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 | 662,970,297 | 7,152 | alexey-milovidov | It's no longer relevant after implementing 100% immutable parts. | 2020-07-23T12:09:16 | 2020-07-23T12:09:16 | {} | MEMBER |
ClickHouse | ClickHouse | 663,195,357 | 12,737 | siradjev | Select is slower by no more than 30% in case of simulated test...
```
drop table if exists foo_compact;
CREATE TABLE foo_compact
(
`f1` String,
`f2` String,
`f3` Int64,
`f4` Float64,
`f5` Int32,
`f6` Int32,
`date` Date
)
ENGINE = MergeTree()
ORDER BY (f2, f3, date, f5, ... | 2020-07-23T19:39:20 | 2020-07-23T19:39:20 | {} | NONE |
ClickHouse | ClickHouse | 663,244,920 | 12,451 | azat | Looks like a duplicate of #5818 | 2020-07-23T21:37:12 | 2020-07-23T21:37:12 | {} | MEMBER |
ClickHouse | ClickHouse | 663,250,738 | 12,740 | grmorozov | Thank you! I'll use this workaround.
My point was that it works with older versions. Was it intended backward incompatible change? | 2020-07-23T21:52:51 | 2020-07-23T21:52:51 | {} | NONE |
ClickHouse | ClickHouse | 663,346,632 | 12,707 | ifedorov | Thanks, I have read doc, but few moments:
1) I have used Final
2) this table merged every hour successfully
3) there is not a lot data, now count(): 26826
And it still do not merge with different date in ver during - 3 days
| 2020-07-24T04:57:40 | 2020-07-24T04:57:40 | {} | NONE |
ClickHouse | ClickHouse | 663,391,676 | 12,709 | meimeitou | > no kafka engine ?
>
> ```
> Code: 56. DB::Exception: Received from localhost:9000. DB::Exception: Unknown table engine Kafka.
> ```
does kafka engine not support arrch64? | 2020-07-24T07:43:53 | 2020-07-24T07:43:53 | {} | NONE |
ClickHouse | ClickHouse | 663,450,094 | 12,661 | jmalicevic | DB::Exception: Invalid namespace: my-namespace: while fetching schema id = X: (Input format doesn't allow to skip errors): (at row 1)
: While executing SourceFromInputStream.
The query ends with:
ENGINE = Kafka()
SETTINGS kafka_broker_list = 'brokre_host:port', kafka_topic_list = 'my-namespace.prod.topic', ka... | 2020-07-24T09:35:27 | 2020-07-24T09:35:27 | {} | NONE |
ClickHouse | ClickHouse | 663,513,294 | 12,734 | alexey-milovidov | Wrong category. | 2020-07-24T12:25:10 | 2020-07-24T12:25:10 | {} | MEMBER |
ClickHouse | ClickHouse | 662,893,819 | 12,709 | meimeitou | change file `cmake/find/zlib.cmake` line 2:
```shell
if (NOT OS_FREEBSD AND NOT ARCH_32)
option (USE_INTERNAL_ZLIB_LIBRARY "Set to FALSE to use system zlib library instead of bundled" ${NOT_UNBUNDLED})
endif ()
```
to
```shell
if (NOT OS_FREEBSD AND NOT ARCH_32)
option (USE_INTERNAL_ZLIB_LIBRARY "Set t... | 2020-07-23T08:54:33 | 2020-07-23T09:53:30 | {} | NONE |
ClickHouse | ClickHouse | 662,921,588 | 12,664 | 4ertus2 | > Seems that clone of ast is still needed.
I've made a solution without `clone()`. Clone after `normalize()` breaks indirect assumption that the aliases links to the same AST object. This assumtion should not be used later. But all our tests are for scenario where we do not break this assumption. So it's very easy t... | 2020-07-23T10:01:13 | 2020-07-23T12:15:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,192,233 | 12,734 | alexey-milovidov | But tests did not run, let's edit some whitespaces in .cpp file :) | 2020-07-23T19:32:03 | 2020-07-23T19:32:14 | {} | MEMBER |
ClickHouse | ClickHouse | 663,353,225 | 12,709 | meimeitou | change file `src/Functions/FunctionBinaryArithmetic.h`:73
```
static void NO_INLINE vector_constant(const A * __restrict a, B b, ResultType * __restrict c, size_t size)
{
return;
for (size_t i = 0; i < size; ++i)
c[i] = Op::template apply<ResultType>(a[i], b);
}
stat... | 2020-07-24T05:28:41 | 2020-07-24T05:28:41 | {} | NONE |
ClickHouse | ClickHouse | 663,436,969 | 12,661 | HristoMohamed | Hi,
(I am from the same company that jmalicevic is from). I investigated this a bit further and this is basically problem with the Avro namespace naming convention. According to the Avro docs:
`The name portion of a fullname, record field names, and enum symbols must:
start with [A-Za-z_] .
subsequently conta... | 2020-07-24T09:14:17 | 2020-07-24T09:20:44 | {
"+1": 1,
"heart": 1
} | NONE |
ClickHouse | ClickHouse | 662,904,564 | 11,501 | andrea-s | So given the above, what would be the current best practice to handle an ETL in which the last x days need be updated daily?
We are currently working with partition swap, I've considered switching to use the alter statements but it doesn't look like there's a way to guarantee atomic updates at this time, correct? As... | 2020-07-23T09:18:39 | 2020-07-23T09:18:39 | {} | NONE |
ClickHouse | ClickHouse | 663,428,895 | 12,661 | azat | >I tried that but the topic does not allow it.
Not sure how is it possible? Can you post the query and the error | 2020-07-24T09:01:21 | 2020-07-24T09:01:21 | {} | MEMBER |
ClickHouse | ClickHouse | 663,554,264 | 12,492 | CurtizJ | Perf tests look ok.
Insert:
Select:
| 2020-07-24T14:02:51 | 2020-07-24T14:02:51 | {} | MEMBER |
ClickHouse | ClickHouse | 662,908,485 | 11,501 | andrea-s | Mine is a little bit easier than yours (just every day we reprocess the last x days for some sources, as data could be fully reliable only with some delay - e.g. google search console data). So our "rewrites" are always based on the primary key... Thanks for the tip about mutation_sync, didn't know about that.
For w... | 2020-07-23T09:28:05 | 2020-07-23T09:28:05 | {} | NONE |
ClickHouse | ClickHouse | 662,927,856 | 12,709 | meimeitou | ```c++
ClickHouse/src/Functions/CRC.cpp:66:14: error: ‘crc32_z’ was not declared in this scope; did you mean ‘crc32’?
66 | { return crc32_z(0L, buf, size); }
```
not found zlib.h | 2020-07-23T10:16:54 | 2020-07-23T10:17:39 | {} | NONE |
ClickHouse | ClickHouse | 662,966,000 | 12,689 | alexey-milovidov | Already backported. | 2020-07-23T11:59:06 | 2020-07-23T11:59:06 | {} | MEMBER |
ClickHouse | ClickHouse | 663,043,916 | 8,979 | alesapin | Hi, sorry for the long answer. It seems like index granularity is not relevant to this issue. This error says that checksum in file `checksums.txt` is absent for file `<column>.mrk`. So probably you have broken checksums in your back, but to investigate I'd like to look at `checksums.txt` file. Also, it's not clear, bu... | 2020-07-23T14:36:34 | 2020-07-23T14:36:34 | {} | MEMBER |
ClickHouse | ClickHouse | 663,194,102 | 12,737 | alexey-milovidov | Yes.
Reading a column from big compact part will involve more disk seeks.
It should be noticeably slower for HDD and just a little slower on SSD.
I'm not 100% sure that 2x difference is Ok, though.
I expect that #12492 will eliminate most of the difference, especially for SSD.
But about 1.1x difference is ex... | 2020-07-23T19:36:21 | 2020-07-23T19:36:21 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 663,373,827 | 12,602 | millin | I have the same error
ver 20.5.3.27 | 2020-07-24T06:51:34 | 2020-07-24T06:51:34 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,529,639 | 12,717 | qoega | Nice. I've made `grep Fatal` after integrational tests and missed that this one was intentional :) | 2020-07-24T13:08:21 | 2020-07-24T13:08:21 | {} | MEMBER |
ClickHouse | ClickHouse | 662,972,094 | 8,661 | alexey-milovidov | > In case of empty table it should create empty structure of folders
This is easy to implement. Maybe separate it into abother task? | 2020-07-23T12:13:37 | 2020-07-23T12:13:37 | {} | MEMBER |
ClickHouse | ClickHouse | 663,025,616 | 12,681 | alexey-milovidov | But I've already backported it. | 2020-07-23T14:03:10 | 2020-07-23T14:03:10 | {} | MEMBER |
ClickHouse | ClickHouse | 663,026,830 | 4,300 | alexey-milovidov | You have mistakenly execute FETCH PARTITION instead of FREEZE PARTITION.
| 2020-07-23T14:05:18 | 2020-07-23T14:05:18 | {} | MEMBER |
ClickHouse | ClickHouse | 663,186,711 | 12,738 | alexey-milovidov | Examples: `ProfileEvents.Names`, `user`, `http_user_agent`... | 2020-07-23T19:19:10 | 2020-07-23T19:19:10 | {} | MEMBER |
ClickHouse | ClickHouse | 663,197,288 | 12,737 | kreuzerkrieg | > Yes.
>
> Reading a column from big compact part will involve more disk seeks.
> It should be noticeably slower for HDD and just a little slower on SSD.
>
> I'm not 100% sure that 2x difference is Ok, though.
> I expect that #12492 will eliminate most of the difference, especially for SSD.
>
> But about 1.1... | 2020-07-23T19:44:15 | 2020-07-23T19:44:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,241,287 | 12,707 | den-crane | https://clickhouse.tech/docs/en/engines/table-engines/mergetree-family/replacingmergetree/
_Data deduplication occurs only during a merge. Merging occurs in the background at an unknown time, so you can’t plan for it. Some of the data may remain unprocessed. Although you can run an unscheduled merge using the OPTIMI... | 2020-07-23T21:28:02 | 2020-07-23T21:28:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,428,581 | 12,663 | 4ertus2 | It should be backported into 20.3-6 in disabled state. | 2020-07-24T09:00:49 | 2020-07-24T09:01:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,527,920 | 12,436 | alex-zaitsev | Related #8089 | 2020-07-24T13:04:22 | 2020-07-24T13:04:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 662,847,520 | 8,331 | vkingnew | this function is great，when release it as a stable version? | 2020-07-23T06:53:15 | 2020-07-23T06:53:15 | {} | NONE |
ClickHouse | ClickHouse | 662,877,871 | 12,667 | MisterTimn | I've checked and found that CH was still using an old Zookeeper URL which was already changed in the CHI resource on the cluster. So I changed that again (to the fully qualified name as well) and I'm not seeing any references to Zookeeper in logs now.
Tables are still in readonly mode on all nodes though. | 2020-07-23T08:15:26 | 2020-07-23T08:15:26 | {} | NONE |
ClickHouse | ClickHouse | 663,039,374 | 12,731 | akuzm | Good intro to gettext: https://www.labri.fr/perso/fleury/posts/programming/a-quick-gettext-tutorial.html
The command I used to reset: `for x in es fa fr ja ru tr zh ; do msginit --input=messages.pot --locale=$x --no-translator --output=$x/LC_MESSAGES/messages.po ; done` | 2020-07-23T14:28:29 | 2020-07-23T14:28:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,207,762 | 12,589 | bobrik | @KochetovNicolai, thanks for taking care of this! One issue I'm seeing with my manual testing is that some logging is missing now, in particular there are no more messages for PK search or skipping indices:
* `Used optimized inclusion search ...`
* `Index X has dropped Y granules`
These were super useful when op... | 2020-07-23T20:09:10 | 2020-07-23T20:09:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,429,809 | 12,709 | alexey-milovidov | > does kafka engine not support arrch64?
Yes, Kafka is disabled on AArch64, see `cmake/find/rdkafka.cmake`.
It's possible to make it work if we provide appropriate version of `contrib/librdkafka-cmake/config.h`. | 2020-07-24T09:02:57 | 2020-07-24T09:04:02 | {} | MEMBER |
ClickHouse | ClickHouse | 662,886,815 | 12,694 | alesapin | **For normal MergeTree:**
The only way is to call `SYSTEM STOP MERGES table_name`. **All merges** will be aborted, after that, you should start them again with `SYSTEM START MERGES table_name`. After that normal merges will start again, without FINAL merges from OPTIMIZE.
**For ReplicatedMergeTree:**
It's not pos... | 2020-07-23T08:37:53 | 2020-07-23T08:41:43 | {
"+1": 2,
"heart": 2
} | MEMBER |
ClickHouse | ClickHouse | 662,924,460 | 12,485 | Jokser | @excitoon Current HTTP client is https://github.com/ClickHouse/ClickHouse/blob/master/src/IO/S3/PocoHTTPClient.cpp . The reason for implementing this on Poco was memory issue when all response is copied into memory | 2020-07-23T10:08:23 | 2020-07-23T10:08:23 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,039,274 | 12,709 | alexey-milovidov | You can edit the file `ClickHouse/src/Functions/CRC.cpp` and disable CRC function if you build with system zlib. | 2020-07-23T14:28:19 | 2020-07-23T14:28:19 | {} | MEMBER |
ClickHouse | ClickHouse | 663,111,693 | 12,492 | CurtizJ | I also added one change from #12183 to this PR. Now compressed block is written for every part of column in granule. It should significanty increase performance of reading, but maybe writing will slow down. Let's look to perf tests. | 2020-07-23T16:42:38 | 2020-07-23T16:42:50 | {} | MEMBER |
ClickHouse | ClickHouse | 663,213,118 | 12,589 | KochetovNicolai | @bobrik ok, I'll check it.
No logging have I removed, so this is suspicious. | 2020-07-23T20:21:33 | 2020-07-23T20:21:33 | {} | MEMBER |
ClickHouse | ClickHouse | 663,253,361 | 12,740 | grmorozov | Also, is it enough to alter existing MV to upgrade CH version, or do I need to recreate it? | 2020-07-23T21:59:58 | 2020-07-23T21:59:58 | {} | NONE |
ClickHouse | ClickHouse | 663,277,719 | 12,740 | den-crane | >My point was that it works with older versions.
I understand.
>Was it intended backward incompatible change?
Probably it was intentional change to fix something. But it became an issue in your case. Unfortunately it seems it does not covered by integration tests. And actually you are the first who complain a... | 2020-07-23T23:18:10 | 2020-07-23T23:18:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,408,818 | 12,661 | jmalicevic | Just to follow up, it can be that the 'dash' in the namespace name is actually not acceptable and the Avro serializer is doing its job. But we cannot remove this at this point from the namespace. If this is the case, is there a way to work around this and disable the name validation? We have successfully read from this... | 2020-07-24T08:30:11 | 2020-07-24T08:30:11 | {} | NONE |
ClickHouse | ClickHouse | 663,515,952 | 12,742 | alexey-milovidov | > I remember that there was discussion around adding unbundled+shared+split build on CI, any progress on this?
Currently no progress.
> Or maybe I can help to speed up this process?
We need to move most of the scripts out of "Sandbox" infrastructure to our repository: https://github.com/ClickHouse/ClickHouse/i... | 2020-07-24T12:32:51 | 2020-07-24T12:32:51 | {} | MEMBER |
ClickHouse | ClickHouse | 662,895,295 | 12,342 | alexey-milovidov | 20.5.3 has been released. | 2020-07-23T08:57:59 | 2020-07-23T08:57:59 | {} | MEMBER |
ClickHouse | ClickHouse | 662,926,740 | 12,716 | xiashaohua | it is my mistake,i did not use ssl | 2020-07-23T10:14:04 | 2020-07-23T10:14:04 | {} | NONE |
ClickHouse | ClickHouse | 663,151,642 | 12,672 | azat | >LGTM, but what's about integration test?
Does related, fixed (pops up after removing `std::bind`) | 2020-07-23T18:03:49 | 2020-07-23T18:03:49 | {} | MEMBER |
ClickHouse | ClickHouse | 663,245,020 | 11,501 | den-crane | When you execute `ALTER TABLE .... WHERE date=today()` mutation locks a table for a microsecond and creates a list of active parts and stores this list into mutations.txt or zookeeper (if the table is replicated). After that mutations begins to mutate these parts from the list and so it does not see new parts inserted ... | 2020-07-23T21:37:27 | 2020-07-23T21:38:30 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 662,885,555 | 12,656 | alesapin | As far as I remember there was several problems with simultaneous MODIFY and RENAME. I'll add more tests. | 2020-07-23T08:34:40 | 2020-07-23T08:34:40 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 663,069,222 | 12,731 | blinkov | @akuzm Docs check has not been launched for some reason while this might easily break the website build. These files were generated by `pybabel` instead of raw gettext tools, so the output differs significantly and `*.mo` are deleted for some reason (their compilation has not been added to the website build pipeline ye... | 2020-07-23T15:21:52 | 2020-07-23T15:21:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,311,898 | 12,694 | zhlovy | thanks ~
Thanks a lot for the trouble you've taken to help me! | 2020-07-24T01:41:14 | 2020-07-24T01:41:14 | {} | NONE |
ClickHouse | ClickHouse | 663,427,428 | 12,709 | alexey-milovidov | > change file src/Functions/FunctionBinaryArithmetic.h:73
Don't do that, consequences will be dreadful.
(Arithmetic operations will return uninitialized memory, what can be worse?)
\+ this code should not compile unless you have disabled warnings about unreachable code.
Internal compiler error means bug in comp... | 2020-07-24T08:59:05 | 2020-07-24T09:06:25 | {} | MEMBER |
ClickHouse | ClickHouse | 663,524,411 | 12,707 | den-crane | >I have used Final
Oops. I've missed that.
group_uniq_aggregated_statistic VS group_uniq_aggregated_statistic_local
So you have the same key at 2 different shards.
Distributed table does not collapse them and it does not understand `Final `.
MergeTree tables also know nothing about sharding and they are comp... | 2020-07-24T12:55:56 | 2020-07-24T12:57:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 662,872,597 | 12,706 | dependabot-preview[bot] | Looks like urllib3 is up-to-date now, so this is no longer needed. | 2020-07-23T08:02:35 | 2020-07-23T08:02:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 662,986,019 | 10,294 | alexey-milovidov | 🎉 SQLancer has been successfuly integrated and already found some issues:
https://twitter.com/sqlancer_dbms/status/1286211595455082497
Thanks, @qoega. | 2020-07-23T12:46:31 | 2020-07-23T12:46:31 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 663,001,881 | 12,672 | alexey-milovidov | LGTM, but what's about integration test? | 2020-07-23T13:18:32 | 2020-07-23T13:18:32 | {} | MEMBER |
ClickHouse | ClickHouse | 663,044,362 | 12,715 | alexey-milovidov | CC @zhang2014 | 2020-07-23T14:37:18 | 2020-07-23T14:37:18 | {} | MEMBER |
ClickHouse | ClickHouse | 663,334,313 | 12,616 | ilovesoup | > Could you describe the feature in more detail?
>
> CH makes distributed aggregation and distributed JOIN/IN in different way. But distributed pipeline is mostly the same: you have server initiator for the query. It passed query (probably in modified form) to other nodes via Distributed table. Then nodes send their... | 2020-07-24T03:45:01 | 2020-07-24T03:45:01 | {} | NONE |
ClickHouse | ClickHouse | 663,513,197 | 12,734 | alexey-milovidov | Looks totaly alright, I'll merge. | 2020-07-24T12:24:52 | 2020-07-24T12:24:52 | {} | MEMBER |
ClickHouse | ClickHouse | 662,971,532 | 8,661 | alexey-milovidov | > Support for ALTER TABLE ... FREEZE (w/o partitions syntax) for non-merge tree tables - Log familty, Join, Set, Distributed, Buffer etc.
It doesn't fit the logic of FREEZE because files in Log* tables are mutable and hardlinks won't work. It cannot be implemented without locking and file copy (or literally, freezin... | 2020-07-23T12:12:16 | 2020-07-23T12:12:52 | {} | MEMBER |
ClickHouse | ClickHouse | 663,046,130 | 12,713 | alexey-milovidov | Confirm, the flush thread of Buffer table does not use the settings from INSERT.
You can adjust `max_partitions_for_insert_block` globally (for default profile) to make effect for background threads. | 2020-07-23T14:40:30 | 2020-07-23T14:40:55 | {} | MEMBER |
ClickHouse | ClickHouse | 663,191,133 | 12,492 | alexey-milovidov | > but maybe writing will slow down. Let's look to perf tests.
And if it's Ok let's apply this change unconditionally (without a setting). | 2020-07-23T19:29:28 | 2020-07-23T19:29:28 | {} | MEMBER |
ClickHouse | ClickHouse | 663,199,841 | 12,737 | kreuzerkrieg | > Select is slower by no more than 30% in case of simulated test...
>
> ```
> drop table if exists foo_compact;
> CREATE TABLE foo_compact
> (
> `f1` String,
> `f2` String,
> `f3` Int64,
> `f4` Float64,
> `f5` Int32,
> `f6` Int32,
> `date` Date
> )
> ENGINE = MergeTree()... | 2020-07-23T19:50:23 | 2020-07-23T19:50:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,217,584 | 12,589 | bobrik | I'm using this branch rebased on top of v20.7.1.4140-testing, so maybe there was a change in master that caused it. | 2020-07-23T20:31:33 | 2020-07-23T20:31:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,426,938 | 12,661 | jmalicevic | I tried that but the topic does not allow it. | 2020-07-24T08:58:22 | 2020-07-24T08:58:22 | {} | NONE |
ClickHouse | ClickHouse | 662,833,804 | 10,851 | zhang2014 | @tavplubix Can you help review this PR again? | 2020-07-23T06:04:11 | 2020-07-23T06:04:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 662,868,384 | 12,685 | alesapin | No errors related to changes. | 2020-07-23T07:52:28 | 2020-07-23T07:52:28 | {} | MEMBER |
ClickHouse | ClickHouse | 662,915,633 | 11,501 | jeanatpi | @andrea-s thanks for sharing this. Not sure if this "is_most_recent" will apply to our use case.
Seems that the CollapsingMergeTree approach is working fine to "delete" data and is simplier than ReplacingMergeTree
Interesting thing also for mutations.
Suppose you execute async : ALTER TABLE .... WHERE date=tod... | 2020-07-23T09:45:52 | 2020-07-23T09:45:52 | {} | NONE |
ClickHouse | ClickHouse | 662,966,004 | 12,615 | npetrusic | UPDATE:
I tried replacing the order of attributes in protobuf schema, it now looks like this:
```
syntax = "proto3";
message User {
int32 timestamp = 1;
string username = 2;
}
```
(only replacing the order of attributes) and it now works. It seems that the Kafka + Protobuf + CH doesn't accept a strin... | 2020-07-23T11:59:07 | 2020-07-23T12:00:11 | {} | NONE |
ClickHouse | ClickHouse | 663,117,167 | 8,331 | Vxider | @vkingnew Thanks, we are waiting for a code review. | 2020-07-23T16:53:48 | 2020-07-23T16:53:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,223,664 | 12,740 | den-crane | >time - duration
because of
```
20.7.1.4140
DESCRIBE TABLE
(
SELECT
now() - 5 AS x,
now() AS y,
toDateTime(x, '') AS z
)
┌─name─┬─type────────────────┬─
│ x │ DateTime('Etc/UTC') │
│ y │ DateTime │
│ z │ DateTime │
└──────┴────────────────... | 2020-07-23T20:45:45 | 2020-07-23T20:49:30 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 662,914,635 | 2,156 | raulgregg | When trying to perform a UNION ALL, I'm getting this error:
`DB::Exception: Block structure mismatch in function connect between Converting and Concat stream: different number of columns: id_item String String(size = 0) (version 20.5.2.7 (official build))`
Both tables have the same size in columns (10), data type... | 2020-07-23T09:43:24 | 2020-07-23T09:43:24 | {} | NONE |
ClickHouse | ClickHouse | 662,955,480 | 7,057 | KochetovNicolai | I don't think that #12589 fully implements what is done in this pr. #12589 just adds thread pool to parallelize index scan by part.
| 2020-07-23T11:31:29 | 2020-07-23T11:31:29 | {} | MEMBER |
ClickHouse | ClickHouse | 662,986,168 | 10,294 | alexey-milovidov | Next step is to add it to CI. | 2020-07-23T12:46:49 | 2020-07-23T12:46:49 | {} | MEMBER |
ClickHouse | ClickHouse | 662,987,810 | 8,076 | alexey-milovidov | We decided to go this way: https://github.com/ClickHouse/ClickHouse/pull/8076#pullrequestreview-389025609 | 2020-07-23T12:50:31 | 2020-07-23T12:50:31 | {} | MEMBER |
ClickHouse | ClickHouse | 663,189,509 | 12,737 | kreuzerkrieg | > 3. We don't have `FireTree` engine in ClickHouse (it was in your report before you edited it).
Search and replace issue :) thats why it was edited :) | 2020-07-23T19:25:44 | 2020-07-23T19:25:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,424,483 | 12,732 | azat | Not sure which one of the following PRs fixes it:
- #12075 (I think this one)
- #12302 (or maybe you need this one and above, it is hard to say w/o line numbers in the stacktrace)
- #12658
But I'm pretty sure that it is fixed already (other PRs left just to let you know about known already issues)
P.S. it is b... | 2020-07-24T08:54:33 | 2020-07-24T08:54:33 | {} | MEMBER |
ClickHouse | ClickHouse | 663,426,115 | 12,661 | azat | Maybe `kafka_skip_broken_messages` may help? | 2020-07-24T08:57:04 | 2020-07-24T08:57:04 | {} | MEMBER |
ClickHouse | ClickHouse | 662,874,810 | 12,705 | azat | Since #12418 has been instantly reverted there are conflicts right now | 2020-07-23T08:08:06 | 2020-07-23T08:08:06 | {} | MEMBER |
ClickHouse | ClickHouse | 662,879,737 | 12,671 | alesapin | Changes don't affect performance. | 2020-07-23T08:20:06 | 2020-07-23T08:20:06 | {} | MEMBER |
ClickHouse | ClickHouse | 662,897,978 | 11,501 | jeanatpi | @den-crane thanks for your answer. Just wanted to add more info on that one.
- For the DELETE part (insert into) it's important to use the FINAL keyword. Else we'll insert too many rows (if the table is not merged between 2 ETL round.
- I had some performance issues using ReplacingMergeTree + argMax view. Querying ... | 2020-07-23T09:04:00 | 2020-07-23T09:04:00 | {} | NONE |
ClickHouse | ClickHouse | 662,986,720 | 10,294 | alexey-milovidov | @mrigger
> A T-shirt would be awesome
Easy. Just send your mailing address to clickhouse-feedback@yandex-team.com and we will send a t-shirt! | 2020-07-23T12:48:07 | 2020-07-23T12:48:29 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 663,043,495 | 12,717 | alexey-milovidov | @qoega This test is sending signal 11 manually to the server. | 2020-07-23T14:35:51 | 2020-07-23T14:35:51 | {} | MEMBER |
ClickHouse | ClickHouse | 663,050,515 | 2,156 | KochetovNicolai | @raulgregg could you please, open a separate issue?
Also please add query and table scheme. | 2020-07-23T14:48:13 | 2020-07-23T14:48:13 | {} | MEMBER |
ClickHouse | ClickHouse | 663,249,478 | 12,451 | azat | FWIW as a workaround `multiIf` can be used:
```sql
WITH 'CCC' AS value_
SELECT CAST(multiIf(value_ = 'AAA', 1, value_ = 'BBB', 2, NULL), 'Nullable(Enum(\'AAA\'=1,\'BBB\'=2))')
```
instead of
```sql
SELECT CAST('CCC', 'Nullable(Enum(\'AAA\'=1,\'BBB\'=2))')
``` | 2020-07-23T21:49:22 | 2020-07-23T21:49:37 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 663,338,018 | 12,709 | meimeitou | > You can edit the file `ClickHouse/src/Functions/CRC.cpp` and disable CRC function if you build with system zlib.
after remove CRC function, i meet:
```
during RTL pass: expand
In file included from /work/ClickHouse/src/Functions/bitRotateLeft.cpp:2:
/work/ClickHouse/src/Functions/FunctionBinaryArithmetic.h: In... | 2020-07-24T04:07:42 | 2020-07-24T05:15:00 | {} | NONE |
ClickHouse | ClickHouse | 663,354,982 | 12,709 | meimeitou | no kafka engine ?
```
Code: 56. DB::Exception: Received from localhost:9000. DB::Exception: Unknown table engine Kafka.
``` | 2020-07-24T05:36:39 | 2020-07-24T05:36:39 | {} | NONE |
ClickHouse | ClickHouse | 663,462,907 | 12,661 | HristoMohamed | I investigated a bit the codebase, from what I see in order to use the option suggested by @azat , allowSyncAfterError needs to be added to the input format. While this exists for some input format (like JSONEachRowRowInputFormat, CSVRowInputFormat.h), it is not implemented for AvroRowInputFormat. In any case, it does ... | 2020-07-24T09:57:47 | 2020-07-24T09:57:47 | {} | NONE |
ClickHouse | ClickHouse | 662,834,708 | 7,057 | zhang2014 | It has been implemented in #12589 | 2020-07-23T06:07:40 | 2020-07-23T06:07:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,101,492 | 5,182 | YoannBuch | Hi all, this may be of interest to some of you here: the APM solution [Instana](https://www.instana.com/) now offers this end-to-end visibility inside the ClickHouse cluster.
We talked about it during this [Altinity webinar](https://www.altinity.com/presentations/2020/07/09/adventures-in-observability-how-in-house-c... | 2020-07-23T16:21:49 | 2020-07-23T16:21:49 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 663,191,015 | 12,737 | kreuzerkrieg | > 1. Compact parts should be slightly slower on read. They are intented for small data parts (under ~10MiB) to speed up INSERTs and not to be enabled by default for all data.
> 2. There are known issues, see #12492
So, is it plainly wrong to use it for more formidable parts, like 2-10Gb?
| 2020-07-23T19:29:12 | 2020-07-23T19:29:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,445,471 | 12,709 | meimeitou | @alexey-milovidov thx, that helps a lot. | 2020-07-24T09:27:50 | 2020-07-24T09:27:50 | {} | NONE |
ClickHouse | ClickHouse | 662,835,979 | 12,663 | zhang2014 | > test failure : 00534_functions_bad_arguments2 ../contrib/libcxx/include/vector:1558:12: runtime error: reference binding to null pointer of type 'const std::__1::__vector_base<DB::ColumnWithTypeAndName, std::__1::allocator<DB::ColumnWithTypeAndName> >::value_type' (aka 'const DB::ColumnWithTypeAndName')
is not rel... | 2020-07-23T06:12:19 | 2020-07-23T06:12:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,059,736 | 12,730 | den-crane | JDBC driver issue https://github.com/ClickHouse/clickhouse-jdbc/issues/290 | 2020-07-23T15:04:43 | 2020-07-23T15:04:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,188,027 | 12,737 | alexey-milovidov | 1. Compact parts should be slightly slower on read. They are intented for small data parts (under ~10MiB) to speed up INSERTs and not to be enabled by default for all data.
2. There are known issues, see https://github.com/ClickHouse/ClickHouse/pull/12492
3. We don't have `FireTree` engine in ClickHouse (it was i... | 2020-07-23T19:22:16 | 2020-07-23T19:22:16 | {} | MEMBER |
ClickHouse | ClickHouse | 663,238,166 | 12,667 | den-crane | it should return rows if you have replicated tables
```
CREATE TABLE testAlter
(
`A` Int64,
`T` DateTime,
`X` Float64
)
ENGINE = ReplicatedReplacingMergeTree('/clickhouse/testAlter', 'xxx')
PARTITION BY toStartOfHour(T)
ORDER BY A
select database,table from system.replicas
┌─database─┬─table... | 2020-07-23T21:19:52 | 2020-07-23T21:19:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 663,544,460 | 12,327 | ghost | @nikitamikhaylov
I finally pass tests and build checks.
Docs check doesn't seem to related this PR.
Please review again. Thank you in advance.
Cc. @alexey-milovidov | 2020-07-24T13:41:27 | 2020-07-24T13:41:27 | {} | NONE |
ClickHouse | ClickHouse | 680,040,680 | 14,043 | markpapadakis | @alexey-milovidov
```cpp
Pipe StorageMemory::read(
const Names & column_names,
const StorageMetadataPtr &metadata_snapshot,
const SelectQueryInfo & /*query_info*/,
const Context & /*context*/,
QueryProcessingStage::Enum /*processed_stage*/,
size_t /*max_block_size*/,
un... | 2020-08-25T13:55:53 | 2020-08-26T10:20:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 680,033,546 | 14,045 | radiatorius | > Check the logs, that the thread [ 10231 ] did right before the crash.
it seems this:
```
...
2020.08.25 13:19:36.739060 [ 10231 ] {} <Debug> db_name: Loading mutation: mutation_3.txt entry, commands size: 1
2020.08.25 13:19:36.739140 [ 10231 ] {} <Debug> db_name: Loading mutation: mutation_4.txt entry, command... | 2020-08-25T13:43:26 | 2020-08-26T10:27:29 | {} | NONE |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.