANGELOG.md) is produced as fast as we physically can. | 2019-08-02T09:09:38 | 2019-08-02T09:09:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,646,960 | 6,241 | den-crane | Your issue is unrelated to the @filimonov ideas probably he missed that you have the JOIN.
53 millions WIDE rows easily able to eat 500GB of memory because of straightforward JOIN you used. | 2019-08-02T10:14:32 | 2019-08-02T10:14:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,648,100 | 6,289 | filimonov | Currently it is not possible. When you run it as root it will finish with exception. See: https://github.com/yandex/ClickHouse/pull/3785#discussion_r240163579
Can i ask why do you need to run clickhouse as root? | 2019-08-02T10:18:03 | 2019-08-02T10:18:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,737,328 | 2,743 | tshev | @adhamelia Do you have any results, which are worth sharing? | 2019-08-02T15:10:42 | 2019-08-02T15:10:42 | {} | NONE |
ClickHouse | ClickHouse | 517,082,965 | 6,250 | alexey-milovidov | 00626_replace_partition_from_table_zookeeper - flappy test. | 2019-08-01T01:34:26 | 2019-08-01T01:34:26 | {} | MEMBER |
ClickHouse | ClickHouse | 517,111,049 | 6,028 | alexey-milovidov | BTW, we can do better in this variant too:
- across all columns of minimum compressed size, select a column with minimum size of value in memory:
```
┌─database─┬─table────────────┬─name──────────────────┬─type───────────┬─default_kind─┬─default_expression─┬─data_compressed_bytes─┬─data_uncompressed_bytes─┬─marks_... | 2019-08-01T04:10:58 | 2019-08-01T04:10:58 | {} | MEMBER |
ClickHouse | ClickHouse | 517,303,805 | 5,042 | den-crane | @MahmoudGoda0
>Actually I cant get your point, but the create of my table as below :
My point you don't have index which is appropriate for your search condition and order by and TOP 1000.
Just try to create another table with `order by(imsi, time)` and test your search.
In case of `order by(time)` databas... | 2019-08-01T14:08:22 | 2019-08-01T14:16:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,721,355 | 6,127 | akuzm | I noticed that we get an empty array for point (45, 45, 45, 45). Is this expected?
```
┌────n─┬─geohashesInBox(n, n, n, n, 1)─┐
│ 44.9 │ ['s'] │
│ 45 │ [] │
│ 45.1 │ ['v'] │
└──────┴───────────────────────────────┘
``` | 2019-08-02T14:27:30 | 2019-08-02T14:27:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,747,679 | 6,248 | AprilSun007 | Thanks for the help,
Comparisons to string work so I re-interpreted your question as:
Does this work?
```
select id, toFloat64(id) = 439385272.0 as compare_indicator
from myTable
where identity(toFloat64(id)) = 439385272.0
```
And the answer is yes it does! Does this help you see why the case withou... | 2019-08-02T15:38:56 | 2019-08-02T15:44:21 | {} | NONE |
ClickHouse | ClickHouse | 517,041,019 | 2,894 | alexey-milovidov | Don't understand these changes. | 2019-07-31T22:09:10 | 2019-07-31T22:09:10 | {} | MEMBER |
ClickHouse | ClickHouse | 517,098,901 | 6,145 | alexey-milovidov | It's git usability issue. Submodules are not updated on merge or pull.
You always have to update them manually with the commands above.
You can blame git for this. | 2019-08-01T03:00:18 | 2019-08-01T03:00:18 | {} | MEMBER |
ClickHouse | ClickHouse | 517,099,507 | 6,111 | alexey-milovidov | It's not possible to have user names with dot characters for now.
But it isn't a fundamental issue, it can be fixed. | 2019-08-01T03:03:20 | 2019-08-01T03:03:20 | {} | MEMBER |
ClickHouse | ClickHouse | 517,113,049 | 6,258 | alexey-milovidov | https://clickhouse-test-reports.s3.yandex.net/0/a109339ce600835fb8437b47f8015c4e479c6311/performance_test/comparison_min_time_gcc_8.html
`SELECT count() FROM hits_100m` query.
It's also significantly slower at formatting Pretty formats, and dealing with DateLUT. | 2019-08-01T04:22:10 | 2019-08-01T04:22:10 | {} | MEMBER |
ClickHouse | ClickHouse | 517,245,797 | 6,215 | peaksnail | 👍 | 2019-08-01T11:34:55 | 2019-08-01T11:34:55 | {} | NONE |
ClickHouse | ClickHouse | 517,287,884 | 5,042 | den-crane | @MahmoudGoda0 what is order by on your table? ( imsi, start_time ) or just start_time?
(I think in case of start_time it is expected behavior).
Can you show extended statistics with [send_logs_level = debug] | 2019-08-01T13:28:10 | 2019-08-01T13:28:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,606,769 | 6,135 | igor-sh8 | Hi.
Sorry, we didn’t determine what the problem was, so we closed tests with `Protobuf` (it was already before your reply).
And for now we just decided to use a different format, in our case `Native`.
Thanks. | 2019-08-02T08:18:07 | 2019-08-02T08:18:07 | {} | NONE |
ClickHouse | ClickHouse | 517,636,192 | 6,241 | Guilherme-Routar | @den-crane JOIN's right table has around 53 million records. And each `snippet` can be quite big. I tried using a subquery as you mentioned and it works perfectly.
Although I would like to understand whether `adaptive index granularity` is being used in my tables. Is there a way I can look into this? @filimonov | 2019-08-02T09:41:19 | 2019-08-02T09:44:24 | {} | NONE |
ClickHouse | ClickHouse | 517,676,335 | 3,443 | 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! | 2019-08-02T12:09:06 | 2019-08-02T12:09:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,783,432 | 6,252 | alexey-milovidov | https://github.com/yandex/ClickHouse/issues/6182 | 2019-08-02T17:28:47 | 2019-08-02T17:28:47 | {} | MEMBER |
ClickHouse | ClickHouse | 517,890,027 | 6,285 | alexey-milovidov | PS. JSON Pointer is great, we should add support for it in ClickHouse :) | 2019-08-03T03:20:08 | 2019-08-03T03:20:08 | {} | MEMBER |
ClickHouse | ClickHouse | 517,209,880 | 6,251 | blinkov | > TODO: add to documentation (after we fix all the remaining issues: #5815).
there's already some documentation on this, we could adapt it without waiting for anything
cc @BayoNet | 2019-08-01T09:39:50 | 2019-08-01T09:39:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,490,964 | 6,275 | den-crane | try to insert with max_insert_block_size=100000 (with MV) (as a test) | 2019-08-01T23:18:21 | 2019-08-02T06:46:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,643,019 | 6,293 | MetalRex101 | Clickhouse converts LowCardinality(String) to LowCardinality(Int8) and then to LowCardinality(UInt64) instead of usual Uint64 | 2019-08-02T10:01:53 | 2019-08-02T10:01:53 | {} | NONE |
ClickHouse | ClickHouse | 517,653,787 | 6,275 | den-crane | Can it be that user `default` and your current user have different max_bytes_before_external_group_by settings?
Check and share output of your insert with send_logs_level = 'debug' | 2019-08-02T10:38:45 | 2019-08-02T10:38:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,670,804 | 6,272 | filimonov | @a-rodin and what do you think about bit more universal approach of multiplexing HTTP data streams with Server-Sent Events?
See #6294 | 2019-08-02T11:46:18 | 2019-08-02T11:46:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,127,823 | 6,246 | filimonov | Is it possible to share the schema or to install on problematic server special clickhouse debug build (with address sanitizer, will work much slower) and try to reproduce the problem in that debug version? | 2019-08-01T05:39:26 | 2019-08-01T05:39:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,219,883 | 6,246 | banburybill | The schema is part of a project the sponsors intends to open source in the future, but unfortunately we cannot currently share. I can, however, certainly install a debug build locally and reproduce the problem. Do I need to build this debug build myself, or is it available somewhere? | 2019-08-01T10:09:19 | 2019-08-01T10:09:19 | {} | NONE |
ClickHouse | ClickHouse | 517,280,396 | 6,241 | filimonov | ClickHouse always use batch/vector processing. I.e. even if you need one row, ClickHouse can't process less than index_granularity rows (by default 8192), if rows are big - you can hit the memory limit when trying to put 8K row in memory.
So index_granularity is minimum size of the block (it just can access smaller... | 2019-08-01T13:07:54 | 2019-08-01T13:07:54 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,437,504 | 6,271 | nvartolomei | Clients should not retry 500 status code either. Only 503 should be retried IMHO and work should be done to correctly report these. | 2019-08-01T20:05:08 | 2019-08-01T20:05:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,660,642 | 6,287 | den-crane | ```
SELECT now() > (today() - 1)
┌─greater(now(), minus(today(), 1))─┐
│ 1 │
└───────────────────────────────────┘
SELECT now() < (today() + 1)
┌─less(now(), plus(today(), 1))─┐
│ 1 │
└───────────────────────────────┘
```
probably index relat... | 2019-08-02T11:04:18 | 2019-08-02T11:04:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,696,335 | 6,242 | nikitamikhaylov | How I think the problem is that aggregate functions don't support constant columns. So in this function https://github.com/yandex/ClickHouse/blob/7015e5ca7df1647a6189a0da4f33bb9f8edc99fa/dbms/src/Functions/array/arrayReduce.cpp#L116 we try to materialize them if they were constant. But they are not constant, because co... | 2019-08-02T13:15:22 | 2019-08-02T13:15:22 | {} | MEMBER |
ClickHouse | ClickHouse | 517,708,514 | 6,289 | alexey-milovidov | If you want to run clickhouse-server as root, you can chown data directory to root:
`chown -R root:root /var/lib/clickhouse`
When you will need to get back normal, you can chown back:
`chown -R clickhouse:clickhouse /var/lib/clickhouse`
| 2019-08-02T13:51:28 | 2019-08-02T13:51:28 | {
"+1": 2,
"-1": 1
} | MEMBER |
ClickHouse | ClickHouse | 517,029,232 | 5,815 | alexey-milovidov | ```
$ docker run -it --rm centos:latest /bin/bash
[root@1b145edc4bb5 /]# yum install yum-utils
Loaded plugins: fastestmirror, ovl
Determining fastest mirrors
* base: mirror.sale-dedic.com
* extras: mirror.sale-dedic.com
* updates: centos-mirror.rbc.ru
base ... | 2019-07-31T21:31:14 | 2019-07-31T21:31:14 | {} | MEMBER |
ClickHouse | ClickHouse | 517,051,451 | 6,205 | alexey-milovidov | Something is wrong with your own test. | 2019-07-31T22:51:11 | 2019-07-31T22:51:11 | {} | MEMBER |
ClickHouse | ClickHouse | 517,118,672 | 6,255 | alexey-milovidov | `00704_drop_truncate_memory_table` - flappy test (fixed in master). | 2019-08-01T04:54:07 | 2019-08-01T04:54:07 | {} | MEMBER |
ClickHouse | ClickHouse | 517,235,760 | 6,191 | CurtizJ | 00974_query_profiler - flappy test failed by timeout. | 2019-08-01T11:01:36 | 2019-08-01T11:01:36 | {} | MEMBER |
ClickHouse | ClickHouse | 517,483,300 | 6,241 | den-crane | Why did you think that memory used on search?
Your query has a JOIN. To execute JOIN CH puts the right table into memory into hash table.
How many rows returns:
```
select count()
from ObserverNodeOccurrence_NodeElements as occ_ne
```
Did you try
```
SELECT COUNT(*)
FROM ObserverNodeOccurrence
where _... | 2019-08-01T22:40:57 | 2019-08-01T22:45:55 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,709,256 | 6,289 | alexey-milovidov | We have this check to avoid situation when the user temporarily runs clickhouse-server as root and it will pollute data directory with some root-owned files, that normally runned clickhouse-server can't manipulate. | 2019-08-02T13:53:29 | 2019-08-02T13:53:50 | {} | MEMBER |
ClickHouse | ClickHouse | 517,714,281 | 6,241 | Guilherme-Routar | Alright, thanks for the input. | 2019-08-02T14:07:33 | 2019-08-02T14:07:33 | {} | NONE |
ClickHouse | ClickHouse | 517,251,669 | 5,042 | MahmoudGoda0 | Hi @alexey-milovidov ,
I tried it with our data, our issues was for 10 queries, 8 of them enhanced but 2 of them become slower than the old version (stable), Can you please help me ?
One query :
select TOP 1000 ext_rec_num,xdr_id,xdr_grp,xdr_type,xdr_subtype,xdr_direction,xdr_location, start_time,stop_time, transact... | 2019-08-01T11:51:09 | 2019-08-01T11:51:09 | {} | NONE |
ClickHouse | ClickHouse | 517,502,073 | 5,601 | bksunday | +1 for Avro support. This schema validated format is super useful on large scale projects involving many separate teams that may not even need to speak to each other to work with the data.
With Kafka and the AvroRegistry it ensures data is of the expected structure and type, and the full doc (schema) on that structu... | 2019-08-02T00:20:09 | 2019-08-02T00:20:09 | {
"+1": 19
} | NONE |
ClickHouse | ClickHouse | 517,635,176 | 6,275 | TH-HA | @den-crane , thanks for the parameter , i've rerun the test with the parameter --max_insert_block_size=100000 add to the clickhouse-client. It's properly taken into account ( i can see the correct value in query_log but it does not change anything - same memory usage.
Regards
| 2019-08-02T09:38:23 | 2019-08-02T09:38:23 | {} | NONE |
ClickHouse | ClickHouse | 517,778,739 | 6,252 | reflection | Actually, reopening for quick discussion: any way to return a better error message about inodes instead of generic "no space left on device?" | 2019-08-02T17:13:33 | 2019-08-02T17:13:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,083,631 | 5,984 | alexey-milovidov | It was implemented in proposed form: https://github.com/yandex/ClickHouse/issues/5984#issuecomment-513411725
> The following one is my favorite.
TABLE name(types..) will be optional
That's more beautiful but adds special syntax. We still have the possibility to implement it :) | 2019-08-01T01:38:16 | 2019-08-01T01:38:16 | {
"hooray": 4
} | MEMBER |
ClickHouse | ClickHouse | 517,140,588 | 2,833 | shivacr | @kshvakov Even i am facing the same issue.
We are running vFlow in one docker which has clickhouse consumer to populate the DB after reading from Kafka.
Clickhouse-server is also running in docker but in another host machine.
After some idle time we are seeing this issue.
Logs:
{"log":"2019/08/01 06:22:17 co... | 2019-08-01T06:33:38 | 2019-08-01T06:33:38 | {} | NONE |
ClickHouse | ClickHouse | 517,145,680 | 5,042 | MahmoudGoda0 | @alexey-milovidov Super clear!
Many thanks, Really appreciated. | 2019-08-01T06:52:00 | 2019-08-01T06:52:00 | {} | NONE |
ClickHouse | ClickHouse | 517,194,706 | 5,317 | alex-zaitsev | So the proposal is:
1) Decrease DBMS_CONNECTION_POOL_WITH_FAILOVER_DEFAULT_DECREASE_ERROR_PERIOD to 60 seconds.
2) Add DBMS_CONNECTION_POOL_WITH_FAILOVER_MAX_ERROR_COUNT = 1000
3) Add 'replica_error_decrease_period' and 'replica_error_max_count' to make those settings configurable.
4) Add current error count to sy... | 2019-08-01T08:58:24 | 2019-08-01T08:58:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,351,184 | 6,241 | Guilherme-Routar | Thank you very much for the input. I updated to v19.11.4 but the error pops up anyway (I did recreate all tables); is there any setting I should change to enable adaptive index granularity?
```
ClickHouse client version 19.11.4.24 (official build).
Connecting to localhost:9000 as user default.
Connected to ClickH... | 2019-08-01T16:02:16 | 2019-08-01T16:04:38 | {} | NONE |
ClickHouse | ClickHouse | 517,480,018 | 6,238 | den-crane | use_uncompressed_cache
https://clickhouse.yandex/docs/en/operations/settings/settings/#setting-use_uncompressed_cache
it is disabled by default | 2019-08-01T22:25:55 | 2019-08-01T22:25:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,574,634 | 6,261 | BayoNet | @alexey-milovidov
Review this pull, please. I have removed information about .config format of the configuration files. If it's incorrect I'll restore it. | 2019-08-02T06:40:21 | 2019-08-02T06:40:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,653,483 | 6,249 | 4ertus2 | "Timed out while waiting for mutation to execute!" in Functional stateless tests (debug) looks like flap | 2019-08-02T10:37:31 | 2019-08-02T10:37:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 517,865,327 | 6,279 | alexey-milovidov | https://github.com/yandex/ClickHouse/issues/6304 | 2019-08-02T22:48:39 | 2019-08-02T22:48:39 | {} | MEMBER |
ClickHouse | ClickHouse | 521,384,875 | 6,446 | alexey-milovidov | Something is wrong:
```
milovidov@milovidov-Pro-P30:~/work/ClickHouse/dbms/tests/queries/0_stateless$ for i in {1..100}; do clickhouse-client --query "SELECT nullIf(1, NULL)"; done
\N
1
1
1
1
1
\N
1
1
1
1
1
\N
1
1
1
1
1
\N
1
1
1
1
1
\N
1
1
1
1
1
\N
1
1
1
1
1
\N
1
1
1
1
1
\... | 2019-08-14T19:27:58 | 2019-08-14T19:27:58 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.