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 | 523,246,324 | 6,564 | alexey-milovidov | Formats are not registered, look at the tests. | 2019-08-21T00:29:01 | 2019-08-21T00:29:01 | {} | MEMBER |
ClickHouse | ClickHouse | 523,317,159 | 6,565 | ChannelerH | > I believe a cause is CREATE MATERIALIZED VIEW.... **SETTINGS index_granularity=8192**
> Try index_granularity=64 ... 16 ... 8
>
> https://gist.github.com/den-crane/32cac4f069f4227b35ab86553bbc649b
>
> Or check adaptive index_granularity in the recent releases.
Firstly, thanks for your reply.I have changed t... | 2019-08-21T06:31:23 | 2019-08-21T06:31:23 | {} | NONE |
ClickHouse | ClickHouse | 523,008,235 | 6,549 | thyn | getfacl
`
file: data/clickhouse/data/actions/actions/
owner: clickhouse
group: clickhouse
user::rwx
group::r-x
other::r-x
`
os CentOS Linux release 7.6.1810 (Core)
umask by default 0022
installed from https://github.com/Altinity/clickhouse-rpm-install
ClickHouse server version 19.13.2 revision 5442... | 2019-08-20T13:13:03 | 2019-08-20T13:16:14 | {} | NONE |
ClickHouse | ClickHouse | 523,017,286 | 6,560 | alexey-milovidov | Minor, because you can use gcc-8 or gcc-9.1 or clang (versions 7, 8, 9 are Ok).
Compiler bugs are quite common. | 2019-08-20T13:34:21 | 2019-08-20T13:34:39 | {} | MEMBER |
ClickHouse | ClickHouse | 523,025,744 | 6,506 | astudnev | Thanks, @den-crane this is probably what i meant. Not sure what "leader" means, however
Does it mean nobody will read data for replicas from this server? | 2019-08-20T13:54:04 | 2019-08-20T13:54:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,162,945 | 6,566 | alexey-milovidov | Duplicate declaration is Ok in C and C++.
But we can add custom rule to our `check-style` script or maybe find something in clang-tidy or PVS-Studio. | 2019-08-20T19:34:55 | 2019-08-20T19:34:55 | {} | MEMBER |
ClickHouse | ClickHouse | 523,256,798 | 6,446 | alexey-milovidov | Previous implementation was correct in terms of idea but technically wrong.
| 2019-08-21T01:25:04 | 2019-08-21T01:25:04 | {} | MEMBER |
ClickHouse | ClickHouse | 523,289,383 | 6,237 | alexey-milovidov | Fixed. But @proller failed to add a test, so test on your own. | 2019-08-21T04:22:30 | 2019-08-21T04:22:30 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 523,296,538 | 6,586 | alexey-milovidov | JOIN in external memory is not implemented. But @4ertus2 is working on this feature right now. | 2019-08-21T05:01:17 | 2019-08-21T05:01:49 | {} | MEMBER |
ClickHouse | ClickHouse | 523,314,391 | 6,562 | MetikovVadim | Thank you for your answer. I will try it soon. | 2019-08-21T06:20:32 | 2019-08-21T06:20:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,403,998 | 5,957 | abyss7 | > @abyss7 is it still relevant?
On the first glance it's not Kafka-related, at least the first case.
Will take a look into the second case. | 2019-08-21T10:48:01 | 2019-08-21T10:48:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,012,822 | 6,549 | thyn | not fixed by
`
setfacl -d -m u::rwx /data/clickhouse/data/actions/actions
`
| 2019-08-20T13:23:52 | 2019-08-20T13:34:42 | {} | NONE |
ClickHouse | ClickHouse | 523,034,243 | 6,572 | 4ertus2 | duplicate #6093 | 2019-08-20T14:13:28 | 2019-08-20T14:13:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,033,064 | 6,572 | den-crane | >Timeout exceeded while receiving data from server. Waited for 300 seconds, timeout is 300 seconds.
>Cancelling query.
It means nothing. Nothing was canceled.
Optimize was working until the successful end.
`optimize` and `alter` are uninterruptible due timeout. It means that client did not receive any messages... | 2019-08-20T14:10:40 | 2019-08-20T14:16:00 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,049,308 | 4,689 | rnz | With
```
# apt-cache policy clickhouse-server
clickhouse-server:
Installed: 19.13.2.19
Candidate: 19.13.2.19
Version table:
```
Error is gone. But dbeaver is unusable (errors on many actions) | 2019-08-20T14:47:39 | 2019-08-20T14:58:38 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 523,141,947 | 6,514 | alexey-milovidov | https://github.com/llvm-mirror/compiler-rt/blob/master/lib/sanitizer_common/sanitizer_deadlock_detector.h#L123
```
LockWithContext all_locks_with_contexts_[64];
``` | 2019-08-20T18:35:04 | 2019-08-20T18:35:04 | {} | MEMBER |
ClickHouse | ClickHouse | 523,250,597 | 6,422 | alexey-milovidov | @abyss7 Waiting for your review. | 2019-08-21T00:51:51 | 2019-08-21T00:51:51 | {} | MEMBER |
ClickHouse | ClickHouse | 523,252,472 | 6,264 | alexey-milovidov | Conflicts are fixed in #6579 | 2019-08-21T01:02:07 | 2019-08-21T01:02:07 | {} | MEMBER |
ClickHouse | ClickHouse | 523,283,957 | 6,461 | alexey-milovidov | @4ertus2 You can continue inserting data into Set engine with subsequent queries until the server will hit OOM. | 2019-08-21T03:49:24 | 2019-08-21T03:49:24 | {} | MEMBER |
ClickHouse | ClickHouse | 523,293,359 | 5,937 | alexey-milovidov | Is it fixed by patching simdjson? https://github.com/lemire/simdjson/pull/276 | 2019-08-21T04:44:33 | 2019-08-21T04:44:33 | {} | MEMBER |
ClickHouse | ClickHouse | 523,000,647 | 6,552 | rafaeldtinoco | https://github.com/yandex/ClickHouse/pull/6569 | 2019-08-20T12:54:09 | 2019-08-20T12:54:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,001,745 | 215 | philodavies | Hi @filimonov, here's a test schema I am using
```CREATE TABLE IF NOT EXISTS test
(
date Date,
datetime DateTime,
unit Enum8('TRUE' = 0, 'FALSE' = 1),
value UInt64
)
ENGINE = MergeTree(date, (date, unit), 8192);
```
I can insert with `insert into test VALUES (18128, 1566305279, 'TRUE', 76854... | 2019-08-20T12:56:59 | 2019-08-20T12:59:23 | {} | NONE |
ClickHouse | ClickHouse | 523,021,594 | 6,506 | den-crane | ```
# cat /etc/clickhouse-server/conf.d/config_substitutes_no_leader.xml
<?xml version="1.0"?>
<yandex>
<merge_tree>
<replicated_can_become_leader>0</replicated_can_become_leader>
</merge_tree>
</yandex>
```
It means this node will not became the leader -- replication / merges coordinator. | 2019-08-20T13:44:25 | 2019-08-20T13:44:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,184,372 | 6,574 | alexey-milovidov | It has some security implications.
If the user has access to the `url` function, it can do server-side requests.
For example, put data to memcached in your company network.
To avoid this issue you have to set `readonly` for malicious users - it will disallow usage of table functions.
Further option to prevent... | 2019-08-20T20:36:43 | 2019-08-20T20:37:44 | {} | MEMBER |
ClickHouse | ClickHouse | 523,281,410 | 6,553 | alexey-milovidov | > And I'm not sure what difference it makes for big files, so maybe we just have to leave it always on.
It will make the order of data non-deterministic.
The task №2 for @nikitamikhaylov is to make an option for order-preserving parallel parsing of data formats. | 2019-08-21T03:34:39 | 2019-08-21T03:34:39 | {} | MEMBER |
ClickHouse | ClickHouse | 523,284,811 | 6,391 | alexey-milovidov | Fixed by @yurriy | 2019-08-21T03:54:11 | 2019-08-21T03:54:11 | {} | MEMBER |
ClickHouse | ClickHouse | 523,285,866 | 6,359 | alexey-milovidov | > Can I set ClickHouse to store latest one month data in the main memory?
Right now it isn't possible out of the box.
Some ugly solutions exist. You can move some data parts to tmps and symlink them from their usual place. Also you can force some data files to be cached in memory using `vmtouch` tool. Nevertheles... | 2019-08-21T04:00:39 | 2019-08-21T04:00:39 | {} | MEMBER |
ClickHouse | ClickHouse | 523,290,498 | 6,098 | alexey-milovidov | @ramun25 Does the issue reproduce in 19.11 or 19.13? | 2019-08-21T04:28:51 | 2019-08-21T04:28:51 | {} | MEMBER |
ClickHouse | ClickHouse | 523,291,970 | 5,969 | alexey-milovidov | Looks like something is wrong in
`(tuple) IN (subquery)` when tuple contains columns from PK.
This issue is most likely not fixed yet... But it will help if you will reproduce it in version 19.13 and post a stack trace here.
Also please install `clickhouse-common-static-dbg` package before reproducing. | 2019-08-21T04:36:55 | 2019-08-21T04:36:55 | {} | MEMBER |
ClickHouse | ClickHouse | 523,321,550 | 5,438 | clemrodriguez | Sorry for the delay. To be honest I wouldn't mind to add integration test. But I don't know how to make them.
Modified code is executed by 'system reload dictionaries' or any call on a dictionary.
| 2019-08-21T06:48:14 | 2019-08-21T06:48:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,350,670 | 6,473 | dvska | Александр,
Yeah, it was my mistake -- an extra trailing line feed in requests. Thanks a lot. | 2019-08-21T08:18:50 | 2019-08-21T08:18:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,448,813 | 6,588 | stavrolia | What kind of table engine do you use? | 2019-08-21T13:08:00 | 2019-08-21T13:09:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,164,551 | 215 | alexey-milovidov | Numeric form should be accepted. | 2019-08-20T19:39:33 | 2019-08-20T19:39:33 | {} | MEMBER |
ClickHouse | ClickHouse | 523,198,012 | 6,552 | rafaeldtinoco | GCC 9 issue is being tracked here:
https://github.com/yandex/ClickHouse/issues/6560
Closing this since the patch was merged.
Thx a lot for considering the fix. | 2019-08-20T21:16:01 | 2019-08-20T21:16:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,380,178 | 6,504 | jetliu1711 | > Client library doesn't support that particular checksums?
> https://github.com/apache/hawq/blob/4156a8760787894ef351de73df3c38287bbba41d/depends/libhdfs3/src/client/RemoteBlockReader.cpp#L148
At first ,I use hdfs table engine to linset data to clickhouse.when I query in clickhouse make error.I think hawq's libhdf... | 2019-08-21T09:37:06 | 2019-08-21T09:37:06 | {} | NONE |
ClickHouse | ClickHouse | 523,469,329 | 6,595 | 4ertus2 | Nothing strange so far. What's the CPU utilisation?
In example, you run 1 query, it uses 100% for 1 second and finishes. Then you run 10 queries simultaneously, they run for 10 seconds and finish. What percentiles would we have? Any.
If the queries have locks they would run one by one and would have 1 sec in 99%. I... | 2019-08-21T13:58:31 | 2019-08-21T14:00:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,009,460 | 6,506 | den-crane | Also https://github.com/yandex/ClickHouse/issues/2642#issuecomment-460037222 | 2019-08-20T13:16:00 | 2019-08-20T13:16:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,031,884 | 6,506 | den-crane | >Does it mean nobody will read data for replicas from this server?
No, it does not.
It's impossible to implement on 100% with replicated tables and internal_replication. Only with replication using Distributed (internal_replication=false). | 2019-08-20T14:08:07 | 2019-08-20T14:08:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,285,257 | 6,359 | alexey-milovidov | @sundy-li Looks like a very interesting feature! I appreciate if you'll make a PR. | 2019-08-21T03:56:58 | 2019-08-21T03:56:58 | {} | MEMBER |
ClickHouse | ClickHouse | 523,139,937 | 6,514 | alexey-milovidov | ```
FATAL: ThreadSanitizer CHECK failed: /build/llvm-toolchain-8-9L7oe3/llvm-toolchain-8-8/projects/compiler-rt/lib/sanitizer_common/sanitizer_deadlock_detector.h:69 "((n_all_locks_)) < (((sizeof(all_locks_with_contexts_)/sizeof((all_locks_with_contexts_)[0]))))" (0x40, 0x40)
#0 __tsan::TsanCheckFailed(char const... | 2019-08-20T18:29:55 | 2019-08-20T18:29:55 | {} | MEMBER |
ClickHouse | ClickHouse | 523,144,734 | 6,549 | filimonov | Maybe restart itself (w/o chmod) also help? | 2019-08-20T18:42:24 | 2019-08-20T18:42:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,169,752 | 6,554 | alexey-milovidov | db55780d21948c2c262e726310de9d6e8553f7aa
c1de51dc5bcb44024953efe7b6e78b1f424dfca2 | 2019-08-20T19:54:42 | 2019-08-20T19:54:42 | {} | MEMBER |
ClickHouse | ClickHouse | 523,292,225 | 5,957 | alexey-milovidov | @abyss7 is it still relevant? | 2019-08-21T04:38:21 | 2019-08-21T04:38:21 | {} | MEMBER |
ClickHouse | ClickHouse | 523,294,559 | 5,891 | alexey-milovidov | @dimarub2000 Need to investigate this issue (how the server <-> client protocol works and is it possible to fix this issue without much effort). | 2019-08-21T04:50:56 | 2019-08-21T04:50:56 | {} | MEMBER |
ClickHouse | ClickHouse | 523,355,738 | 1,102 | tetafro | Waiting for `REPLACE`. The usecase is switching application to another table by re-creating view like this
```
CREATE VIEW my_table AS SELECT * FROM my_table_v1
...
REPLACE TABLE my_table WITH
CREATE VIEW AS SELECT * FROM my_table_v2
```
and application (which queries `my_table`) seamlessly switches from one tab... | 2019-08-21T08:32:17 | 2019-08-21T08:32:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,045,293 | 4,689 | rnz | Same error with
```
DBeaver version 6.1.5.201908191349
```
```
# lsb_release -a
No LSB modules are available.
Distributor ID: Debian
Description: Debian GNU/Linux 10 (buster)
Release: 10
Codename: buster
# apt-cache policy clickhouse-server
clickhouse-server:
Installed: 18.16.1+ds-4
Candidate: 18... | 2019-08-20T14:38:46 | 2019-08-20T14:41:15 | {} | NONE |
ClickHouse | ClickHouse | 523,121,595 | 5,618 | alexey-milovidov | > Because final aggregation result should fit into memory.
So for group by SOME_HIGH_CARDINALITY_SET_OF_COLUMNS, ch will do intermediate partial aggregations on a disk, but in the end CH will gather the result into one part in memory.
But that's just untrue. | 2019-08-20T17:44:20 | 2019-08-20T17:44:20 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 523,288,286 | 6,308 | alexey-milovidov | The issue has been fixed by @abyss7 in version 19.11.7. | 2019-08-21T04:16:02 | 2019-08-21T04:16:02 | {} | MEMBER |
ClickHouse | ClickHouse | 523,430,176 | 6,414 | abyss7 | The reason is about new processors code - the `ReadSuffix()` method is now called less frequently than it was, causing Kafka commits to occur only when a message stream is "dried up". | 2019-08-21T12:15:52 | 2019-08-21T12:15:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 522,999,929 | 6,549 | thyn | I've come back to problem. Re test everything and chmod -R 755 fixed the problem. Next mutation is ok, then I've run another mutation it started to show error again.
1. stop clickhouse-server
2. chmod -R 755
3. start clickhouse-server
4. run one or two mutations -ok
5. error
6. goto 1
| 2019-08-20T12:52:28 | 2019-08-20T13:03:25 | {} | NONE |
ClickHouse | ClickHouse | 523,006,647 | 6,562 | 4ertus2 | > We write identical data in both tables (reversed metric Path to second table).
Could you check the time of 'insert-select + optimize final' for tables t1(Path String) and t2 (reversedPath String) with your data? Just to be sure it's not the reason of difference. | 2019-08-20T13:09:11 | 2019-08-20T13:11:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,239,017 | 6,518 | alexey-milovidov | This is definitely not the right fix for the issue.
I will prepare different fix in a day. | 2019-08-20T23:51:54 | 2019-08-20T23:51:54 | {} | MEMBER |
ClickHouse | ClickHouse | 523,251,568 | 6,343 | alexey-milovidov | Please prepare for merge. | 2019-08-21T00:57:26 | 2019-08-21T00:57:26 | {} | MEMBER |
ClickHouse | ClickHouse | 523,257,052 | 6,563 | yingfeng | Corruptions for parametric aggregation functions:
```
SELECT id, retention(pickup_datetime = '2018-01-01', pickup_datetime = '2018-01-02', pickup_datetime = '2018-01-03') AS r FROM trips;
```
Where the `trips` table has very few records:
```
SELECT * FROM trips
┌─id─┬─cab_type_id─┬─────pickup_datetime─┐
│ 1 ... | 2019-08-21T01:26:21 | 2019-08-21T01:26:54 | {} | NONE |
ClickHouse | ClickHouse | 523,275,085 | 6,563 | alexey-milovidov | I cannot reproduce it in Docker:
https://gist.github.com/alexey-milovidov/4ddbe9efe9e93041042ae9aff32acfc5 | 2019-08-21T02:59:42 | 2019-08-21T02:59:42 | {} | MEMBER |
ClickHouse | ClickHouse | 523,313,535 | 6,572 | filimonov | Proper way would be just to skip both timeout / and cancellation logic in case of non-breakable queries.
But actually the main problem here is lack of transparency / misleading messages. So logic can even stay the same, but at least messages should give better feeling of what happens inside.
Closing (let's cont... | 2019-08-21T06:17:01 | 2019-08-21T06:18:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,366,158 | 6,343 | dimarub2000 | > Please prepare for merge.
Added a T_test.h file like in performance-test PR, and made options --host --port similar to performance-test PR, think it is ready now.
UPD: Not sure if i should fix style of a huge table (Style check — Found 101 errors) | 2019-08-21T08:59:40 | 2019-08-21T12:30:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,474,323 | 6,595 | gasabr | Thank you for such a fast response, got it now! If that is expected behavior, could you, please, explain, why does this query consumes so much resources? Or is there probably a way to rewrite it to be simpler? | 2019-08-21T14:10:26 | 2019-08-21T14:10:26 | {} | NONE |
ClickHouse | ClickHouse | 523,028,111 | 6,571 | maqroll | 19.9.5.36 | 2019-08-20T13:59:26 | 2019-08-20T13:59:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,105,414 | 6,575 | 4ertus2 | It's an parser error on remote node: it is parsing query from text representation and find something wrong with it. At this stage any function would be just a name.
Most probably it's not related with bitmap fucntions but with text representation of ```gender='男'```.
Would be the query OK if you change it to ```gende... | 2019-08-20T17:02:56 | 2019-08-20T17:08:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,289,170 | 6,238 | alexey-milovidov | > 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?
It allows to cache blocks of data in decompressed form to avoid the cost of decompressing data and reading data from disk for repeating queries. Uncompressed cache is l... | 2019-08-21T04:21:15 | 2019-08-21T04:21:51 | {} | MEMBER |
ClickHouse | ClickHouse | 523,292,879 | 5,968 | alexey-milovidov | Please convert addresses from stack trace to file:line.
Example:
```
addr2line -afiCe ./clickhouse-server 0xfa01384 0x9eb8ef3
```
And you will see a place where unhandled exception was thrown. | 2019-08-21T04:41:48 | 2019-08-21T04:42:08 | {} | MEMBER |
ClickHouse | ClickHouse | 490,612,516 | 5,222 | Crazylearner30 | @den-crane i havent changed anything since the last load and it was working fine
below is the table schema
CREATE TABLE in.staging_contact (column 1, column 2,.......) ENGINE = MergeTree PARTITION BY category ORDER BY (topic, domain) SETTINGS index_granularity = 8192
above is the show create table TABLE
... | 2019-05-08T19:08:27 | 2019-08-21T09:25:30 | {} | NONE |
ClickHouse | ClickHouse | 523,480,833 | 6,595 | den-crane | Please show a line with statistics from clickhouse-client
>PARTITION BY (toYYYYMM(created), group_id) ORDER BY (group_id, idxid);
>WHERE idxid == '...' GROUP BY idxid;
How many `group_id` do you have? CH scans many partitions in this case.
try PARTITION BY (toYYYYMM(created)) ORDER BY (idxid, group_id)
Or ... | 2019-08-21T14:25:33 | 2019-08-21T14:25:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,000,748 | 6,552 | rafaeldtinoco | Pull request made! Thx for reviewing this. | 2019-08-20T12:54:22 | 2019-08-20T12:54:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,121,018 | 6,576 | alexey-milovidov | GROUP BY result doesn't have to fit in memory, look at https://clickhouse.yandex/docs/en/query_language/select/#select-group-by-in-external-memory | 2019-08-20T17:42:51 | 2019-08-20T17:42:51 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 523,135,586 | 6,514 | alexey-milovidov | The only thing left is internal assert in thread sanitizer. | 2019-08-20T18:18:59 | 2019-08-20T18:18:59 | {} | MEMBER |
ClickHouse | ClickHouse | 523,137,792 | 6,577 | alexey-milovidov | The description is not descriptive enough. | 2019-08-20T18:24:33 | 2019-08-20T18:24:33 | {} | MEMBER |
ClickHouse | ClickHouse | 523,142,332 | 6,514 | alexey-milovidov | https://github.com/google/sanitizers/issues/950 | 2019-08-20T18:36:09 | 2019-08-20T18:36:09 | {} | MEMBER |
ClickHouse | ClickHouse | 523,247,798 | 6,429 | alexey-milovidov | (incomplete) | 2019-08-21T00:35:49 | 2019-08-21T00:35:49 | {} | MEMBER |
ClickHouse | ClickHouse | 523,261,250 | 6,518 | alexey-milovidov | #6580 | 2019-08-21T01:47:36 | 2019-08-21T01:47:36 | {} | MEMBER |
ClickHouse | ClickHouse | 523,287,819 | 6,341 | alexey-milovidov | > Are some events removed in 19.6.2?
No, `MergedRows` metric is in place.
Example from version 19.11.3.11:
```
:) SELECT * FROM system.events WHERE event = 'MergedRows'
SELECT *
FROM system.events
WHERE event = 'MergedRows'
┌─event──────┬───────value─┬─description────────────────────────────────────────... | 2019-08-21T04:13:05 | 2019-08-21T04:13:05 | {} | MEMBER |
ClickHouse | ClickHouse | 523,294,249 | 5,866 | alexey-milovidov | #5695 | 2019-08-21T04:49:13 | 2019-08-21T04:49:13 | {} | MEMBER |
ClickHouse | ClickHouse | 523,323,607 | 6,549 | thyn | No, restart without chmod doesn't help. Just tested
only
`
systemctl stop clickhouse-server && chmod -R 755 /data/clickhouse/data && systemctl start clickhouse-server && sleep 10 && clickhouse-client
`
Also there is no any notable problem with files on other operations
| 2019-08-21T06:55:21 | 2019-08-21T06:57:33 | {} | NONE |
ClickHouse | ClickHouse | 523,261,161 | 6,580 | alexey-milovidov | CC @YiuRULE | 2019-08-21T01:47:08 | 2019-08-21T01:47:08 | {} | MEMBER |
ClickHouse | ClickHouse | 523,450,253 | 6,588 | 4ertus2 | There's a RENAME query: https://clickhouse.yandex/docs/en/query_language/misc/#misc_operations-rename
It recuires that your source and destination tables are on the same device (to be able move data by hard links). Could you print ```show create metadata.income``` and your rename query to check if you made an error. | 2019-08-21T13:11:51 | 2019-08-21T13:11:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,004,866 | 6,569 | rafaeldtinoco | Fixes: https://github.com/yandex/ClickHouse/issues/6552 | 2019-08-20T13:04:42 | 2019-08-20T13:04:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,019,991 | 6,554 | alexey-milovidov | NumbersMultiThreadedBlockInputStream? | 2019-08-20T13:40:39 | 2019-08-20T13:40:39 | {} | MEMBER |
ClickHouse | ClickHouse | 523,052,219 | 6,571 | 4ertus2 | FYI @CurtizJ | 2019-08-20T14:54:01 | 2019-08-20T14:54:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,150,228 | 6,569 | rafaeldtinoco | Failed test is being skipped in our builds...
I got another issue (seems irrelevant) in our builds:
```
- result differs with reference:
11: --- /<<BUILDDIR>>/clickhouse-18.16.1+ds/dbms/tests/queries/0_stateless/00536_int_exp.reference 2018-12-20 16:38:50.000000000 +0000
11: +++ /tmp/clickhouse.test..aS0HS/tmp... | 2019-08-20T18:57:35 | 2019-08-20T18:57:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,284,995 | 6,381 | alexey-milovidov | @dimarub2000 Please try to reproduce this issue. | 2019-08-21T03:55:14 | 2019-08-21T03:55:14 | {} | MEMBER |
ClickHouse | ClickHouse | 523,294,739 | 5,844 | alexey-milovidov | And it was optimized by @4ertus2 | 2019-08-21T04:51:59 | 2019-08-21T04:51:59 | {} | MEMBER |
ClickHouse | ClickHouse | 523,336,704 | 6,484 | TomatoBoy90 | > @JonLeeCSDN
> `cluster-1` is interpreted as arithmetic expression. You have to put in in backticks:
>
> ```
> `cluster-1`
> ```
>
> or use different cluster name.
thank you ,in fact ,code as follow will be true:
`node03 :) CREATE TABLE ontime_all AS ontime_local_2 ENGINE = Distributed('cluster-1', 'h2', ... | 2019-08-21T07:38:06 | 2019-08-21T07:38:06 | {} | NONE |
ClickHouse | ClickHouse | 523,017,702 | 6,506 | astudnev | I did not find any documentation on how to set replicated_can_become_leader and what it does | 2019-08-20T13:35:18 | 2019-08-20T13:35:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,157,914 | 6,569 | alexey-milovidov | See comments in `preciseExp10.h`:
```
/** exp10 from GNU libm fails to give precise result for integer arguments.
* For example, exp10(3) gives 1000.0000000000001
* despite the fact that 1000 is exactly representable in double and float.
* Better to always use implementation from MUSL.
*
* Note: the... | 2019-08-20T19:19:48 | 2019-08-20T19:19:48 | {} | MEMBER |
ClickHouse | ClickHouse | 523,284,415 | 6,480 | alexey-milovidov | Could you please run clickhouse-server with gdb:
```
gdb --args clickhouse-server --config-file=/etc/clickhouse-server/config.xml
```
In gdb, type `r`. After stopped by signal, type `t apply all bt`. | 2019-08-21T03:51:59 | 2019-08-21T03:51:59 | {} | MEMBER |
ClickHouse | ClickHouse | 523,312,411 | 5,968 | last2win | one mounth passed, and i cannot find this version of clickhouse. | 2019-08-21T06:12:36 | 2019-08-21T06:12:36 | {} | NONE |
ClickHouse | ClickHouse | 523,386,293 | 6,451 | kaijianding | @CurtizJ I think the new ttl behavior is still not good enough and is still confusing. It's better to purge out-dated data automatically. Maybe a new background scanner thread to scan parts to do the purge periodically. A part should have statistics about min/max value of a Date/Datetime column, so the scanner should n... | 2019-08-21T09:53:43 | 2019-08-21T09:53:43 | {} | NONE |
ClickHouse | ClickHouse | 523,387,494 | 974 | sc750531323 | how to fix this issue ?
+1 | 2019-08-21T09:57:02 | 2019-08-21T09:57:02 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 523,295,681 | 5,925 | alexey-milovidov | @geldot That's right! I can fix it after merge. | 2019-08-21T04:56:56 | 2019-08-21T04:56:56 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 523,005,589 | 6,565 | den-crane | I believe a cause is CREATE MATERIALIZED VIEW.... **SETTINGS index_granularity=8192**
Try index_granularity=64 ... 16 ... 8
https://gist.github.com/den-crane/32cac4f069f4227b35ab86553bbc649b
Or check adaptive index_granularity in the recent releases. | 2019-08-20T13:06:37 | 2019-08-20T13:06:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,027,223 | 6,571 | den-crane | It's supposed to be fixed a long time ago https://github.com/yandex/ClickHouse/issues/3253
What CH version do you use? | 2019-08-20T13:57:27 | 2019-08-20T13:57:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,078,358 | 6,563 | weiqxu | I also have the same problem in Centos 7.3.
I have some investigation, it crashed in these code.
const ElfW(Word) * hash = reinterpret_cast<const ElfW(Word) *>(correct_address(info->dlpi_addr, it->d_un.d_ptr));
sym_cnt = hash[1];
the code used to compute the address when the t... | 2019-08-20T15:54:16 | 2019-08-20T15:54:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,122,041 | 5,618 | alexey-milovidov | ClickHouse can send aggregation result to client without allocating memory for the whole result.
It will do it by buckets of 1 / 256 of total result size. | 2019-08-20T17:45:27 | 2019-08-20T17:45:27 | {} | MEMBER |
ClickHouse | ClickHouse | 523,260,213 | 6,575 | yuzhichang | `gender='M'` doesn't help. Here's cli output:
Note that `gender='M'` gives an empty bitmap.
```
shoyclpgrepl01.mcd.com.cn :) SELECT toDate(order_complete_time) AS day, sum(order_total_sales) AS amount FROM dist_cdp_orders_partition_by_day WHERE bitmapCotains((SELECT groupBitmapState(mid_seq) FROM cdp_cust_tag_yzc ... | 2019-08-21T01:42:08 | 2019-08-21T01:42:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 523,276,953 | 6,582 | alexey-milovidov | https://clickhouse-test-reports.s3.yandex.net/0/a691bf3c9195419002364853feac011c2b05c926/functional_stateless_tests_(address).html | 2019-08-21T03:10:06 | 2019-08-21T03:10:06 | {} | MEMBER |
ClickHouse | ClickHouse | 523,279,736 | 5,273 | alexey-milovidov | > Have no idea why 00395_nullable fails, locally (under ub sanitizer) works fine
It's our problem, look at https://github.com/yandex/ClickHouse/pull/6580 | 2019-08-21T03:25:33 | 2019-08-21T03:25:33 | {} | MEMBER |
ClickHouse | ClickHouse | 523,285,398 | 6,359 | alexey-milovidov | Multi-volumes (in development) feature will help: https://github.com/yandex/ClickHouse/pull/6489
You will be able to create a volume for fresh data in tmpfs. | 2019-08-21T03:57:49 | 2019-08-21T03:57:49 | {} | MEMBER |
ClickHouse | ClickHouse | 523,285,914 | 5,925 | geldot | The rest of Clickhouse uses US English spelling (materialize, color, center) rather than British English (materialise, colour, centre) so for consistency this function should probably be 'neighbor' without the 'u'
Not my preference, but mixing the two dialects gets confusing.
| 2019-08-21T04:00:50 | 2019-08-21T04:00:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 525,359,967 | 6,602 | amosbird | https://stackoverflow.com/a/51436025 | 2019-08-27T15:40:48 | 2019-08-27T15:40:48 | {} | COLLABORATOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.