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 | 588,196,134 | 9,200 | mvcalder-xbk | Thanks @den-crane,
I guess I misunderstood how the partitioning mechanism worked. In our actual database (the example was simplified for the question) we were computing the date from the timestamp (both are strings here) via `substr(timestamp,1,10)`. I thought CH would "figure out" the partition, in hindsight that ... | 2020-02-19T12:19:06 | 2020-02-19T12:19:06 | {} | NONE |
ClickHouse | ClickHouse | 588,451,317 | 9,227 | alesapin | Why tests for alters don't have word `alter` in their names? | 2020-02-19T20:47:26 | 2020-02-19T20:47:26 | {} | MEMBER |
ClickHouse | ClickHouse | 587,449,089 | 9,195 | BernieJiangDL | My table engine is ReplicatedMergeTree and the records are inserted into ReplicatedMergeTree table directly rather than distributed table. How does it work without transaction guaranties when those exception happens? Do these 1000 records all fail or some success some fail? | 2020-02-18T13:02:02 | 2020-02-18T13:02:02 | {} | NONE |
ClickHouse | ClickHouse | 587,644,728 | 9,193 | den-crane | There is no example with `where` and no explanation about implicit/explicit string conversion
```
SELECT *
FROM dt
WHERE timestamp = '2019-01-01 03:00:00.000'
Received exception from server (version 20.1.4):
Code: 43. DB::Exception: Received from localhost:9000. DB::Exception: Illegal types of arguments (Date... | 2020-02-18T18:46:58 | 2020-02-18T18:49:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,777,553 | 5,488 | alexey-milovidov | @ygrek It's Ok for farmHash64 function in ClickHouse to be non stable between different machines / releases if it will be clearly stated in docs. | 2020-02-18T20:17:27 | 2020-02-18T20:17:27 | {} | MEMBER |
ClickHouse | ClickHouse | 588,123,401 | 9,195 | BernieJiangDL | > > How does clickhouse deduplicate the same rows without so-called “primary key”? Is there a md5 operation for each row or something like that? Does it happen in merge-tree merging stage?
>
> You do insert: `insert into someREPLICATEDtable values (1),(2),(3);`
> CH calculates a checksum from `someREPLICATEDtable|c... | 2020-02-19T09:33:52 | 2020-02-19T09:33:52 | {} | NONE |
ClickHouse | ClickHouse | 588,290,035 | 7,274 | clemrodriguez | Hi,
Once again (see #5438), I polluted the history logs for this PR due to a rebase : code was conflicting as master had several new commits and refactoring.
As it is pending since a while, would you prefer that I close it and reopen a new one ?
Regards. | 2020-02-19T15:53:35 | 2020-02-19T15:53:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,392,502 | 9,162 | kreuzerkrieg | 74 minutes clean build with PCH
76 without
Do we have any new ideas? | 2020-02-19T19:05:38 | 2020-02-19T19:05:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,397,640 | 9,208 | dependabot-preview[bot] | OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting `@dependabot ignore this major version` or `@dependabot ignore this minor version`.
If you change your mind, just re-... | 2020-02-19T19:16:38 | 2020-02-19T19:16:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,401,219 | 8,915 | alexey-milovidov | @zhang2014 But why it fails in the example from this issue?
Shouldn't HAVING be moved to WHERE if possible? | 2020-02-19T19:23:32 | 2020-02-19T19:23:32 | {} | MEMBER |
ClickHouse | ClickHouse | 588,586,052 | 9,195 | BernieJiangDL | > > Any solutions for this case? Many thanks~
>
> Just do not use spark jdbc to insert into CH. Generate TSV files by Spark and upload them using custom code (&jdbc) which understands CH limits and constraints.
Got it. Decoupling the spark task and clickhouse writer module. Thanks~~~ | 2020-02-20T03:05:02 | 2020-02-20T03:05:02 | {} | NONE |
ClickHouse | ClickHouse | 587,369,343 | 9,159 | nikitamikhaylov | @vpanfilov Try to install package clickhouse-common-static-dbg to get normal stacktrace. | 2020-02-18T09:42:43 | 2020-02-18T09:42:43 | {} | MEMBER |
ClickHouse | ClickHouse | 587,672,237 | 9,198 | nikitamikhaylov | both are appropriate, but stuff is slang and quite rude | 2020-02-18T19:05:43 | 2020-02-18T19:05:43 | {} | MEMBER |
ClickHouse | ClickHouse | 587,685,979 | 8,989 | alexey-milovidov | It's about dangling reference (use after free) of Context in StorageFile:
```
#0 0x1e22a18e in DB::Context::getApplicationType() const /build/obj-x86_64-linux-gnu/../dbms/src/Interpreters/Context.cpp:1969:20
#1 0x1f2a673d in DB::getInputFormatSetting(DB::Settings const&, DB::Context const&) /build/obj-x86_... | 2020-02-18T19:14:52 | 2020-02-18T19:14:52 | {} | MEMBER |
ClickHouse | ClickHouse | 588,009,647 | 9,203 | onine007 | alter table dm_user_behavior_events MODIFY SETTING index_granularity = 8192, enable_mixed_granularity_parts = 1 ; | 2020-02-19T02:56:16 | 2020-02-19T02:56:16 | {} | NONE |
ClickHouse | ClickHouse | 588,077,267 | 8,990 | vigneshrajarr | Can anyone help me here ? | 2020-02-19T07:36:54 | 2020-02-19T07:36:54 | {} | NONE |
ClickHouse | ClickHouse | 587,666,830 | 9,198 | alexey-milovidov | staff vs stuff | 2020-02-18T19:01:55 | 2020-02-18T19:01:55 | {} | MEMBER |
ClickHouse | ClickHouse | 587,776,197 | 5,488 | alexey-milovidov | @JacobHayes Let's do it.
| 2020-02-18T20:16:33 | 2020-02-18T20:16:33 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 588,022,033 | 9,195 | BernieJiangDL | > > My table engine is ReplicatedMergeTree and the records are inserted into ReplicatedMergeTree table directly rather than distributed table. How does it work without transaction guaranties when those exception happens? Do these 1000 records all fail or some success some fail?
>
> https://clickhouse.tech/docs/en/op... | 2020-02-19T03:51:42 | 2020-02-19T03:51:42 | {} | NONE |
ClickHouse | ClickHouse | 588,092,900 | 9,161 | filimonov | @Enmk has set of changes related to DateLUT, i add him to review. Please sync it with him to avoid complicated painful merges | 2020-02-19T08:19:21 | 2020-02-19T08:19:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,214,563 | 9,217 | alesapin | We don't need a test here, because test `01076_parallel_alter_replicated_zookeeper` checks exactly this bug and flaps because the previous fix was incorrect. | 2020-02-19T12:57:28 | 2020-02-19T12:57:28 | {} | MEMBER |
ClickHouse | ClickHouse | 588,423,782 | 9,200 | mvcalder-xbk | Yes, I think we are agreeing, my example was intended to demonstrate our original mistake. | 2020-02-19T20:07:55 | 2020-02-19T20:07:55 | {} | NONE |
ClickHouse | ClickHouse | 588,857,577 | 9,235 | zhang2014 | test fail reason(not related to this change):
```
2020-02-20 07:48:27 Code: 75. DB::Exception: Received from localhost:9000. DB::Exception: Cannot write to file /var/lib/clickhouse/metadata/default/mem1.sql.tmp, errno: 28, strerror: No space left on device.
``` | 2020-02-20T09:41:29 | 2020-02-20T09:41:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,872,162 | 9,223 | Akazz | Failed tests are false negatives. | 2020-02-20T09:52:34 | 2020-02-20T09:52:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,363,874 | 9,161 | abyss7 | > In `daemon` and `mysqlxx` the directories `include` and `src` remain.
> Let's merge together?
Sure, I will do this in next PR | 2020-02-18T09:31:23 | 2020-02-18T09:31:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,184,331 | 8,925 | morozovsk | Возможно проблема была в структуре таблицы, она содержит текстовое поле query, но не содержит его в "ORDER BY". Возможно для движка SummingMergeTree в новых версиях кликхауса это вызывает проблему.
Исправил структуру таблицы, пока что ошибка не воспроизводится. | 2020-02-19T11:51:03 | 2020-02-19T11:51:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,270,430 | 8,929 | compasses | Hello, cannot open the address: https://yandex.ru/legal/cla/?lang=en, which need yandex passport, so it's unavailable outside? | 2020-02-19T14:47:32 | 2020-02-19T14:47:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,277,342 | 8,990 | abyss7 | We don't support build with Apple Clang. Please, get the Clang Compiler Suite from LLVM - via Brew, downloading binary from their website, or building it on your own. | 2020-02-19T15:00:42 | 2020-02-19T15:00:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,306,681 | 9,228 | yohannj | I do not have the rights to update the PR labels.
Can someone specify the label `documentation` please? | 2020-02-19T16:21:56 | 2020-02-19T16:21:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,319,017 | 9,200 | den-crane | If you do like this
```
create table XX (timestamp String, date String default substring(timestamp,1,10), K Int64, S String )
Engine=MergeTree
partition by date
order by (date, K, timestamp);
insert into XX(timestamp, K) select '2019-01-01 00:00:00', number from numbers(1000000);
insert into XX(timestamp, K) s... | 2020-02-19T16:44:37 | 2020-02-19T16:45:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,884,682 | 9,211 | alexey-milovidov | ClickHouse build check — 16/16 builds are OK | 2020-02-20T10:02:16 | 2020-02-20T10:02:16 | {} | MEMBER |
ClickHouse | ClickHouse | 587,950,974 | 8,505 | den-crane | @weiyongyuan you put pictures in the wrong order | 2020-02-18T23:11:15 | 2020-02-18T23:11:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,397,961 | 9,205 | dependabot-preview[bot] | OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting `@dependabot ignore this major version` or `@dependabot ignore this minor version`.
If you change your mind, just re-... | 2020-02-19T19:17:19 | 2020-02-19T19:17:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,428,922 | 9,217 | alesapin | own tests failed :( | 2020-02-19T20:18:51 | 2020-02-19T20:18:51 | {} | MEMBER |
ClickHouse | ClickHouse | 588,573,071 | 8,915 | zhang2014 | @alexey-milovidov
This looks like an `InterpreterExplain` bug (SelectOption is not pushed into the analysis of the Join expression). I will try to fix it .
``` sql
ANALYZE SELECT *
FROM
(
SELECT *
FROM test_00597
)
ANY RIGHT JOIN
(
SELECT *
FROM
(
SELECT *
FROM ... | 2020-02-20T02:10:41 | 2020-02-20T02:10:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,507,625 | 9,181 | den-crane | >Does inserting data into the distribution table cause stress on zookeeper？
No. Distributed table does not work with ZK.
Only Replicated* tables use ZK for replication / merge / mutation coordination
| 2020-02-18T15:11:12 | 2020-02-18T15:11:12 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,556,864 | 9,188 | alexey-milovidov | @Slach You can specify TTL for system tables, see https://github.com/ClickHouse/ClickHouse/pull/8794
@qoega But metric_log is very small table, it's unlikely that anyone will ever need TTL for it. | 2020-02-18T16:44:18 | 2020-02-18T16:44:18 | {} | MEMBER |
ClickHouse | ClickHouse | 588,096,577 | 9,161 | alexey-milovidov | @azat Ok. | 2020-02-19T08:28:22 | 2020-02-19T08:28:22 | {} | MEMBER |
ClickHouse | ClickHouse | 588,123,579 | 7,340 | alexey-milovidov | @akuzm All `median` functions are synonims of the corresponding `quantile` functions with parameter `level` set to 0.5. | 2020-02-19T09:34:17 | 2020-02-19T09:34:17 | {} | MEMBER |
ClickHouse | ClickHouse | 588,391,307 | 9,219 | alexey-milovidov | ```
../contrib/FastMemcpy/FastMemcpy.c:24:2: error: it can only be compiled under windows or unix
#error it can only be compiled under windows or unix
^
``` | 2020-02-19T19:03:06 | 2020-02-19T19:03:06 | {} | MEMBER |
ClickHouse | ClickHouse | 588,414,436 | 9,200 | den-crane | >except our where clause was also using substring(timestamp,1,10).
So you doing it wrong again. See in my example `Processed 1.00 million rows` VS your `Processed 2.00 million rows`
CH stores in parts minmax values of timestamp '2019-01-01 00:00:00' and stores substring(timestamp,1,10) '2019-01-01' **NOWHERE**. S... | 2020-02-19T19:48:08 | 2020-02-19T19:48:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,576,479 | 9,162 | kreuzerkrieg | > We made analysis some time ago on what consumes the most time while compiling CH: https://pastebin.com/VQsJXKkm
>
> Looks like you did everything correct, but just try to precompile the most expensive headers to get some improvements.
Will try these under `Expensive headers`
The question how frequently it chan... | 2020-02-18T17:23:08 | 2020-02-18T17:23:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,933,871 | 9,176 | den-crane | >(event_dt >= 20191123) AND (event_dt <= 20191125)
also OK.
check
```
select latest_fail_reason from system.mutations where is_done=0 limit 3;
select count() from system.replication_queue
```
at all other nodes.
You can restart mutations.
`detach table dm_user_behavior_events; attach table dm_user_... | 2020-02-18T22:22:47 | 2020-02-18T22:23:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,424,762 | 9,200 | den-crane | ah. OK. :thumbsup: | 2020-02-19T20:10:05 | 2020-02-19T20:10:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,731,792 | 9,236 | dependabot-preview[bot] | OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting `@dependabot ignore this major version` or `@dependabot ignore this minor version`.
If you change your mind, just re-... | 2020-02-20T08:10:00 | 2020-02-20T08:10:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,878,083 | 9,229 | alexey-milovidov | Have question in TG chat. | 2020-02-20T09:57:02 | 2020-02-20T09:57:02 | {} | MEMBER |
ClickHouse | ClickHouse | 588,955,442 | 9,198 | nikitamikhaylov | `Build failed without build log. It's inner CI problem. Maybe something wrong with docker or host. Just wait for build restart.`
The last commit is only style improvements. Other test are broken in master. | 2020-02-20T10:55:01 | 2020-02-20T10:55:13 | {} | MEMBER |
ClickHouse | ClickHouse | 587,473,431 | 8,962 | alexey-milovidov | The issue is related to this PR https://github.com/ClickHouse/ClickHouse/pull/8370
It is questionable and we're considering that it can be partially or fully reverted. | 2020-02-18T14:00:16 | 2020-02-18T14:00:16 | {} | MEMBER |
ClickHouse | ClickHouse | 587,704,720 | 9,193 | DarkWanderer | @den-crane пример про where есть, в конце "creating a table" | 2020-02-18T19:27:24 | 2020-02-18T19:27:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,026,495 | 9,195 | den-crane | >How does clickhouse deduplicate the same rows without so-called “primary key”? Is there a md5 operation for each row or something like that? Does it happen in merge-tree merging stage?
You do insert: `insert into someREPLICATEDtable values (1),(2),(3);`
CH calculates a checksum from `someREPLICATEDtable|columns|ty... | 2020-02-19T04:14:35 | 2020-02-19T04:18:24 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,065,319 | 8,505 | weiyongyuan | > @weiyongyuan you put pictures in the wrong order
oh sorry,I already put them correctly | 2020-02-19T06:56:51 | 2020-02-19T06:56:51 | {} | NONE |
ClickHouse | ClickHouse | 588,274,616 | 9,223 | alesapin | LGTM | 2020-02-19T14:55:40 | 2020-02-19T14:55:40 | {} | MEMBER |
ClickHouse | ClickHouse | 588,275,365 | 8,929 | alexey-milovidov | @compasses it's not required to sign the CLA. You can just read and copy the https://developercertificate.org/ instead. | 2020-02-19T14:57:04 | 2020-02-19T14:57:04 | {} | MEMBER |
ClickHouse | ClickHouse | 588,391,850 | 9,198 | alexey-milovidov | Style check... | 2020-02-19T19:04:11 | 2020-02-19T19:04:11 | {} | MEMBER |
ClickHouse | ClickHouse | 588,563,810 | 9,234 | templarzq | [1297/4003] Building CXX object contrib/arrow-cmake/CMakeFiles/thrift_static.dir/__/thrift/lib/cpp/src/thrift/concurrency/Util.cpp.o
FAILED: contrib/arrow-cmake/CMakeFiles/thrift_static.dir/__/thrift/lib/cpp/src/thrift/concurrency/Util.cpp.o
/data/zhangqiang/softwares/gcc-9.2.0/bin/g++ -DARROW_WITH_LZ4 -DARROW_WITH_... | 2020-02-20T01:31:57 | 2020-02-20T01:31:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,628,734 | 6,267 | winter7 | I'm wondering how do you calculate the number of znodes
> When there are 300 tables(20 shards and 2 replicas) and the partition expression is yyyyMMdd type, each table ttl is 1 year, then we get 20 * 300 * 365 + 20 * 300 * 365 + 20 * 2 * 300 * 365 * 2 + = 13,140,000 znodes, 130million znodes when 200 shards! | 2020-02-20T06:15:26 | 2020-02-20T06:15:26 | {} | NONE |
ClickHouse | ClickHouse | 588,726,080 | 9,159 | vpanfilov | Sorry, but I can't provide more detailed stack trace. But I am pretty sure that the actual error (_Block structure mismatch in RemoteBlockOutputStream stream: different types_) is triggered here: https://github.com/ClickHouse/ClickHouse/blob/575ddefa6c32c692f2db7ee8688ca586f3b1d622/dbms/src/Core/Block.cpp#L479-L481
... | 2020-02-20T08:05:59 | 2020-02-20T08:08:15 | {} | NONE |
ClickHouse | ClickHouse | 587,431,350 | 8,505 | jackpgao | Same problem | 2020-02-18T12:13:03 | 2020-02-18T12:13:03 | {} | NONE |
ClickHouse | ClickHouse | 587,560,980 | 8,821 | alexey-milovidov | There are two issues in one:
1. https://github.com/ClickHouse/ClickHouse/pull/9048 (fixed)
2. Index analysis works wrong when there are non-strict comparisons inside NOT (that is used to interpret ALTER DELETE). It's in progress, the fix from @Akazz is almost ready. | 2020-02-18T16:51:35 | 2020-02-18T16:51:35 | {} | MEMBER |
ClickHouse | ClickHouse | 587,718,809 | 5,488 | JacobHayes | This would be great to guarantee (we're starting to generate and load data into clickhouse with fingerprints from other systems). I'll try to add a `farmFingerprint64` like [this one](https://github.com/ClickHouse/ClickHouse/blob/5e29fe3425a6dcb331a7a7abdb2d2973e4685d65/dbms/src/Functions/FunctionsHashing.h#L434-L444) ... | 2020-02-18T19:36:53 | 2020-02-18T19:36:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,058,552 | 9,178 | yuzhichang | @qoega @alexey-milovidov The failing case 0_stateless/00974_query_profiler.sql looks unstable. It passes on my local env, and doesn't relate to this pr change. | 2020-02-19T06:31:18 | 2020-02-19T06:31:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,585,608 | 9,195 | BernieJiangDL | > There is undocumented MergeTree engine setting `replicated_deduplication_window`.
Thanks for your respond. I will check it out. | 2020-02-20T03:02:53 | 2020-02-20T03:02:53 | {} | NONE |
ClickHouse | ClickHouse | 588,972,540 | 9,245 | qoega | @BayoNet do we use table syntax in docs? Probably it will be easier to read as a table | 2020-02-20T11:24:16 | 2020-02-20T11:24:16 | {} | MEMBER |
ClickHouse | ClickHouse | 587,210,240 | 8,994 | filimonov | > @filimonov The best available domain-specific generator is in `clickhouse-obfuscator` tool.
Disadvantages: string fragments leakage (unacceptable in many cases as it's hard to prove the absence of sensitive data), breaks JSON / URLs / other well-formatted strings. (I know I want too much :) ), lack of possibility ... | 2020-02-18T00:09:17 | 2020-02-18T09:53:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,402,800 | 9,161 | azat | After `libs/libdaemon` has been moved to `base/libdaemon`, `libs/libdaemon` can be removed from the include directories (seems that it does not used anyway), i.e. here https://github.com/ClickHouse/ClickHouse/blob/9e4b29fe6ce271993b1d07a6d520e19a6adf8cca/dbms/programs/server/CMakeLists.txt#L26 | 2020-02-18T10:57:08 | 2020-02-18T10:57:08 | {} | MEMBER |
ClickHouse | ClickHouse | 587,506,223 | 9,180 | den-crane | >Is the clickhouse cluster too dependent on zookeeper？
YES. CH supports and works properly only with ZK.
https://github.com/ClickHouse/ClickHouse/issues/777
https://github.com/ClickHouse/ClickHouse/issues/479
>With the growth of shard and data, will the pressure of zookeeper increase linearly?
With proper ... | 2020-02-18T15:08:18 | 2020-02-18T15:08:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,276,279 | 9,195 | den-crane | >Any solutions for this case? Many thanks~
Just do not use spark jdbc to insert into CH. Generate TSV files by Spark and upload them using custom code (&jdbc) which understands CH limits and constraints. | 2020-02-19T14:58:46 | 2020-02-19T14:58:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,393,252 | 9,230 | filimonov | #8915 ? | 2020-02-19T19:07:08 | 2020-02-19T19:07:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,474,027 | 8,962 | alexey-milovidov | About the question "Type name case sensitivity".
We support case insensitive names when they correspond to the SQL standard or when they correspond to the implementation of some popular DBMS (MySQL or Postgres). | 2020-02-18T14:01:36 | 2020-02-18T14:01:36 | {} | MEMBER |
ClickHouse | ClickHouse | 587,492,122 | 8,989 | akuzm | UBSan fuzz test report: https://clickhouse-test-reports.s3.yandex.net/0/3c1735d996594c222a03046eaa033d92f6e6bac1/functional_stateless_tests_(ubsan).html
-- cannot reproduce.
Tried last two queries:
```
:) SELECT CAST(CAST(((SELECT dictGetDate('\0', NULL)), (SELECT log(-708, CAST(CAST(-2.597, 'String'), 'Date'))) ... | 2020-02-18T14:39:59 | 2020-02-18T14:39:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,686,606 | 8,989 | alexey-milovidov | > probably it just calls exit()
Yes, all sanitizers print to stderr and exit. | 2020-02-18T19:15:18 | 2020-02-18T19:15:18 | {} | MEMBER |
ClickHouse | ClickHouse | 588,109,834 | 8,836 | alexey-milovidov | What is the status of this PR? | 2020-02-19T09:01:44 | 2020-02-19T09:01:44 | {} | MEMBER |
ClickHouse | ClickHouse | 588,126,393 | 6,462 | MrSsunlight | I also encountered the same problem. I modified the ttl of the original table to shorten it, and then used OPTIMIZE FINAL, but the expired data was not deleted | 2020-02-19T09:40:48 | 2020-02-19T09:40:48 | {} | NONE |
ClickHouse | ClickHouse | 588,246,951 | 9,215 | alexey-milovidov | @abyss7 FATAL errors are reports for deadly signal (segfault, abort, etc). We transfer data that will be logged through the pipe. See https://github.com/ClickHouse/ClickHouse/issues/5826
It's possible to also transfer pointer to the query and print it. | 2020-02-19T14:02:45 | 2020-02-19T14:02:45 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 587,925,457 | 9,200 | den-crane | You made two mistakes.
* You did not mention partition key (date) in WHERE
* You put date as the first field but again did not mention date in WHERE
Right now you can fix it
```
where
date between '2020-02-08' and '2020-02-08' and
timestamp >= '2020-02-08 00:00:00' and timestamp <= '2020-02-08 23:59:... | 2020-02-18T22:02:55 | 2020-02-18T22:02:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,109,610 | 8,840 | alexey-milovidov | Still failed. | 2020-02-19T09:01:12 | 2020-02-19T09:01:12 | {} | MEMBER |
ClickHouse | ClickHouse | 588,114,274 | 7,859 | alexey-milovidov | Outdated. | 2020-02-19T09:12:27 | 2020-02-19T09:12:27 | {} | MEMBER |
ClickHouse | ClickHouse | 588,227,381 | 9,195 | abyss7 | There is undocumented MergeTree engine setting `replicated_deduplication_window`. | 2020-02-19T13:24:21 | 2020-02-19T13:24:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,398,798 | 9,200 | mvcalder-xbk | Thanks @den-crane for a great explanation. What we had been doing was effectively as you do in your last example, with the partition key `substring(timestamp,1,10)` except our `where` clause was also using `substring(timestamp,1,10)`. Doing that:
```
SELECT count()
FROM XX
WHERE (substring(timestamp, 1, 10) >... | 2020-02-19T19:19:04 | 2020-02-19T19:19:04 | {} | NONE |
ClickHouse | ClickHouse | 587,442,457 | 8,505 | weiyongyuan | same problem.Here is my test result.
First picture is the old version(1.1.54380)'s test result.
Second picture is the new version(19.17.6.36)'s test result
. As I did not replicate the tables, I couldn't restore data. | 2020-02-18T10:38:01 | 2020-02-18T10:38:18 | {} | NONE |
ClickHouse | ClickHouse | 587,472,174 | 9,162 | abyss7 | We made analysis some time ago on what consumes the most time while compiling CH: https://pastebin.com/VQsJXKkm
Looks like you did everything correct, but just try to precompile the most expensive headers to get some improvements. | 2020-02-18T13:57:20 | 2020-02-18T13:57:20 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,484,114 | 9,183 | dependabot-preview[bot] | OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting `@dependabot ignore this major version` or `@dependabot ignore this minor version`.
If you change your mind, just re-... | 2020-02-18T14:23:10 | 2020-02-18T14:23:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,109,974 | 8,817 | alexey-milovidov | ? | 2020-02-19T09:02:04 | 2020-02-19T09:02:04 | {} | MEMBER |
ClickHouse | ClickHouse | 588,110,996 | 8,733 | alexey-milovidov | This PR is incomplete. It doesn't implement multiplexing of connection initiation and handshakes and lead to slower queries. The author has switched to another task. The task is still relevant. | 2020-02-19T09:04:24 | 2020-02-19T09:04:24 | {} | MEMBER |
ClickHouse | ClickHouse | 588,279,105 | 8,290 | CurtizJ | > Need one more test with compact parts and skip indices.
I've added a smoke test to check basic functionality with compact parts and skip indices.
| 2020-02-19T15:03:49 | 2020-02-19T15:03:49 | {} | MEMBER |
ClickHouse | ClickHouse | 588,844,412 | 9,243 | alesapin | Maybe you can provide server logs after `ALTER` query begins? Also, it would be nice if you can provide data to reproduce the OOM. You can send it privately to our mailing lists clickhouse-feedback@yandex-team.ru. | 2020-02-20T09:31:41 | 2020-02-20T09:31:41 | {} | MEMBER |
ClickHouse | ClickHouse | 588,931,982 | 8,989 | akuzm | > It's about dangling reference (use after free) of Context in StorageFile:
The stack is similar to the thing that we've seen under msan recently: https://pastebin.com/raw/tTxCDFKp
@tavplubix knows the reason, but I think he haven't fixed it yet. | 2020-02-20T10:38:08 | 2020-02-20T10:38:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,971,838 | 8,990 | vigneshrajarr | Even tried with llvm compiler, error occurs again
-- The C compiler identification is Clang 8.0.1
-- The CXX compiler identification is Clang 8.0.1
-- Check for working C compiler: /usr/local/Cellar/llvm@8/8.0.1_1/bin/clang
-- Check for working C compiler: /usr/local/Cellar/llvm@8/8.0.1_1/bin/clang -- works
-- D... | 2020-02-20T11:22:37 | 2020-02-20T11:22:37 | {} | NONE |
ClickHouse | ClickHouse | 587,501,462 | 9,195 | den-crane | > My table engine is ReplicatedMergeTree and the records are inserted into ReplicatedMergeTree table directly rather than distributed table. How does it work without transaction guaranties when those exception happens? Do these 1000 records all fail or some success some fail?
https://clickhouse.tech/docs/en/operat... | 2020-02-18T14:58:39 | 2020-02-18T14:58:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,747,562 | 9,193 | den-crane | > @den-crane пример про where есть, в конце "creating a table"
аа, не заметил. Но я бы упомянул что необходимо явно задавать для WHERE всегда toDateTime64 или cast()
на этом примере неочевидно,
`WHERE timestamp = toDateTime64('2019-01-01 00:00:00', 3, 'Europe/Moscow')`
что придется писать `timestamp = toDateTim... | 2020-02-18T19:56:39 | 2020-02-18T19:56:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 587,995,148 | 9,176 | onine007 | @den-crane I restart mutations ,but also no data in table system.replication_queue and no change
in table system.mutations ;
It has not this problem in my test environment .
Test environment has less partitions than prod , but the same partition data in 2 environments .
| 2020-02-19T01:55:43 | 2020-02-19T01:55:43 | {} | NONE |
ClickHouse | ClickHouse | 588,113,412 | 8,377 | alexey-milovidov | ? | 2020-02-19T09:10:26 | 2020-02-19T09:10:26 | {} | MEMBER |
ClickHouse | ClickHouse | 588,142,537 | 8,817 | KochetovNicolai | I still don't know if this pr is needed. Will fix it later. | 2020-02-19T10:12:54 | 2020-02-19T10:12:54 | {} | MEMBER |
ClickHouse | ClickHouse | 588,321,994 | 9,176 | den-crane | I advised to restart mutations and read clickhouse log (trace) what is going with mutations and where they stuck | 2020-02-19T16:50:25 | 2020-02-19T16:50:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,400,510 | 9,198 | nikitamikhaylov | I know :) They are not mine. | 2020-02-19T19:22:19 | 2020-02-19T19:22:19 | {} | MEMBER |
ClickHouse | ClickHouse | 588,586,505 | 9,234 | templarzq | operations:
git clone clickhouse on windows
then copy files to centos7
| 2020-02-20T03:07:17 | 2020-02-20T03:07:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 588,880,862 | 9,221 | alexey-milovidov | > No space left on device
> Cannot find name of filesystem by mount point /var/lib/clickhouse/metadata/default/ints.sql.tmp
This is CI glitch. | 2020-02-20T09:59:08 | 2020-02-20T09:59:08 | {} | MEMBER |
ClickHouse | ClickHouse | 604,111,605 | 9,837 | filimonov | Please replace the entypoint with something simple like bash, and check if $CLICKHOUSE_CONFIG file is readable and `clickhouse extract-from-config --config-file $CLICKHOUSE_CONFIG --key=logger.log` returns good value. | 2020-03-25T22:04:31 | 2020-03-25T22:04:31 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.