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 | 812,281,177 | 18,813 | tkionshao | I found the server it crashing using LXC, so I think #20608 #21594 are relevant with it.
I tried to use docker contain where Clickhouse is running inside on LXC server. The time of segmentation fault happening is getting lower, but if too many process units are working such as inserting data, dropping data and heavy... | 2021-04-02T02:24:47 | 2021-04-02T02:27:43 | {} | NONE |
ClickHouse | ClickHouse | 812,300,536 | 9,604 | alexey-milovidov | Finalize each block. | 2021-04-02T03:43:12 | 2021-04-02T03:43:12 | {} | MEMBER |
ClickHouse | ClickHouse | 607,388,776 | 7,467 | Tasselmi | No, thanks
On 04/02/2020 00:54, Ivan Blinkov wrote:
@Tasselmi, do you have any further questions?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or unsubscribe. | 2020-04-01T17:33:36 | 2021-04-02T04:10:33 | {} | NONE |
ClickHouse | ClickHouse | 811,982,235 | 22,431 | mdianjun | > Can't you just do the following:
>
> ```
> INSERT INTO demo (....,dt) SELECT *, toDate('20200310') as dt ....
> ```
No, because I want to use the insert statement with clickhouse-jdbc, and read data from a file stream, like this:
```
conn.createStatement
.write()
.sql("insert into dem... | 2021-04-01T15:22:00 | 2021-04-01T15:22:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,327,679 | 14,298 | filimonov | https://presentations.clickhouse.tech/cpp_russia_2020/index.html | 2021-04-02T05:30:21 | 2021-04-02T05:30:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 811,954,457 | 22,442 | vzakaznikov | Same issue with `expr FOLLOWING`. | 2021-04-01T14:39:36 | 2021-04-01T14:39:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,286,478 | 22,480 | alexey-milovidov | > Testflows check — failed: 0, passed: 1, other: 0 | 2021-04-02T02:44:27 | 2021-04-02T02:44:27 | {} | MEMBER |
ClickHouse | ClickHouse | 812,320,373 | 22,471 | qoega | It useful to add secret to a vault from SQL. | 2021-04-02T05:07:12 | 2021-04-02T05:07:12 | {} | MEMBER |
ClickHouse | ClickHouse | 812,331,545 | 21,694 | filimonov | BTW most tfaffic is created by few hundreds of different useragents, lru cache on hashed user agent string allows not to apply any regexp most of the time. | 2021-04-02T05:42:13 | 2021-04-02T05:42:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,331,650 | 9,604 | alexey-milovidov | You just specify -State explicitly if needed. | 2021-04-02T05:42:35 | 2021-04-02T05:42:35 | {} | MEMBER |
ClickHouse | ClickHouse | 812,350,935 | 22,483 | vsop-479 | My Clickhouse version is 20.6.
I will test it on a new version. | 2021-04-02T06:37:01 | 2021-04-02T06:37:01 | {} | NONE |
ClickHouse | ClickHouse | 812,356,842 | 13,957 | filimonov | JSONAsString + parsing of the json in materialized view using JSON function is an option. | 2021-04-02T06:53:05 | 2021-04-02T06:53:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 811,973,018 | 22,444 | den-crane | Not related to LZ4 and JDBC compression.
> .data(createHadoopInputStream(args), ClickHouseFormat.ORC)
> .dataCompression(ClickHouseCompression.gzip)
This means that your data stream with ORC is compressed by gzip.
If your stream is not compressed by gzip but ORC has internal compression then you shoul... | 2021-04-01T15:08:04 | 2021-04-01T15:08:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,065,305 | 22,423 | Enmk | I like the idea of deriving encryption\decryption keys from the server key, however, I doubt that it is safe to reuse the same key for the encryption\decrption functions.
IMO this piece of code would be a great starting point for one of the secret backends proposed in #22471, namely a `server_key_backend` or any oth... | 2021-04-01T17:41:11 | 2021-04-01T17:41:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,146,707 | 21,693 | filimonov | I mean that we still have unfixed perf regressions in 21.2 (same exists in 21.3) | 2021-04-01T20:14:54 | 2021-04-01T20:14:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,196,965 | 22,478 | akuzm | From `CD 9075-2:201?(E) 7.14 <window clause>`:
```
13) If WDEF's <window frame clause> specifies <window frame preceding> or <window frame following>,
then let UVS be the <unsigned value specification> simply contained in the <window frame preceding>
or <window frame following>.
Case:
a)
If RANGE is specified, t... | 2021-04-01T21:58:10 | 2021-04-01T21:58:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,332,226 | 21,694 | alexey-milovidov | Yes, this technique was implemented in our internal library (the number of unique user-agents was about several millions). | 2021-04-02T05:44:23 | 2021-04-02T05:44:23 | {} | MEMBER |
ClickHouse | ClickHouse | 812,334,500 | 22,460 | filimonov | Please do
```
grep '####' -C50 /var/log/clickhouse-server/clickhouse-server.log
``` | 2021-04-02T05:51:29 | 2021-04-02T05:51:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 810,758,567 | 22,398 | alexey-milovidov | Brown university: https://github.com/crottyan/mgbench
Abandoned, but still quite fresh. | 2021-03-31T04:50:17 | 2021-04-01T14:30:26 | {} | MEMBER |
ClickHouse | ClickHouse | 811,957,406 | 22,398 | alexey-milovidov | PostgreSQL vs ClickHouse via clickhouse_fdw in PostgreSQL:
https://www.percona.com/blog/2019/05/09/improving-olap-workload-performance-for-postgresql-with-clickhouse-database/ | 2021-04-01T14:44:07 | 2021-04-01T14:44:07 | {} | MEMBER |
ClickHouse | ClickHouse | 811,993,275 | 22,398 | alex-zaitsev | ClickHouse vs Druid vs Rockset:
https://altinity.com/blog/clickhouse-nails-cost-efficiency-challenge-against-druid-rockset | 2021-04-01T15:38:36 | 2021-04-01T15:38:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,175,823 | 22,475 | mastertheknife | \+ I also have a patch to disable FastMemcpy, but it seems this doesn't exist in master anymore.
```
--- contrib/FastMemcpy/CMakeLists.txt.orig 2021-03-15 20:23:07.888493694 -0400
+++ contrib/FastMemcpy/CMakeLists.txt 2021-03-15 20:22:46.919278044 -0400
@@ -1,6 +1,6 @@
option (ENABLE_FASTMEMCPY "Enable FastMemc... | 2021-04-01T21:13:08 | 2021-04-01T21:13:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,198,770 | 22,478 | vzakaznikov | Intuitively one would expect it to work. Not a big deal though. It is more about user experience and having a consistent behavior. We can close the issue if you want.
Here is a similar corner case `RANGE BETWEEN 0 FOLLOWING AND 0 FOLLOWING`
```
user-host :) SELECT number,sum(number) OVER (RANGE BETWEEN 0 FOLLOW... | 2021-04-01T22:03:08 | 2021-04-01T22:03:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,218,794 | 22,475 | alexey-milovidov | Yes, we replaced it with another memcpy, that is only enabled with `glibc_compatibility`. | 2021-04-01T22:52:17 | 2021-04-01T22:52:17 | {} | MEMBER |
ClickHouse | ClickHouse | 812,302,524 | 14,298 | filimonov | > You can find it here
I see only readme there. Can you commit spec file somewhere? It's quite silly to unpack srpm for that.
Upd. Find it here https://build.opensuse.org/package/show/home:mastertheknife:clickhouse-stable/ClickHouse | 2021-04-02T03:52:51 | 2021-04-02T03:57:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 811,958,322 | 22,398 | alexey-milovidov | ClickHouse vs. clickhouse_fdw in PostgreSQL:
https://www.percona.com/blog/2019/05/01/benchmark-clickhouse-database-and-clickhousedb_fdw/ | 2021-04-01T14:45:34 | 2021-04-01T14:45:34 | {} | MEMBER |
ClickHouse | ClickHouse | 812,278,641 | 22,476 | mastertheknife | Interesting. I checked on the POWER9 system, looks like it does have such header, but it probably needs to be included differently, e.g. with "ppc_wrappers/xmmintrin.h" instead of "xmmintrin.h", because the first one is really SSE2 and has no includes to the ppc_wrappers one.
```
[root@removed /]# find /usr -name '*m... | 2021-04-02T02:15:14 | 2021-04-02T02:15:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,291,201 | 19,766 | ghost | @vdimir
I fixed some codes.
- Remove `AggregateFunctionNullVariadic::insertResultInto`
- Prevent `forward` with `tail` and `backward` with `head`
- Minor fix for `exception messages`
Thank you. | 2021-04-02T03:03:16 | 2021-04-02T03:04:23 | {
"eyes": 1
} | NONE |
ClickHouse | ClickHouse | 812,339,915 | 22,460 | likang-github | 
| 2021-04-02T06:07:12 | 2021-04-02T06:07:12 | {} | NONE |
ClickHouse | ClickHouse | 812,366,813 | 22,485 | Slach | oops | 2021-04-02T07:10:28 | 2021-04-02T07:10:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 811,950,126 | 22,398 | alexey-milovidov | TSBS benchmark: https://altinity.com/blog/clickhouse-for-time-series
(2018 version, does not have read in order optimization)
ClickHouse vs Kudu on TSBS (from Cloudera, in favor of Kudu):
https://blog.cloudera.com/benchmarking-time-series-workloads-on-apache-kudu-using-tsbs/ | 2021-04-01T14:33:16 | 2021-04-01T14:33:16 | {} | MEMBER |
ClickHouse | ClickHouse | 811,961,384 | 22,398 | alexey-milovidov | SparkSql, Presto, Impala, HAWQ, ClickHouse, GreenPlum on TPC-DS inspired benchmark:
https://programmersought.com/article/7939525005/ | 2021-04-01T14:50:17 | 2021-04-01T14:50:17 | {} | MEMBER |
ClickHouse | ClickHouse | 812,132,177 | 21,101 | kitaisreal | @keerthana-kannan-ch
If you create dictionary with query
```
CREATE DICTIONARY ip_prefix_dict(
prefix String DEFAULT '',
as_number UInt32 DEFAULT 0,
as_name String DEFAULT ''
)
PRIMARY KEY prefix
SOURCE(CLICKHOUSE(HOST 'localhost' PORT 9000 USER 'default' TABLE 'ip_prefix' PASSWORD '' DB 'tests'))
LIFET... | 2021-04-01T19:46:50 | 2021-04-01T19:46:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,193,480 | 22,460 | filimonov | Normally that should not happen. Core file is created when the process segfault.
It is alway a sign of some bug.
To invistigate it we need clickhouse version and fragment of clickhouse logs (see /var/log/clickhouse-server/clickhouse-server.log) - lines around '####'.
Optionally the coredump files can also be... | 2021-04-01T21:55:07 | 2021-04-01T21:55:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,264,753 | 22,482 | alexey-milovidov | Explore the contents with:
`zstd -d < RC_2019-12.zst | head -n10 | jq .`
`zstd -d < RC_2019-12.zst | jq '.all_awardings' | uniq | less`
`zstd -d < RC_2019-12.zst | jq '.subreddit_type' | sort | uniq -c`
The format is mostly regular. New fields did appear through the time.
There is some redundancy...
This feat... | 2021-04-02T01:18:46 | 2021-04-02T01:25:05 | {} | MEMBER |
ClickHouse | ClickHouse | 812,346,136 | 22,460 | likang-github | 
| 2021-04-02T06:24:29 | 2021-04-02T06:24:29 | {} | NONE |
ClickHouse | ClickHouse | 812,359,691 | 14,896 | compasses | @UnamedRus Hi, Do you know which MR fix that? | 2021-04-02T06:59:58 | 2021-04-02T06:59:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,360,571 | 19,873 | filimonov | > you guys use parquetreader --> json
Sure we don't. We use https://github.com/apache/arrow/blob/master/cpp/src/parquet/arrow/reader.h | 2021-04-02T07:02:08 | 2021-04-02T07:02:08 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,367,987 | 22,472 | alexey-milovidov | @KochetovNicolai Unhandled exception in destructor in PipelineExecutor:
```
2021.04.02 04:39:42.752226 [ 308 ] {} <Fatal> BaseDaemon: (version 21.4.1.6421, build id: 4B4A05810C453280D81F1B1B5A6E37FA46BFEE56) (from thread 128749) Terminate called for uncaught exception:
2021.04.02 04:40:18.651512 [ 131993 ] {} <Fatal... | 2021-04-02T07:11:37 | 2021-04-02T07:11:37 | {} | MEMBER |
ClickHouse | ClickHouse | 812,384,241 | 22,460 | likang-github | change table engine？ use MergeTree？ | 2021-04-02T07:27:18 | 2021-04-02T07:27:18 | {} | NONE |
ClickHouse | ClickHouse | 812,143,957 | 22,375 | paeolo | @kitaisreal Woaw that was fast! Thank you for the great work;) | 2021-04-01T20:09:57 | 2021-04-01T20:10:41 | {} | NONE |
ClickHouse | ClickHouse | 812,173,019 | 22,402 | gyuton | Internal documentation ticket: DOCSUP-8181. | 2021-04-01T21:07:18 | 2021-04-01T21:07:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,218,011 | 22,476 | alexey-milovidov | I can remember why only gcc was enabled.
It because gcc has special support to compile SSE intrinsics (x86) for ppc64.
(most likely with special version of `*intrin.h`)
So, you can compile for Power while enabling `__SSE2__`, etc.
It was since gcc-8. Maybe clang has also kept up?
Nevertheless, the build withou... | 2021-04-01T22:50:00 | 2021-04-01T22:51:01 | {} | MEMBER |
ClickHouse | ClickHouse | 812,282,948 | 22,408 | alexey-milovidov | @filimonov I checked it:
> Instead of fancy sorting by weight, I changed the sort to be stable.
We are already using stable sort (Radix LSD), so the fix is irrelevant for us. | 2021-04-02T02:31:34 | 2021-04-02T02:31:34 | {} | MEMBER |
ClickHouse | ClickHouse | 812,313,205 | 22,460 | likang-github | yes version 20.12.3.3 | 2021-04-02T04:39:51 | 2021-04-02T04:39:51 | {} | NONE |
ClickHouse | ClickHouse | 811,951,404 | 22,398 | alexey-milovidov | ClickHouse vs Spark on Wikipedia data:
https://www.percona.com/blog/2017/02/13/clickhouse-new-opensource-columnar-database/ | 2021-04-01T14:35:09 | 2021-04-01T14:35:09 | {} | MEMBER |
ClickHouse | ClickHouse | 811,957,927 | 22,398 | alexey-milovidov | ClickHouse vs. MariaDB Column Store vs. Spark vs. MySQL on Wikipedia data:
https://www.percona.com/blog/2017/03/17/column-store-database-benchmarks-mariadb-columnstore-vs-clickhouse-vs-apache-spark/ | 2021-04-01T14:44:58 | 2021-04-01T14:44:58 | {} | MEMBER |
ClickHouse | ClickHouse | 811,994,500 | 22,398 | alex-zaitsev | ClickHouse vs ScyllaDB on 500B rows dataset:
https://altinity.com/blog/2020/1/1/clickhouse-cost-efficiency-in-action-analyzing-500-billion-rows-on-an-intel-nuc | 2021-04-01T15:40:27 | 2021-04-01T15:40:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,026,908 | 20,380 | akuzm | Another one https://clickhouse-test-reports.s3.yandex.net/22300/c7bf01fa33fa1f8043f97b40fe55d5acb641ca14/functional_stateless_tests_(thread).html#fail1 | 2021-04-01T16:33:08 | 2021-04-01T16:33:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,107,864 | 22,379 | alvarotuso | @filimonov that seems to fix it:
```
SELECT *
FROM test.test_collation
ORDER BY v ASC COLLATE 'en'
SETTINGS optimize_read_in_order = 0
Query id: 7291aa7d-9365-4069-b847-b8fe9223f501
┌─v─┬─v2─┐
│ a │ a │
│ A │ A │
│ b │ b │
│ B │ B │
│ c │ c │
│ C │ C │
└───┴────┘
6 rows in set. Elapsed: 0.003... | 2021-04-01T18:59:15 | 2021-04-01T18:59:15 | {} | NONE |
ClickHouse | ClickHouse | 812,139,057 | 22,325 | filimonov | I think it should be backported to 21.3. So maybe mark it a bugfix? | 2021-04-01T20:00:52 | 2021-04-01T20:00:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,300,783 | 22,483 | den-crane | What Clickhouse version do you use?
Modern Clickhouse recalculate TTL when alter table modify ttl applied.
Also use `ALTER TABLE ... MATERIALIZE TTL` to materialize ttl. | 2021-04-02T03:44:29 | 2021-04-02T03:44:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 811,993,060 | 22,398 | alex-zaitsev | ClickHouse vs RedShift on fintech use case:
https://altinity.com/blog/clickhouse-vs-redshift-performance-for-fintech-risk-management | 2021-04-01T15:38:15 | 2021-04-01T15:38:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,096,947 | 22,221 | traceon | @Blackmorse what does your role mapping config look like? | 2021-04-01T18:38:57 | 2021-04-01T18:38:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,128,621 | 21,412 | rhuddleston | previous_value would be very helpful | 2021-04-01T19:39:42 | 2021-04-01T19:39:42 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 812,161,899 | 21,101 | kitaisreal | I checked issue is reproduced on current master.
Currently during database load it first load dictionaries, then load tables, that way if table column depends on dictGet function call it will be loaded successfully because dictionary will exists in database, in your case your table ip_test is in `default` database, an... | 2021-04-01T20:44:49 | 2021-04-01T20:45:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,172,848 | 21,329 | gyuton | Internal documentation ticket: DOCSUP-8181. | 2021-04-01T21:06:57 | 2021-04-01T21:06:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,298,641 | 22,404 | lxswbr | @KochetovNicolai
More than 600 lines are all the same information | 2021-04-02T03:33:52 | 2021-04-02T03:33:52 | {} | NONE |
ClickHouse | ClickHouse | 812,326,066 | 14,298 | filimonov | > Thank you! If possible please continue this effort, modify the spec file accordingly and implement official builds (as you started)
Actually we want to have only single option for official build for all distributions (it depends only on old version of glibc). Packaging can be improved, but binary should be the sam... | 2021-04-02T05:25:19 | 2021-04-02T05:25:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,356,639 | 22,460 | alexey-milovidov | It's related to HDFS, the issue is duplicate of:
#22115
#13406 | 2021-04-02T06:52:31 | 2021-04-02T06:52:31 | {} | MEMBER |
ClickHouse | ClickHouse | 812,362,535 | 19,873 | alanpaulkwan | Don't know what the problem is then! | 2021-04-02T07:06:13 | 2021-04-02T07:06:13 | {} | NONE |
ClickHouse | ClickHouse | 812,363,450 | 22,487 | alexey-milovidov | > Stress test (memory) — Fatal message in clickhouse-server.log
OOM. | 2021-04-02T07:07:09 | 2021-04-02T07:07:09 | {} | MEMBER |
ClickHouse | ClickHouse | 811,952,878 | 22,398 | alexey-milovidov | ClickHouse vs. Redshift on NYC taxi dataset:
https://altinitydb.medium.com/clickhouse-vs-amazon-redshift-benchmark-e223429f4f95 | 2021-04-01T14:37:12 | 2021-04-01T14:37:12 | {} | MEMBER |
ClickHouse | ClickHouse | 812,027,161 | 22,300 | akuzm | Functional stateless tests (thread) https://github.com/ClickHouse/ClickHouse/issues/20380 | 2021-04-01T16:33:32 | 2021-04-01T16:33:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,086,414 | 22,472 | akuzm | Ohhh I forgot it's not so easy, and the container for fast test is not rebuilt in CI. At least I see that the LLVM_VERSION from the container is picked up properly. | 2021-04-01T18:19:19 | 2021-04-01T18:19:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,232,314 | 20,499 | alexey-milovidov | > I've managed to build server locally, but it has problems with PHDR cache code (debug+shared+ASAN build):
`getHashOfLoadedBinary` can be simply disabled for shared build. | 2021-04-01T23:34:06 | 2021-04-01T23:34:06 | {} | MEMBER |
ClickHouse | ClickHouse | 812,232,554 | 20,499 | alexey-milovidov | > What problem was inside this log?
Something with LTO (only works with the same versions of clang/lld). | 2021-04-01T23:34:59 | 2021-04-01T23:35:31 | {} | MEMBER |
ClickHouse | ClickHouse | 812,345,622 | 22,460 | likang-github |
[users.txt](https://github.com/ClickHouse/ClickHouse/files/6248005/users.txt)
[config.txt](https://github.com/ClickHouse/ClickHouse/files/6248006/config.txt)
| 2021-04-02T06:22:59 | 2021-04-02T06:22:59 | {} | NONE |
ClickHouse | ClickHouse | 812,361,284 | 22,484 | alexey-milovidov | Ok, it's cross-compiled successfully for AArch64. | 2021-04-02T07:03:52 | 2021-04-02T07:03:52 | {} | MEMBER |
ClickHouse | ClickHouse | 812,262,840 | 1,158 | alexey-milovidov | #22428 | 2021-04-02T01:11:33 | 2021-04-02T01:11:33 | {} | MEMBER |
ClickHouse | ClickHouse | 812,264,249 | 22,482 | alexey-milovidov | The source is http://files.pushshift.io/reddit/comments/
It is available from Dec 2005 thru Dec 2019 (almost 15 years)
and represented by one file per month, 169 files in total.
Every file is in JSONEachRow format with `bz2`, `xz` or `zstd` compression.
The size of downloaded compressed data is 687 GB.
Downl... | 2021-04-02T01:17:01 | 2021-04-02T02:00:31 | {} | MEMBER |
ClickHouse | ClickHouse | 812,300,625 | 22,422 | marising | It's seems that the 3 ZooKeeper processes are created by the runner. But I don’t understand why clickhouse can’t connect to zookeeper. | 2021-04-02T03:43:41 | 2021-04-02T03:43:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 811,960,311 | 22,398 | alexey-milovidov | Mark Litwintschik's benchmarks: different systems on different hardware at different time:
https://tech.marksblogg.com/benchmarks.html
| 2021-04-01T14:48:44 | 2021-04-01T14:48:44 | {} | MEMBER |
ClickHouse | ClickHouse | 812,280,076 | 22,482 | alexey-milovidov | ```
clickhouse-client --query "SELECT count() FROM reddit"; sleep 100; clickhouse-client --query "SELECT count() FROM reddit"
```
The loading speed on old 16-core server (E5-2560v2) is 358 000 rows/sec and the bottleneck is bz2 decompression. | 2021-04-02T02:20:27 | 2021-04-02T02:21:02 | {} | MEMBER |
ClickHouse | ClickHouse | 812,287,690 | 22,476 | alexey-milovidov | > it probably needs to be included differently, e.g. with "ppc_wrappers/xmmintrin.h" instead of "xmmintrin.h"
If you add `ppc_wrappers` directory to include paths before other includes, it should work as a replacement.
Also maybe there is some mechanism that enables it automatically.
Adding ifdefs in every sourc... | 2021-04-02T02:49:14 | 2021-04-02T02:50:52 | {} | MEMBER |
ClickHouse | ClickHouse | 597,812,232 | 9,604 | alexey-milovidov | @dimarub2000 will take this task.
Update: not happened. | 2020-03-11T19:01:31 | 2021-04-02T03:44:34 | {} | MEMBER |
ClickHouse | ClickHouse | 812,314,444 | 14,298 | mastertheknife | > Great effort! I think that it worth comminting (at least some, or maybe most of) cmake patches to clickhouse main repo.
Thank you!
I have created pull requests for all ppc64le patches except two, which i am not sure how they should be handled.
@alexey-milovidov also for you. These i'm not sure how to handle, bec... | 2021-04-02T04:45:19 | 2021-04-02T04:45:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,346,536 | 22,485 | filimonov | Duplicate of #7659
Also #8158 | 2021-04-02T06:25:35 | 2021-04-02T06:26:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,369,859 | 22,472 | alexey-milovidov | Looks like we are waiting for ThreadPool in destructor and exception was thrown in one of threads. | 2021-04-02T07:13:20 | 2021-04-02T07:13:20 | {} | MEMBER |
ClickHouse | ClickHouse | 811,984,217 | 22,431 | den-crane | Actually you can: you can use input function. | 2021-04-01T15:25:09 | 2021-04-01T15:25:09 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,020,297 | 22,221 | Blackmorse | @traceon
> do you disconnect the first session, when you try to authenticate the second time (with reassigned roles at LDAP side)?
I do the both - result is the same
> does role2 exist in ClickHouse?
Yes, it does. After rebooting Clickhouse user can log in and correctly gain role_2
> is it reproduced ... | 2021-04-01T16:22:07 | 2021-04-01T16:22:34 | {} | NONE |
ClickHouse | ClickHouse | 812,055,334 | 22,405 | Jirennor | Thanks for the replies.
The suggestion from @den-crane seems to work and I will do some further testing with that but I do have some follow up questions.
- Is there a reason why this needs to be at the end of the index?
- Why do I need to order it twice on the `Timeslot` field?
To elaborate on the last question... | 2021-04-01T17:22:41 | 2021-04-01T17:22:41 | {} | NONE |
ClickHouse | ClickHouse | 812,187,474 | 22,477 | vzakaznikov | The same issue with `RANGE 0 FOLLOWING AND 0 PRECEDING`
```
user-host :) SELECT number,sum(number) OVER (RANGE BETWEEN 0 FOLLOWING AND 0 PRECEDING) FROM values('number Int8', (1),(1),(2),(3))
SELECT
number,
sum(number) OVER (RANGE BETWEEN 0 FOLLOWING AND 0 PRECEDING)
FROM values('number Int8', 1, 1, 2... | 2021-04-01T21:41:12 | 2021-04-01T21:41:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 811,594,063 | 14,893 | dengc367 | > @dengc367 No, it's in early stage of development.
> BTW, what sketches do you need the most?
>
> ClickHouse already have multiple sketches for COUNT DISTINCT, quantiles and topK.
Thanks for your reply, @alexey-milovidov. I need the sketches like the `Intersect/Except` functionality in Presto 。Ref: https://gith... | 2021-04-01T02:35:25 | 2021-04-01T22:56:04 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 812,283,862 | 22,482 | alexey-milovidov | It's expected that the dataset has approximately 7 bn records and the data load will take 6 hours. | 2021-04-02T02:35:00 | 2021-04-02T02:36:09 | {} | MEMBER |
ClickHouse | ClickHouse | 812,389,467 | 22,465 | alesapin | Flaky test | 2021-04-02T07:33:18 | 2021-04-02T07:33:18 | {} | MEMBER |
ClickHouse | ClickHouse | 811,955,692 | 22,398 | alexey-milovidov | ClickHouse vs. Redshift (price-performance), no many details:
https://altinity.com/blog/2017/7/3/clickhouse-vs-redshift-2 | 2021-04-01T14:41:28 | 2021-04-01T14:41:38 | {} | MEMBER |
ClickHouse | ClickHouse | 812,272,466 | 22,460 | likang-github | how can i specify the core.* file generation path through the config.xml | 2021-04-02T01:50:55 | 2021-04-02T01:50:55 | {} | NONE |
ClickHouse | ClickHouse | 812,334,981 | 22,221 | Blackmorse | @traceon
```
<role_mapping>
<base_dn>cn=groups,cn=accounts,dc=company,dc=com</base_dn>
<scope>subtree</scope>
<search_filter>(&(member=uid={user_name},cn=users,cn=accounts,dc=company,dc=com))</search_filter>
<attribute>cn</attribute>
</role_mapping>
``` | 2021-04-02T05:52:56 | 2021-04-02T05:52:56 | {} | NONE |
ClickHouse | ClickHouse | 811,949,338 | 21,018 | UnamedRus | Probably related https://github.com/ClickHouse/ClickHouse/issues/22438 | 2021-04-01T14:32:07 | 2021-04-01T14:32:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 811,951,838 | 22,398 | alexey-milovidov | ClickHouse vs. MariaDB ColumnStore on Star Schema Benchmark (TPC-H derivative):
https://www.percona.com/blog/2020/07/27/clickhouse-and-columnstore-in-the-star-schema-benchmark/ | 2021-04-01T14:35:47 | 2021-04-01T14:35:47 | {} | MEMBER |
ClickHouse | ClickHouse | 812,172,595 | 21,901 | gyuton | Internal documentation ticket: DOCSUP-8181. | 2021-04-01T21:06:27 | 2021-04-01T21:06:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 812,298,458 | 9,604 | amosbird | > Results will be incorrect (not fully merged)
What's the definition of "not fully merged" here? Do we merge to the `-State` or finalize each block? | 2021-04-02T03:33:05 | 2021-04-02T03:33:05 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 812,301,124 | 22,460 | alexey-milovidov | Do you use the official ClickHouse release binaries? | 2021-04-02T03:46:09 | 2021-04-02T03:46:09 | {} | MEMBER |
ClickHouse | ClickHouse | 812,308,794 | 9,604 | amosbird | > Finalize each block.
Hmm, how does it work with "Materialized view will merge data in background." and non-SimpleAggregateFunction. | 2021-04-02T04:21:50 | 2021-04-02T04:22:20 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 812,352,481 | 22,460 | likang-github |
[clickhouse-server.zip](https://github.com/ClickHouse/ClickHouse/files/6248084/clickhouse-server.zip)
| 2021-04-02T06:41:18 | 2021-04-02T06:41:18 | {} | NONE |
ClickHouse | ClickHouse | 829,216,685 | 23,296 | tavplubix | Could you please share output of the following query?
```
WITH
'd23fba2922c1a8f3adafceb40a86c34c_45540_46193_7_46162' AS p1,
'd23fba2922c1a8f3adafceb40a86c34c_46190_46194_1' AS p2
SELECT
materialize(hostName()) AS host,
*
FROM remote('replica1,replica2,replica3', 'system.part_log')
WHERE (datab... | 2021-04-29T12:59:20 | 2021-04-29T13:06:53 | {} | MEMBER |
ClickHouse | ClickHouse | 829,296,213 | 23,537 | filimonov | Actually the task is very easy, but it's a bit questionable - it will pollute logs with more data, but the extra value of that is not 100% clear - normally certain types of messages can happen in certain types of threads anyway.
Also - did you try system.text_log? | 2021-04-29T14:40:56 | 2021-04-29T14:40:56 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 829,443,564 | 23,771 | alexey-milovidov | PVS: false positive. | 2021-04-29T17:16:27 | 2021-04-29T17:16:27 | {} | MEMBER |
ClickHouse | ClickHouse | 829,476,149 | 23,746 | bnaecker | Good catch, thank you. | 2021-04-29T18:06:19 | 2021-04-29T18:06:19 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.