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 | 513,333,325 | 6,081 | den-crane | Check disk free space. `r1 │ global_events_with_properties ` merge reserved 2*35.55 Gb | 2019-07-19T18:36:26 | 2019-07-19T18:36:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,585,193 | 6,091 | alexey-milovidov | `Stress test (thread)` - known race condition in zlib-ng (used while compressing log files), irrelevant. | 2019-07-21T20:29:18 | 2019-07-21T20:29:18 | {} | MEMBER |
ClickHouse | ClickHouse | 513,591,760 | 5,942 | stavrolia | Better attempt in PR #6092 | 2019-07-21T22:10:39 | 2019-07-21T22:10:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,733,390 | 6,098 | akuzm | Hi @ramun25, thanks for reporting! Is it a one-time issue or a frequent one? If you can reproduce it reliably, it would be helpful to have a small test case, so that we can reproduce it ourselves and debug the problem.
We recently fixed a similar issue #4402. The fix is in master branch, and is going to be released ... | 2019-07-22T10:20:52 | 2019-07-22T10:20:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,848,411 | 5,695 | alexey-milovidov | Probably it should not because of FINAL modifier?
```
milovidov-Pro-P30 :) SELECT * FROM (SELECT * FROM bill) WHERE storeId = '079912d5-b297-4634-a2b5-c7adb5dec5f4'
SELECT *
FROM
(
SELECT *
FROM bill
)
WHERE storeId = '079912d5-b297-4634-a2b5-c7adb5dec5f4'
┌─storeId─────────────────────────────... | 2019-07-22T16:01:02 | 2019-07-22T16:01:02 | {} | MEMBER |
ClickHouse | ClickHouse | 513,919,170 | 6,059 | BayoNet | Этот пулл решает ту же задачу, что и более ранний https://github.com/yandex/ClickHouse/pull/5989/. Я разрешу конфликты и возьму лучшее из обоих. | 2019-07-22T19:22:43 | 2019-07-22T19:22:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,260,757 | 6,014 | filimonov | Hmmm...
ENABLE_MONGODB is poco build flag (contrib/poco)
ENABLE_POCO_MONGODB is ClickHouse flag.
| 2019-07-19T14:55:07 | 2019-07-19T15:01:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,276,309 | 6,071 | arruzk | @abyss7 thank you!
Maybe a full log will help:
[clickhouse-server.log.txt](https://github.com/yandex/ClickHouse/files/3411740/clickhouse-server.log.txt)
Before restarting clickhouse read ~700 million rows from kafka.
It looks like it was able to start only after finishing of removing old parts.
```
2019.07.19 13:... | 2019-07-19T15:38:30 | 2019-07-19T15:39:56 | {} | NONE |
ClickHouse | ClickHouse | 513,386,777 | 6,043 | alexey-milovidov | Perf test:
```
Cannot write to file /var/lib/clickhouse/data/default/table_MySQLWire/data.MySQLWire,
errno: 28, strerror: No space left on device.
``` | 2019-07-19T21:39:45 | 2019-07-19T21:39:45 | {} | MEMBER |
ClickHouse | ClickHouse | 513,557,864 | 6,080 | alexey-milovidov | All ZooKeeper tests have timed out: https://nda.ya.ru/3VmJmN | 2019-07-21T14:11:01 | 2019-07-21T14:11:01 | {} | MEMBER |
ClickHouse | ClickHouse | 513,737,069 | 3,796 | filimonov | > Note, another reason for getDatabaseName in IStorage
See https://github.com/yandex/ClickHouse/pull/5953 | 2019-07-22T10:33:12 | 2019-07-22T10:33:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,800,999 | 6,098 | Imomoi | Hi there.
Looks like we have similar issue on 19.9.3.31 (official build) under Ubuntu on AWS EC2.
2019.07.22 12:46:00.869158 [ 1560 ] {} <Error> BaseDaemon: ########################################
2019.07.22 12:46:00.869244 [ 1560 ] {} <Error> BaseDaemon: (version 19.9.3.31 (official build)) (from threa... | 2019-07-22T13:54:09 | 2019-07-22T13:54:09 | {} | NONE |
ClickHouse | ClickHouse | 513,980,905 | 6,034 | den-crane | The reason is excessive final aggregation at an initiator.
distributed_group_by_no_merge =1 gives 900ms and correct result https://t.me/clickhouse_en/13176 | 2019-07-22T22:36:29 | 2019-07-22T22:36:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,282,030 | 6,066 | abyss7 | Isn't it the same #6042 ? | 2019-07-19T15:56:03 | 2019-07-19T15:56:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,385,911 | 6,077 | alexey-milovidov | `ClickHouse Stress Test (thread)` - fixed in #6072 | 2019-07-19T21:35:52 | 2019-07-19T21:35:52 | {} | MEMBER |
ClickHouse | ClickHouse | 513,392,211 | 6,081 | abraithwaite | We have 5TB free disk space on the server.
```
$ df -h
Filesystem Size Used Avail Use% Mounted on
overlay 49G 2.5G 45G 6% /
tmpfs 64M 0 64M 0% /dev
tmpfs 47G 0 47G 0% /sys/fs/cgroup
/dev/nvme2n1p1 97G 9.0G 88G 10% /var/log
shm 64M ... | 2019-07-19T21:57:58 | 2019-07-19T21:59:39 | {} | NONE |
ClickHouse | ClickHouse | 513,415,743 | 6,055 | alexey-milovidov | Should we rename `format_csv_unquoted_null_literal_as_null`
to `input_format_csv_unquoted_null_literal_as_null`
to have a chance to introduce a setting
`output_format_csv_unquoted_null_literal_as_null`
to write NULLs as NULL? | 2019-07-20T00:05:12 | 2019-07-20T00:05:12 | {} | MEMBER |
ClickHouse | ClickHouse | 513,551,794 | 6,031 | yurriy | It turned out that though mysqlnd does not support mysql-8 default authentication method, it does support similar authentication method which is used by ClickHouse. I added an [example](https://github.com/yandex/ClickHouse/pull/6091/files#diff-1a5a3073c85d14baec4412ccbdcc04fb) with the same version of PHP (7.3.7) and m... | 2019-07-21T12:51:32 | 2019-07-21T12:51:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,568,661 | 5,757 | wawaka | ```
ClickHouse client version 19.11.3.11 (official build).
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 19.11.3 revision 54423.
clickhouse1.example.com :) OPTIMIZE TABLE sms_service.events_v2_local ON CLUSTER statistics FINAL
OPTIMIZE TABLE sms_service.events_v2_local ON... | 2019-07-21T16:32:22 | 2019-07-21T16:32:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,847,846 | 5,695 | alexey-milovidov | Confirmed. It doesn't work at all:
```
CREATE TABLE bill (storeId String) ENGINE = ReplacingMergeTree ORDER BY storeId;
INSERT INTO bill SELECT generateUUIDv4() FROM numbers(1000000);
SELECT * FROM bill ORDER BY rand() LIMIT 10;
│ ba1534cc-a9cb-46e8-8715-9236fc6189f6 │
...
create view bill_determine_view a... | 2019-07-22T15:56:43 | 2019-07-22T15:56:43 | {} | MEMBER |
ClickHouse | ClickHouse | 513,861,165 | 3,796 | alexey-milovidov | BTW, Why we are not using `InterpreterInsertQuery` while pushing to views?
It will allow both recursive materialized views and do defaults/materialized calculation automatically. | 2019-07-22T16:42:07 | 2019-07-22T16:42:07 | {} | MEMBER |
ClickHouse | ClickHouse | 513,925,608 | 5,695 | alexey-milovidov | CC @zhang2014 | 2019-07-22T19:42:04 | 2019-07-22T19:42:04 | {} | MEMBER |
ClickHouse | ClickHouse | 513,926,690 | 5,695 | alexey-milovidov | It was specified by me:
https://github.com/yandex/ClickHouse/pull/2015#issuecomment-374283452 | 2019-07-22T19:45:12 | 2019-07-22T19:45:12 | {} | MEMBER |
ClickHouse | ClickHouse | 513,994,701 | 5,985 | den-crane | It has different behaviour with different RDMS.
PG does not allow to change to not nullable until column contains NULLs. Probably it possible because PG DDL are transactional and PG able to check and alter in one transaction. So PG makes you to update the column to replace all nulls with some value before alter.
... | 2019-07-22T23:39:03 | 2019-07-22T23:43:50 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,226,207 | 6,069 | den-crane | BTW, wrong file. It should be fixed in XDBCDictionarySource.cpp
https://github.com/temoon/ClickHouse/blob/master/dbms/src/Dictionaries/XDBCDictionarySource.cpp#L131
And obviously it is should be generalized for all dictionaries.
And the condition `where $updated_field >= ` can be omitted for the first call.
| 2019-07-19T13:21:22 | 2019-07-19T13:23:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,305,697 | 6,052 | 4ertus2 | There's also a test that should check insertion in distributed (including distributed over replicated) and it OK with #6043
https://github.com/yandex/ClickHouse/blob/master/dbms/tests/integration/test_insert_into_distributed/test.py | 2019-07-19T17:11:55 | 2019-07-19T17:11:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,418,677 | 5,759 | alexey-milovidov | The test is wrong and the code doesn't build. | 2019-07-20T00:31:46 | 2019-07-20T00:31:46 | {} | MEMBER |
ClickHouse | ClickHouse | 513,464,872 | 5,975 | alexey-milovidov | But I strongly recommend to upgrade to 19.11 instead. | 2019-07-20T12:49:17 | 2019-07-20T12:49:17 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 513,545,063 | 6,031 | yurriy | @foolfish3 could you please temporary replace user and password (and other sensitive data) in test.php and provide a result of ``strace -s 1000 -- php -r test.php``? | 2019-07-21T11:20:09 | 2019-07-21T11:20:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,587,635 | 6,093 | alexey-milovidov | BTW, 300 seconds is default socket read timeout. | 2019-07-21T21:07:48 | 2019-07-21T21:07:48 | {} | MEMBER |
ClickHouse | ClickHouse | 513,767,133 | 6,099 | alexey-milovidov | 19.11 is the most stable version in all terms except Kafka integration.
You should evaluate any new version in testing environment before installing in production. | 2019-07-22T12:17:37 | 2019-07-22T12:17:37 | {} | MEMBER |
ClickHouse | ClickHouse | 513,922,558 | 5,950 | alexey-milovidov | Fixed. | 2019-07-22T19:32:50 | 2019-07-22T19:32:50 | {} | MEMBER |
ClickHouse | ClickHouse | 513,925,361 | 5,695 | alexey-milovidov | It was disabled intentionally:
https://github.com/yandex/ClickHouse/blob/master/dbms/src/Interpreters/PredicateExpressionsOptimizer.cpp#L141
We should figure out what wrong can happen if we will enable it... | 2019-07-22T19:41:17 | 2019-07-22T19:41:17 | {} | MEMBER |
ClickHouse | ClickHouse | 513,960,756 | 5,959 | den-crane | pass this parameter to ch-client.
clickhouse-client --max_insert_block_size=1024 | 2019-07-22T21:24:34 | 2019-07-22T21:24:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,410,373 | 6,004 | alexey-milovidov | Right side of JOIN is collected into hash table in memory.
Then while reading from left side of JOIN we'll do lookups in hash table and find rows to be joined. | 2019-07-19T23:25:20 | 2019-07-19T23:26:29 | {} | MEMBER |
ClickHouse | ClickHouse | 514,018,305 | 6,099 | ghost | @alexey-milovidov
Thank you for your advice! :)
| 2019-07-23T01:37:51 | 2019-07-23T01:37:51 | {} | NONE |
ClickHouse | ClickHouse | 513,229,888 | 6,069 | alexey-milovidov | Let's change it to `1971-01-01 00:00:00`. | 2019-07-19T13:32:32 | 2019-07-19T13:32:32 | {} | MEMBER |
ClickHouse | ClickHouse | 513,257,258 | 6,073 | alexey-milovidov | PS. Probably it may create excessive CPU load (number of threads waiting for CPU) in presence of many databases? | 2019-07-19T14:45:49 | 2019-07-19T14:46:43 | {} | MEMBER |
ClickHouse | ClickHouse | 513,259,751 | 6,074 | alexey-milovidov | Time is spent for filesystem operations (stat), reading small files (columns.txt, checksums.txt) and reading index (primary.idx) that is also small. It usually isn't bounded by CPU neither by full disk utilization, but bounded by latency of filesystem operations, and performance can be significantly improved by paralle... | 2019-07-19T14:52:25 | 2019-07-19T14:52:25 | {} | MEMBER |
ClickHouse | ClickHouse | 513,360,753 | 6,042 | florida-hehe | > @florida-hehe can you provide your Kafka server configuration regarding the auth settings?
`<kafka>`
`<debug>all</debug>`
`<auto_offset_reset>latest</auto_offset_reset>`
`<security_protocol>sasl_plaintext</security_protocol>`
`<sasl_mechanism>PLAIN</sasl_mechanism>`
`<compression_type>snap... | 2019-07-19T20:06:12 | 2019-07-19T20:06:12 | {} | NONE |
ClickHouse | ClickHouse | 513,529,145 | 5,759 | Felixoid | Indeed, haven't adjusted it after the latest tiny change. | 2019-07-21T07:01:26 | 2019-07-21T07:01:26 | {} | MEMBER |
ClickHouse | ClickHouse | 513,620,887 | 6,063 | zhang2014 | I'll try to fix it. you can currently drop the database using the following command:
```
clickhouse :) DETACH DATABASE {need drop database name}
clickhouse :) exit
~ cd {clickhouse data path}
~ rm -rf metadata/{need drop database name}
```
| 2019-07-22T02:51:18 | 2019-07-22T03:38:30 | {
"+1": 6
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,925,508 | 5,695 | alexey-milovidov | I think that nothing wrong can happen, because FINAL is evaluated before WHERE. | 2019-07-22T19:41:46 | 2019-07-22T19:41:46 | {} | MEMBER |
ClickHouse | ClickHouse | 513,267,612 | 6,071 | abyss7 | It's still not good - will investigate. | 2019-07-19T15:13:55 | 2019-07-19T15:13:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,417,112 | 6,026 | alexey-milovidov | @abyss7 Tomorrow? | 2019-07-20T00:17:02 | 2019-07-20T00:17:02 | {} | MEMBER |
ClickHouse | ClickHouse | 513,464,820 | 5,975 | alexey-milovidov | Done: https://github.com/yandex/ClickHouse/pull/5626 | 2019-07-20T12:48:40 | 2019-07-20T12:48:40 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 513,813,075 | 6,098 | Imomoi | @akuzm - registered #6105 | 2019-07-22T14:23:00 | 2019-07-22T14:23:00 | {} | NONE |
ClickHouse | ClickHouse | 513,833,298 | 6,105 | alexey-milovidov | Does it reproduce on version 19.11? | 2019-07-22T15:10:33 | 2019-07-22T15:10:33 | {} | MEMBER |
ClickHouse | ClickHouse | 513,917,143 | 5,815 | alexey-milovidov | Yum tells me that one of the packages is not signed:
```
[root@1692dc7fa4cd /]# yum install clickhouse-server clickhouse-client
Loaded plugins: product-id, subscription-manager ... | 2019-07-22T19:16:07 | 2019-07-22T19:16:07 | {} | MEMBER |
ClickHouse | ClickHouse | 513,226,076 | 4,586 | tishkinis | Code: 48, e.displayText() = DB::Exception: DatabaseDictionary: removeTable() is not supported (version 19.9.2.4) | 2019-07-19T13:21:02 | 2019-07-19T13:21:02 | {} | NONE |
ClickHouse | ClickHouse | 513,234,286 | 6,071 | arruzk | This issue was solved by itself after 30 minutes of restart loop. | 2019-07-19T13:45:22 | 2019-07-19T13:45:22 | {} | NONE |
ClickHouse | ClickHouse | 513,253,089 | 6,072 | alexey-milovidov | Ok.
BTW. Do you have PR to upstream (to follow)? | 2019-07-19T14:34:36 | 2019-07-19T14:34:36 | {} | MEMBER |
ClickHouse | ClickHouse | 513,278,742 | 5,999 | MahmoudGoda0 | Thank you very much for your kind support. | 2019-07-19T15:46:02 | 2019-07-19T15:46:02 | {} | NONE |
ClickHouse | ClickHouse | 513,585,533 | 6,075 | alexey-milovidov | Segfault in stress test (nullptr dereference):
```
2019.07.21 02:11:19.245828 [ 358 ] {3a0c2198-33f6-4403-a373-60895003c6b0} <Debug> executeQuery: (from [::1]:47076) SELECT * FROM merge('', defaultValueOfArgumentType('\0'))
2019.07.21 02:11:19.247688 [ 359 ] {000c3d04-b785-47ad-9bf8-4ea45d89e918} <Trace> Interpret... | 2019-07-21T20:34:36 | 2019-07-21T20:34:36 | {} | MEMBER |
ClickHouse | ClickHouse | 513,661,966 | 5,959 | renkun2020 | > > [root@master tmp]# du -sh passed_data.csv
> > 2.1G passed_data.csv
> > [root@master tmp]# wc -l passed_data.csv
> > 36953 passed_data.csv
>
> That explains everything. You have avg size of the row about 57Kb. It's very big rows.
>
> ClickHouse tries to form blocks of max_insert_block_size (by default 10485... | 2019-07-22T06:43:55 | 2019-07-22T06:43:55 | {} | NONE |
ClickHouse | ClickHouse | 513,309,714 | 6,042 | abyss7 | My bad, haven't looked at the version. There are no plans to backport Kafka+SASL support to 19.7 | 2019-07-19T17:24:37 | 2019-07-19T17:24:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,658,554 | 6,052 | millin | I checked my cluster configuration and found that `internal_replication` is also enabled.
This error occurs only with Replicated engine with more than one shard. | 2019-07-22T06:30:11 | 2019-07-22T06:30:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,696,324 | 6,096 | Rostish | I Have the same problem:
It happens random. I do batch calculations with day granularity, 2 from 4 dates have ended correctly, but others not.
> 2019.07.22 07:35:51.014610 [ 32 ] {} <Error> void DB::StorageKafka::streamThread(): std::exception. Code: 1001, type: cppkafka::HandleException, e.what() = Local: Timed ou... | 2019-07-22T08:34:53 | 2019-07-22T08:38:46 | {} | NONE |
ClickHouse | ClickHouse | 513,765,451 | 4,402 | alexey-milovidov | > ASan report inside libunwind is normal (by design).
We have also fixed ASan reports in libunwind by https://github.com/yandex/ClickHouse/pull/4828 | 2019-07-22T12:11:56 | 2019-07-22T12:12:03 | {} | MEMBER |
ClickHouse | ClickHouse | 513,826,384 | 6,101 | abyss7 | > Do I understand correctly that it will dynamically load .so from the system?
> We should avoid it for production build.
It's used to load plugins specified in config. We don't use it. | 2019-07-22T14:53:55 | 2019-07-22T14:53:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,922,718 | 5,950 | alexey-milovidov | Upstream PR is up to you. | 2019-07-22T19:33:20 | 2019-07-22T19:33:20 | {} | MEMBER |
ClickHouse | ClickHouse | 513,926,890 | 5,695 | alexey-milovidov | We can simply remove restriction on FINAL. | 2019-07-22T19:45:46 | 2019-07-22T19:45:46 | {} | MEMBER |
ClickHouse | ClickHouse | 514,000,413 | 4,247 | alexey-milovidov | TODO:
- change FunctionsIntrospection;
- enrich trace_log;
- check that profiler is set up for every query thread; | 2019-07-23T00:07:05 | 2019-07-23T00:07:05 | {} | MEMBER |
ClickHouse | ClickHouse | 513,307,693 | 6,080 | abyss7 | Should fix #6071 | 2019-07-19T17:18:23 | 2019-07-19T17:18:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,385,557 | 6,075 | alexey-milovidov | TODO: check integration tests. | 2019-07-19T21:34:15 | 2019-07-19T21:34:15 | {} | MEMBER |
ClickHouse | ClickHouse | 513,409,683 | 6,031 | alexey-milovidov | CC @yurriy | 2019-07-19T23:20:44 | 2019-07-19T23:20:44 | {} | MEMBER |
ClickHouse | ClickHouse | 513,512,122 | 5,759 | alexey-milovidov | But the test doesn't work. | 2019-07-21T01:46:23 | 2019-07-21T01:46:23 | {} | MEMBER |
ClickHouse | ClickHouse | 513,844,427 | 6,103 | nikitamikhaylov | Please add some lines to server config to create system.text_log table. | 2019-07-22T15:36:11 | 2019-07-22T15:36:11 | {} | MEMBER |
ClickHouse | ClickHouse | 513,218,708 | 5,981 | alexey-milovidov | 98b0d08bf3 | 2019-07-19T12:57:38 | 2019-07-19T12:57:38 | {} | MEMBER |
ClickHouse | ClickHouse | 513,294,472 | 6,052 | 4ertus2 | Your case is 'Replicated + Distributed + input_format_defaults_for_omitted_fields'. I've tried combination 'MergeTree + Distributed + input_format_defaults_for_omitted_fields' and it looks working. It also interesting if you have 'internal_replication' set for your cluster configuration.
Could you get more details o... | 2019-07-19T16:34:54 | 2019-07-19T16:36:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,410,002 | 6,021 | alexey-milovidov | This issue is fixed by @zhang2014 in commit https://github.com/yandex/ClickHouse/commit/17864076 that was released in version 19.11. | 2019-07-19T23:23:03 | 2019-07-19T23:23:03 | {} | MEMBER |
ClickHouse | ClickHouse | 513,512,066 | 6,031 | yurriy | @foolfish3 thank you for a bug report, could you please provide a result of ``php -r 'phpinfo();'``?
It is a known issue that few clients which does not support mysql-8 default authentication method cannot connect to ClickHouse. It will be fixed in ClickHouse by implementing older authentication method. As soon as we ... | 2019-07-21T01:45:25 | 2019-07-21T01:52:52 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,557,971 | 5,759 | alexey-milovidov | Wow, it's green :) Thank you! | 2019-07-21T14:12:38 | 2019-07-21T14:12:38 | {} | MEMBER |
ClickHouse | ClickHouse | 513,559,770 | 6,046 | etah000 | It is actually queries related. When tried some other queries, the cpu usage for some nodes reached more then 95%. Any tips for locating the bottleneck for a query?
why the adjustment of "uncompressed_cache_size" has no effect for the query performance? | 2019-07-21T14:35:27 | 2019-07-21T14:35:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,587,892 | 6,055 | tavplubix | > Should we rename `format_csv_unquoted_null_literal_as_null`
> to `input_format_csv_unquoted_null_literal_as_null`
> to have a chance to introduce a setting
> `output_format_csv_unquoted_null_literal_as_null`
> to write NULLs as NULL?
We can use `format_csv_unquoted_null_literal_as_null` for input and output (l... | 2019-07-21T21:11:58 | 2019-07-21T21:11:58 | {} | MEMBER |
ClickHouse | ClickHouse | 513,622,214 | 6,031 | foolfish3 | @yurriy your example still not work, still same error (MySQL server has gone away )
i think --with-pdo-mysql should be --with-pdo-mysql=mysqlnd in ur compile param
=====================================
strace -s 1000 -- php test.php
=====================================
getcwd("/root", 4096) =... | 2019-07-22T02:59:42 | 2019-07-22T02:59:42 | {} | NONE |
ClickHouse | ClickHouse | 513,568,034 | 5,951 | alexey-milovidov | What I have changed:
1. Made parentheses mandatory to allow `table.COLUMNS('regexp')` in future.
2. Made regexp non-anchored to be consistent with `merge` table function.
3. Renamed ColumnsClause to ColumnsMatcher, because it doesn't look similar to SELECT, WHERE, HAVING... clauses.
4. Fixed error in AST formatti... | 2019-07-21T16:23:44 | 2019-07-22T19:38:56 | {} | MEMBER |
ClickHouse | ClickHouse | 513,913,982 | 6,055 | alexey-milovidov | > We can use format_csv_unquoted_null_literal_as_null for input and output (like format_csv_delimiter), but it may be inconvenient if user wants to read nulls as \N or NULL and write nulls as \N.
Let's split. | 2019-07-22T19:06:50 | 2019-07-22T19:06:50 | {} | MEMBER |
ClickHouse | ClickHouse | 514,025,866 | 5,959 | renkun2020 | > pass this parameter to ch-client.
>
> clickhouse-client --max_insert_block_size=1024
It works! Thanks so much!
[root@master tmp]# cat passed_data.csv | clickhouse-client --user=ckuser --password=*** --host=*** --port=9000 --database=*** '--format_csv_delimiter=|' '--query=INSERT INTO matchs FORMAT CSV' ... | 2019-07-23T02:17:34 | 2019-07-23T02:17:34 | {} | NONE |
ClickHouse | ClickHouse | 514,034,844 | 6,031 | yurriy | @foolfish3 I noticed that packets validation became slightly stronger in version of ClickHouse which I used, and I had a different error, but for the same reason. The error occurs because the implementation of authentication plugin in PHP extension doesn't assume that plugin used by ClickHouse will be set as a default ... | 2019-07-23T03:03:36 | 2019-07-23T03:03:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,408,779 | 6,083 | alexey-milovidov | Note: I don't like the code of partial dictionary update, because it is
the classical example of wrong usage of time in distributed systems.
CC @hakobyana
It's impossible to correctly use client side time to filter data by server side time. | 2019-07-19T23:15:17 | 2019-07-19T23:15:17 | {} | MEMBER |
ClickHouse | ClickHouse | 513,415,217 | 6,083 | den-crane | There is no good solution.
We always use the client side time with huge lag -5(15) min.
Because modification time inconsequent in multy-user system with transactions.
1st Transaction begins and updates last_modified = now() = t
2nd Transaction begins and update slast_modified = now() = t + 1
2nd Transaction ... | 2019-07-20T00:01:24 | 2019-07-20T00:03:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,504,591 | 5,759 | Felixoid | @alexey-milovidov I've fixed tests and now both simple and replicated MergeTree behaves identically | 2019-07-20T22:45:32 | 2019-07-20T22:45:32 | {} | MEMBER |
ClickHouse | ClickHouse | 513,809,714 | 6,098 | akuzm | @Imomoi the call stack is completely different, so it's probably another bug. Could you create a separate issue? The test case would be helpful as well, because from the call stack alone it's hard to understand the reason, so we must reproduce it locally. | 2019-07-22T14:15:27 | 2019-07-22T14:15:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,839,657 | 6,026 | alexey-milovidov | Today? | 2019-07-22T15:24:43 | 2019-07-22T15:24:43 | {} | MEMBER |
ClickHouse | ClickHouse | 514,046,255 | 6,031 | foolfish3 | your param --with-pdo-mysql will link mysql lib in /usr
in my case ,you should specify --with-pdo-mysql=mysqlnd , use mysqlnd instead of mysql lib
if you configure with --with-pdo-mysql=mysqlnd, "MySQL server has gone away" will appear
clickhouse server version is 19.9.2.4
################################
$... | 2019-07-23T04:08:44 | 2019-07-23T04:08:44 | {} | NONE |
ClickHouse | ClickHouse | 513,259,000 | 6,073 | akuzm | > PS. Probably it may create excessive CPU load (number of threads waiting for CPU) in presence of many databases?
Databases are loaded sequentially, and the number of threads per database is capped with the number of physical cores (as before), so it's probably OK. | 2019-07-19T14:50:25 | 2019-07-19T14:50:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,301,741 | 6,076 | blinkov | сам патч вроде норм, но тесты сломаны | 2019-07-19T16:58:37 | 2019-07-19T16:58:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,410,948 | 5,985 | alexey-milovidov | Intended behaviour: convert NULL values to default values while ALTERing (?). | 2019-07-19T23:28:57 | 2019-07-19T23:29:16 | {
"+1": 6,
"heart": 1,
"hooray": 1
} | MEMBER |
ClickHouse | ClickHouse | 513,411,725 | 5,984 | alexey-milovidov | Add support for case-insensitive table functions in TableFunctionFactory similar to FunctionFactory, AggregateFunctionFactory.
Implement `VALUES('a UInt64, s String', (1, 'one'), (2, 'two'), (3, 'three'))` as a table function.
Parse first argument as table structure; for the rest arguments - evaluate constant exp... | 2019-07-19T23:34:07 | 2019-07-19T23:34:31 | {
"+1": 4
} | MEMBER |
ClickHouse | ClickHouse | 513,526,174 | 6,031 | foolfish3 | test.php
$pdo= new PDO(......);
php test.php
Fatal error: Uncaught PDOException: PDO::__construct(): MySQL server has gone away in /root/test.php:15
Stack trace:
#0 /root/test.php(15): PDO->__construct('....
tail clickhouse-server.err.log
2019.07.21 15:14:51.654231 [ 133 ] {} <Error> MySQLHandler: DB::Exce... | 2019-07-21T06:15:46 | 2019-07-21T06:15:46 | {} | NONE |
ClickHouse | ClickHouse | 513,814,168 | 6,102 | filimonov | Please start with official tutorial: https://clickhouse.yandex/tutorial.html
Also please read that: https://github.com/yandex/ClickHouse/issues/2161
Also you can watch one of many presentation:
https://www.youtube.com/channel/UChtmrD-dsdpspr42P_PyRAw
https://github.com/yandex/clickhouse-presentations
For... | 2019-07-22T14:25:32 | 2019-07-22T14:25:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,988,554 | 6,108 | alexey-milovidov | `00626_replace_partition_from_table_zookeeper` - flappy test. | 2019-07-22T23:09:22 | 2019-07-22T23:09:22 | {} | MEMBER |
ClickHouse | ClickHouse | 513,996,851 | 6,075 | alexey-milovidov | > Segfault in stress test (nullptr dereference):
Tech debt, will be addressed asap. | 2019-07-22T23:49:28 | 2019-07-22T23:49:28 | {} | MEMBER |
ClickHouse | ClickHouse | 513,997,121 | 6,055 | alexey-milovidov | > 00411_long_accurate_number_comparison
Will be fixed after merge with master. | 2019-07-22T23:50:53 | 2019-07-22T23:50:53 | {} | MEMBER |
ClickHouse | ClickHouse | 513,260,640 | 6,074 | alexey-milovidov | It is also very important for FUSE: S3, MooseFS. | 2019-07-19T14:54:49 | 2019-07-19T14:54:49 | {} | MEMBER |
ClickHouse | ClickHouse | 513,443,770 | 5,545 | k0st1an | Play does not work. Nearby stood two versions of 19.7 and 19.11. 19.11 the problem were found much more often in almost all the topic. After rollback to version 19.7 no problems so far. | 2019-07-20T07:10:18 | 2019-07-20T07:10:18 | {} | NONE |
ClickHouse | ClickHouse | 513,445,583 | 6,076 | BayoNet | Тесты по мёржа мастера пофиксились. | 2019-07-20T07:40:07 | 2019-07-20T07:40:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 513,766,227 | 6,037 | nikitamikhaylov | > server version and revision fields similar to query_log;
Which fields? I see this information only about client. | 2019-07-22T12:14:31 | 2019-07-22T12:14:31 | {} | MEMBER |
ClickHouse | ClickHouse | 503,282,142 | 5,388 | alexey-milovidov | Closed as incomplete/accidential PR. | 2019-06-18T19:41:14 | 2019-06-18T19:41:14 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.