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 | 694,772,419 | 14,859 | 4ertus2 | Fuzzer error is in fetch phase: wget ERROR 404: Not Found. | 2020-09-18T09:51:26 | 2020-09-18T09:51:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,805,787 | 14,684 | alexey-milovidov | Ok. I also worried because DDLWorker is typical pain point - it does not scale for large number of queries for several other reasons. I doubt if parallel execution will solve it. | 2020-09-18T11:08:25 | 2020-09-18T11:08:25 | {} | MEMBER |
ClickHouse | ClickHouse | 694,870,901 | 11,684 | alesapin | performance not affected. | 2020-09-18T13:34:17 | 2020-09-18T13:34:17 | {} | MEMBER |
ClickHouse | ClickHouse | 694,889,702 | 8,889 | aubweb9 | Hello,
I am having similar issues :
```
2020.09.18 15:38:53.096703 [ 7346 ] {} <Fatal> BaseDaemon: (version 20.5.4.40 (official build), build id: 98286EF6C386A9AC) (from thread 5169) (query_id: 830ed3b7-fdcd-4fb7-af38-2a9c50bfe905) Received signal Segmentation fault (11)
2020.09.18 15:38:53.096838 [ 7346 ] {} <... | 2020-09-18T14:07:14 | 2020-09-18T14:07:14 | {} | NONE |
ClickHouse | ClickHouse | 695,033,228 | 14,808 | alexey-milovidov | Wait only for build check. | 2020-09-18T18:56:57 | 2020-09-18T18:56:57 | {} | MEMBER |
ClickHouse | ClickHouse | 694,626,644 | 14,786 | liangli1213 | @BohuTANG
```
bdcdev09.bdcdesign.analog.com :) select version();
SELECT version()
┌─version()────┐
│ 20.10.1.4678 │
└──────────────┘
```
Have you reproduce follow my steps? one important step is, you must setup MySQL Master and Slave first, then ClickHouse sync from Slave.
| 2020-09-18T03:26:50 | 2020-09-18T03:26:50 | {} | NONE |
ClickHouse | ClickHouse | 694,729,846 | 14,962 | akuzm | https://github.com/ClickHouse/ClickHouse/issues/14570
Is it going to fix this issue as well? | 2020-09-18T08:21:16 | 2020-09-18T08:21:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,853,186 | 14,535 | tavplubix | `Yandex third-party checks` - unrelated (broken in master) | 2020-09-18T13:00:40 | 2020-09-18T13:00:40 | {} | MEMBER |
ClickHouse | ClickHouse | 695,033,014 | 14,808 | alexey-milovidov | OOM in build - ignore. | 2020-09-18T18:56:25 | 2020-09-18T18:56:25 | {} | MEMBER |
ClickHouse | ClickHouse | 695,206,705 | 14,976 | alexey-milovidov | Yes, it is integrated into MergeTree tables - it should remove the need for Buffer. And we don't plan to extend Buffer tables. | 2020-09-19T12:22:13 | 2020-09-19T12:22:13 | {} | MEMBER |
ClickHouse | ClickHouse | 695,208,602 | 4,901 | sxpsxp12 | @alexey-milovidov | 2020-09-19T12:42:57 | 2020-09-19T12:42:57 | {} | NONE |
ClickHouse | ClickHouse | 695,216,449 | 13,652 | alexey-milovidov | @excitoon What it happening here? Can we remove `do not test` label? | 2020-09-19T13:55:02 | 2020-09-19T13:55:02 | {} | MEMBER |
ClickHouse | ClickHouse | 694,536,156 | 14,963 | CurtizJ | Currently it's the early stage of implementing #14196. | 2020-09-17T22:36:02 | 2020-09-17T22:36:02 | {} | MEMBER |
ClickHouse | ClickHouse | 694,680,055 | 14,786 | liangli1213 | @BohuTANG , it is my fault, I just set binlog_checksum to NONE in slave.
I have restarted the replication from MySQL to ClickHouse on my test server, so far I didn't see any error.
| 2020-09-18T06:22:41 | 2020-09-18T06:22:41 | {} | NONE |
ClickHouse | ClickHouse | 694,697,991 | 11,116 | alesapin | > Superseded by #12270#11883
Ok, can be closed. | 2020-09-18T07:08:50 | 2020-09-18T07:08:50 | {} | MEMBER |
ClickHouse | ClickHouse | 694,784,700 | 14,924 | akuzm | Ast fuzzer is fixed in master (move to clang-11). | 2020-09-18T10:17:44 | 2020-09-18T10:17:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,841,468 | 14,979 | zhang2014 | @liangli1213 @tavplubix Yes, we have this plan. and that should implement in the near future. automatically convert MySQL's secondary index to ClickHouse's skipping index(index type using MaterializeMySQL database settings, default it is bloom_filter) | 2020-09-18T12:35:25 | 2020-09-18T12:42:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,865,285 | 9,624 | alesapin | `00157_cache_dictionary` -- flaky.
`01081_demangle` with memory sanitizer -- not important.
`test_storage_kafka/test.py::test_kafka_unavailable` -- flap
```
Failed to start cluster:
[Errno 2] No such file or directory: '/ClickHouse/tests/integration/test_storage_kafka/_instances/instance/database/format_schemas/k... | 2020-09-18T13:23:35 | 2020-09-18T13:23:35 | {} | MEMBER |
ClickHouse | ClickHouse | 694,897,400 | 12,171 | alexey-milovidov | @azat
I want it to reuse our formats and their settings. So, format Template can be used for this purpose (and any other format as well).
Maybe we can add some fields like `query_id_color String`, `thread_id_color String`, ... , `color_end String` that the user can use to highlight logs.
Then we have two opti... | 2020-09-18T14:20:53 | 2020-09-18T14:20:53 | {} | MEMBER |
ClickHouse | ClickHouse | 694,897,737 | 12,300 | alexey-milovidov | Discussion: https://github.com/ClickHouse/ClickHouse/pull/12171#issuecomment-688029276 | 2020-09-18T14:21:29 | 2020-09-18T14:21:29 | {} | MEMBER |
ClickHouse | ClickHouse | 694,718,900 | 14,297 | alesapin | Ok, will wait for release tests. | 2020-09-18T07:58:07 | 2020-09-18T07:58:07 | {} | MEMBER |
ClickHouse | ClickHouse | 694,737,636 | 14,201 | amuluowin | > I was thinking about just put the errors (together with unparsed messages) inside a rows generated by Kafka engine.
>
> And later on MV level you can decide what to do with them (drop/store in table/push somewhere etc)
>
> Logfile - not sure if good (they are not accessible if you have only clickhouse shell).
... | 2020-09-18T08:38:12 | 2020-09-18T08:38:12 | {} | NONE |
ClickHouse | ClickHouse | 694,836,766 | 14,976 | nvartolomei | ClickHouse recently (since v20.6.3.28-stable, not enabled by default) added support for in memory parts for MergeTree* engines (these in theory would make it unnecessary to have a Buffer table for batching writes if the target is a MergeTree table) #10697. In-memory parts use a WAL. | 2020-09-18T12:24:04 | 2020-09-18T12:28:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 695,147,303 | 14,979 | liangli1213 | @BohuTANG , thanks for your reply, I shall open a feature request since I really need this conversion. | 2020-09-19T01:43:38 | 2020-09-19T01:43:38 | {} | NONE |
ClickHouse | ClickHouse | 694,455,800 | 14,297 | alesapin | 
:( Fixed in the latest commit. But we have to wait again( | 2020-09-17T19:36:49 | 2020-09-17T19:36:49 | {} | MEMBER |
ClickHouse | ClickHouse | 694,191,531 | 14,927 | den-crane | And how to add or drop / modify a column without re-writing this file?
And I guess it would be much more seek operations when CH reads a single column.
And BTW CH supports compact format parts and it performance is lower than wide format
https://github.com/ClickHouse/ClickHouse/issues/12737 https://github.com/Cl... | 2020-09-17T12:16:30 | 2020-09-17T21:01:55 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,787,865 | 14,953 | tavplubix | Because it has merge conflicts and probably depends on some PR, which was not marked as "pr-bugfix" and was not backported. | 2020-09-18T10:24:28 | 2020-09-18T10:24:28 | {} | MEMBER |
ClickHouse | ClickHouse | 694,831,070 | 14,297 | alesapin | Report OK https://clickhouse-test-reports.s3.yandex.net/14297/3371a62513dfa6abbb065f51502ae86e2930c820/integration_tests_(release).html | 2020-09-18T12:10:00 | 2020-09-18T12:10:00 | {} | MEMBER |
ClickHouse | ClickHouse | 694,946,233 | 14,973 | amosbird | > We already do it, just add new settings here https://github.com/ClickHouse/ClickHouse/blob/master/src/Storages/MergeTree/MergeTreeDataMergerMutator.cpp#L1079-L1081
Thanks! Done. | 2020-09-18T15:51:31 | 2020-09-18T15:51:31 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 695,030,708 | 14,977 | alexey-milovidov | LGTM.
But I forget why it will work correctly if we write, for example `indexOf(arr, x) = 0`? | 2020-09-18T18:51:17 | 2020-09-18T18:51:17 | {} | MEMBER |
ClickHouse | ClickHouse | 695,198,225 | 14,685 | alexey-milovidov | Now we have only 176 unstable queries (was 418), so maybe it has helped
(or other recent attempts to lower binary size have helped). | 2020-09-19T10:49:37 | 2020-09-19T10:49:37 | {} | MEMBER |
ClickHouse | ClickHouse | 695,214,748 | 14,521 | alexey-milovidov | @hexiaoting I don't remember what has changed but now it's Ok:
```
milovidov-desktop :) SELECT name FROM system.databases WHERE name ILIKE 'De%'
SELECT name
FROM system.databases
WHERE name ILIKE 'De%'
┌─name────┐
│ default │
└─────────┘
``` | 2020-09-19T13:39:25 | 2020-09-19T13:39:25 | {} | MEMBER |
ClickHouse | ClickHouse | 694,727,304 | 13,499 | alesapin | `Stress test (undefined) ` -- old version of stress-test script. | 2020-09-18T08:15:51 | 2020-09-18T08:15:51 | {} | MEMBER |
ClickHouse | ClickHouse | 694,869,026 | 13,617 | akuzm | ERROR: type should be string, got "https://clickhouse-test-reports.s3.yandex.net/13617/bb8ce1fbb7be14f5840ed39675f87b394dfcde25/performance_comparison/report.html#fail1\r\n\r\nEnabled the profiler back, +20% again." | 2020-09-18T13:30:46 | 2020-09-18T13:30:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,919,029 | 9,363 | jathael | Below added to odbc.ini worked for me:
`MaxCharSize = 2000000`
`MaxVarcharSize = 2000000`
`MaxLongVarcharSize = 2000000` | 2020-09-18T14:59:36 | 2020-09-18T14:59:59 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 694,437,039 | 11,844 | alexey-milovidov | ```
clickhouse-benchmark <<< "SELECT DISTINCT Title = decrypt('aes-128-cbc', encrypt('aes-128-cbc', Title, 'key-key-key-key-'), 'key-key-key-key-') FROM test.hits"
```
16 cores and 100 MB/sec.
```
OSSL_PARAM_locate
EVP_CipherInit_ex
__pthread_rwlock_rdlock !!!
OPENSSL_cleanse
```
Not quite acceptable. | 2020-09-17T19:02:50 | 2020-09-17T19:02:50 | {} | MEMBER |
ClickHouse | ClickHouse | 694,627,685 | 14,786 | liangli1213 | Sync from master has no problem, I have tried. | 2020-09-18T03:30:52 | 2020-09-18T03:30:52 | {} | NONE |
ClickHouse | ClickHouse | 694,647,044 | 14,786 | bohutang | I can't reproduce it, what's your slave my.cnf?
In principle, syncing from mysql master or slave there is no difference.
Here are some configs related to my mysql:
```
sudo docker network create --subnet=172.18.0.0/16 selnet
sudo docker run --net selnet --ip 172.18.0.3 -d -e MYSQL_ROOT_PASSWORD=123 mysql:5.7 mys... | 2020-09-18T04:43:31 | 2020-09-18T04:43:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,697,417 | 14,742 | alesapin | AST fuzzer — Task failed: $?=8 Race condition in docker images with master. | 2020-09-18T07:07:26 | 2020-09-18T07:07:26 | {} | MEMBER |
ClickHouse | ClickHouse | 694,704,366 | 7,228 | alesapin | Probably #13450 will improve DDL tasks processing, so try the latest 20.6+ version. If the issue reproduces just reopen. | 2020-09-18T07:23:51 | 2020-09-18T07:23:51 | {} | MEMBER |
ClickHouse | ClickHouse | 694,705,602 | 12,450 | alesapin | Have you seen similar errors in the latest clickhouse versions? I cannot reproduce the issue :( | 2020-09-18T07:26:52 | 2020-09-18T07:26:59 | {} | MEMBER |
ClickHouse | ClickHouse | 694,806,680 | 14,972 | 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-... | 2020-09-18T11:10:34 | 2020-09-18T11:10:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,836,512 | 14,201 | ramazanpolat | > BTW, it's Ok to implement this feature with another SystemLog table.
I think proposal @filimonov 's proposal (using a MV to get parse errors like `Select * from kafka_table where _error_code=0;`) is way better than a SystemLog table since it will give users more freedom.
| 2020-09-18T12:23:29 | 2020-09-18T12:23:29 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,944,000 | 9,624 | alesapin | Stress test -- race, but it exists for a long time. | 2020-09-18T15:46:48 | 2020-09-18T15:46:48 | {} | MEMBER |
ClickHouse | ClickHouse | 695,028,598 | 14,977 | alexey-milovidov | ```
MergeTreeIndexConditionBloomFilter.cpp:235:19: error: loop variable is copied but only used as const reference; consider making it a const reference [performance-for-range-copy,-warnings-as-errors]
2020-09-18 10:09:16 for (auto arg : arguments)
2020-09-18 10:09:16 ^
2020-09-18 10:09:16... | 2020-09-18T18:46:59 | 2020-09-18T18:46:59 | {} | MEMBER |
ClickHouse | ClickHouse | 695,103,431 | 10,562 | azat | >BTW, we already decided to migrate all Python scripts to Python 3.
@alexey-milovidov do you have any progress on this? (I already have some changes for this, but I can wait if you are working on this) | 2020-09-18T21:55:54 | 2020-09-18T21:55:54 | {} | MEMBER |
ClickHouse | ClickHouse | 695,216,325 | 13,869 | alexey-milovidov | `sysctl` variant should be Ok. | 2020-09-19T13:53:56 | 2020-09-19T13:53:56 | {} | MEMBER |
ClickHouse | ClickHouse | 694,437,581 | 11,844 | alexey-milovidov | ```
SELECT DISTINCT Title = decrypt('aes-128-cbc', encrypt('aes-128-cbc', Title, 'key-key-key-key-'), 'key-key-key-key+')
FROM test.hits
→ Progress: 131.07 thousand rows, 10.46 MB (1.27 million rows/s., 101.30 MB/s.) 1%
Received exception from server (version 20.10.1):
Code: 454. DB::Exception: Received from lo... | 2020-09-17T19:03:52 | 2020-09-17T19:03:52 | {} | MEMBER |
ClickHouse | ClickHouse | 694,563,947 | 14,786 | liangli1213 | @BohuTANG @zhang2014 , I just used build 20.10.1.4678, and test with steps I mentioned above, unfortunately, the issue is even worse, when doing step 3, below error occur very sooner:
`2020.09.18 07:56:00.172767 [ 16800 ] {} <Error> MaterializeMySQLSyncThread(): Query EXTERNAL DDL FROM MySQL(clickhouse_test, clickho... | 2020-09-18T00:01:34 | 2020-09-18T00:01:34 | {} | NONE |
ClickHouse | ClickHouse | 694,840,139 | 14,896 | l1t1 | create table a engine=Mergetree()order by number as select a.number*10000000+b.number*1000+c.number from numbers(1000)a ,numbers(1000) b,numbers(1000) c, took 80 seconds | 2020-09-18T12:32:34 | 2020-09-18T12:32:34 | {} | NONE |
ClickHouse | ClickHouse | 694,873,999 | 14,979 | liangli1213 | @tavplubix , thanks for your reply, I begin to use ClickHouse not so long, this afternoon, I make replication from MySQL to ClickHouse successfully, and do a perf test for one of the table.
I found the perf is very low, a query will take 40s, that is because I removed all index from the MySQL table before doing the ... | 2020-09-18T13:40:04 | 2020-09-18T13:40:04 | {} | NONE |
ClickHouse | ClickHouse | 694,915,806 | 14,973 | alesapin | We already do it, just add new settings here https://github.com/ClickHouse/ClickHouse/blob/master/src/Storages/MergeTree/MergeTreeDataMergerMutator.cpp#L1079-L1081 | 2020-09-18T14:53:26 | 2020-09-18T14:53:26 | {} | MEMBER |
ClickHouse | ClickHouse | 694,692,801 | 14,297 | alesapin | Release build killed by OOM, I'll just rerun it and release. | 2020-09-18T06:56:13 | 2020-09-18T06:56:13 | {} | MEMBER |
ClickHouse | ClickHouse | 694,726,693 | 13,499 | alesapin | Integration tests report https://clickhouse-test-reports.s3.yandex.net/13499/1fc6fc7590ff4ee4cd46ea3d923c9a61d37f55f1/integration_tests_(asan).html. Some flaky test failed? | 2020-09-18T08:14:37 | 2020-09-18T08:14:56 | {} | MEMBER |
ClickHouse | ClickHouse | 694,794,750 | 14,843 | akuzm | OK, so we don't have this in the integration tests. The exact scenario where it fails is trying to load config from ZK node (`from_zk` includes). The fix is trivial (initialize thread pool earlier than this). The bad part is writing an integration test for this -- not sure how I'll be able to bring myself to do it. | 2020-09-18T10:40:48 | 2020-09-18T10:40:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,806,665 | 14,972 | alexey-milovidov | As far as I understand it is for syntax highlight in documentation. We have no issues with syntax highlight. | 2020-09-18T11:10:31 | 2020-09-18T11:10:31 | {} | MEMBER |
ClickHouse | ClickHouse | 695,141,717 | 14,979 | liangli1213 | I have one primary key and 2 index in my table, as follows:
```
PRIMARY KEY (`id`),
KEY `vendor_feature_checkDate` (`vendor`,`feature`,`checkDate`),
KEY `site_vendor_checkDate` (`site`,`vendor`,`checkDate`)
```
And ClickHouse convert it to order by clause:
```
ORDER BY (feature, checkDate, site, vendor, id)
``... | 2020-09-19T00:59:40 | 2020-09-19T00:59:40 | {} | NONE |
ClickHouse | ClickHouse | 695,209,761 | 14,886 | alexey-milovidov | Azat already started working on it: https://github.com/ClickHouse/ClickHouse/pull/15007 | 2020-09-19T12:54:32 | 2020-09-19T12:54:32 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 694,495,531 | 14,880 | vitlibar | AST fuzzer fail is known, see https://github.com/ClickHouse/ClickHouse/issues/13393 | 2020-09-17T20:56:42 | 2020-09-17T20:56:42 | {} | MEMBER |
ClickHouse | ClickHouse | 694,708,022 | 14,684 | azat | >AST fuzzer — Task failed: $?=8
```
2020-09-18 08:01:03 + wget -nv -nd -c https://clickhouse-builds.s3.yandex.net/14684/c9baceb7605799dca12ed9450fc2ec230baed523/clickhouse_build_check/clang-10_debug_none_bundled_unsplitted_disable_False_binary/clickhouse
2020-09-18 08:01:03 https://clickhouse-builds.s3.yandex.ne... | 2020-09-18T07:32:41 | 2020-09-18T07:32:41 | {} | MEMBER |
ClickHouse | ClickHouse | 694,803,818 | 14,201 | alexey-milovidov | BTW, it's Ok to implement this feature with another SystemLog table. | 2020-09-18T11:03:17 | 2020-09-18T11:03:17 | {} | MEMBER |
ClickHouse | ClickHouse | 690,752,900 | 14,711 | myrrc | TODO scripting:
- [x] Links must point to line numbers
- [x] Investigate whether https://regex101.com/r/R6iogw/8 operates on all lines correctly but the sed on my computer misses a few options (and doesn't parse comments).
- [x] Less write-only code (rewrite to Python).
- [x] Treat missing default values as `OFF`... | 2020-09-10T21:55:38 | 2020-09-18T13:46:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 695,157,102 | 14,535 | zhang2014 | > Maybe change changelog category to "Improvement" or "New feature"? The entry "Support decimal data type for MaterializedMySQL" doesn't look like a bugfix.
Done. | 2020-09-19T03:22:14 | 2020-09-19T03:22:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 695,205,269 | 14,892 | vzakaznikov | @alexey-milovidov, thanks merging this PR. Just note that ` test_distributed_over_live_view_order_by_node` might still fail from time to time due to the bug described above. For now, we might need to disable or mark this test as a known fail. | 2020-09-19T12:06:09 | 2020-09-19T12:06:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 695,208,574 | 4,901 | sxpsxp12 | i'm sorry, i cant understand，so, finally, how resolve it | 2020-09-19T12:42:38 | 2020-09-19T12:42:38 | {} | NONE |
ClickHouse | ClickHouse | 695,215,661 | 14,588 | alexey-milovidov | All issues resolved. | 2020-09-19T13:47:40 | 2020-09-19T13:47:40 | {} | MEMBER |
ClickHouse | ClickHouse | 694,713,715 | 13,953 | man-brain | Requirements lack the need to control resource usage during the backup. We definitely want to have the ability to make the backup as quickly as possible, but we also don't want production workload suffering due to CPU or I/O starvation during backup. | 2020-09-18T07:46:25 | 2020-09-18T07:46:25 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 694,715,145 | 10,083 | alesapin | > ```sql
> SELECT version()
>
> ┌─version()───┐
> │ 20.3.12.112 │
> └─────────────┘
>
> 2020.07.31 16:29:15.446095 [ 19442 ] {4f3b7ca1-9eb0-4b08-a7b4-e0c7c7f98ae6} <Error> executeQuery: Code: 49, e.displayText() = DB::Exception: Could not find a column of minimum size in MergeTree, part /data1/clickhouse/data/... | 2020-09-18T07:49:39 | 2020-09-18T07:49:39 | {} | MEMBER |
ClickHouse | ClickHouse | 694,804,948 | 10,083 | alexey-milovidov | If you are using non-Replicated MergeTree, the error can be caused by data corruption in filesystem.
Check the contents of `/data1/clickhouse/data/xxx/xxx_msg/20190228_700983_713697_27_3953449/`, also check the server `uptime` and look for any suspicious messages in `dmesg` and `/var/log/kern.log`.
If you will us... | 2020-09-18T11:06:16 | 2020-09-18T11:06:16 | {} | MEMBER |
ClickHouse | ClickHouse | 694,833,564 | 14,979 | tavplubix | 1. "indexes" in MySQL and "indexes" in ClickHouse are very different things. However, ClickHouse automatically choose `ORDER BY` and `PARTITION BY` keys for synced table when it rewrites the `CREATE` query from MySQL syntax to ClickHouse syntax.
2. No, it's not possible to `ALTER` MaterializeMySQL table manually. Pro... | 2020-09-18T12:16:21 | 2020-09-18T12:16:21 | {} | MEMBER |
ClickHouse | ClickHouse | 695,042,907 | 11,913 | alexey-milovidov | @akuzm AST Fuzzer:
`2020-09-18 08:00:36 ERROR 404: Not Found.` | 2020-09-18T19:19:50 | 2020-09-18T19:19:50 | {} | MEMBER |
ClickHouse | ClickHouse | 694,700,752 | 14,786 | zhang2014 | Maybe we can close this issue. | 2020-09-18T07:15:23 | 2020-09-18T07:15:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,842,823 | 14,953 | zhang2014 | Got it. by the way, we will manually commit backport to 20.8(include some other MaterializeMySQL bugs fix ) | 2020-09-18T12:38:36 | 2020-09-18T12:38:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,858,317 | 14,535 | tavplubix | Maybe change changelog category to "Improvement" or "New feature"? The entry "Support decimal data type for MaterializedMySQL" doesn't look like a bugfix. | 2020-09-18T13:10:19 | 2020-09-18T13:10:19 | {} | MEMBER |
ClickHouse | ClickHouse | 694,863,656 | 14,921 | alexey-milovidov | Now it's 11% slower than gcc, was 21%. | 2020-09-18T13:20:24 | 2020-09-18T13:20:24 | {} | MEMBER |
ClickHouse | ClickHouse | 694,864,849 | 14,921 | alexey-milovidov | > Yandex third-party checks (only for Yandex employees)
Already broken, don't care... | 2020-09-18T13:22:43 | 2020-09-18T13:22:43 | {} | MEMBER |
ClickHouse | ClickHouse | 694,870,767 | 11,684 | alesapin | Test already finished https://clickhouse-test-reports.s3.yandex.net/11684/c8a5d9e0a2da0a30b2f5881a727126a2a5ac9530/functional_stateless_tests_(thread).html. | 2020-09-18T13:34:01 | 2020-09-18T13:34:01 | {} | MEMBER |
ClickHouse | ClickHouse | 695,199,034 | 14,892 | alexey-milovidov | I will merge and @KochetovNicolai will answer later. | 2020-09-19T10:59:01 | 2020-09-19T10:59:01 | {} | MEMBER |
ClickHouse | ClickHouse | 695,198,972 | 14,962 | alexey-milovidov | test_distributed_over_live_view - known bad test @vzakaznikov
Probably fixed in https://github.com/ClickHouse/ClickHouse/pull/14892 | 2020-09-19T10:58:18 | 2020-09-19T10:59:20 | {} | MEMBER |
ClickHouse | ClickHouse | 695,215,586 | 14,045 | alexey-milovidov | Fixed in #14646 | 2020-09-19T13:46:53 | 2020-09-19T13:46:53 | {} | MEMBER |
ClickHouse | ClickHouse | 694,439,759 | 14,776 | tavplubix | Probably we should allow attaching another instance of storage if it doesn't store data on disk, but I'm not sure. | 2020-09-17T19:08:03 | 2020-09-17T19:08:03 | {} | MEMBER |
ClickHouse | ClickHouse | 694,629,049 | 14,954 | zhang2014 | Why not backport? | 2020-09-18T03:36:59 | 2020-09-18T03:36:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,638,704 | 14,930 | somewheve | alright...... | 2020-09-18T04:11:01 | 2020-09-18T04:11:01 | {} | NONE |
ClickHouse | ClickHouse | 694,899,018 | 12,300 | alexey-milovidov | You can configure `text_log` to use `ENGINE = File` with format of your choice. | 2020-09-18T14:23:34 | 2020-09-18T14:23:34 | {} | MEMBER |
ClickHouse | ClickHouse | 695,015,076 | 14,831 | bharatnc | Related to #4821 | 2020-09-18T18:17:05 | 2020-09-18T18:17:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,627,476 | 14,786 | bohutang | Can you try sync from the master and have a check? | 2020-09-18T03:30:08 | 2020-09-18T03:30:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,690,756 | 14,926 | alesapin | Performance not affected. | 2020-09-18T06:50:52 | 2020-09-18T06:50:52 | {} | MEMBER |
ClickHouse | ClickHouse | 694,704,861 | 14,786 | liangli1213 | I will close it later, the test is still in progress. | 2020-09-18T07:25:11 | 2020-09-18T07:25:11 | {} | NONE |
ClickHouse | ClickHouse | 694,958,153 | 14,999 | den-crane | STR:
```
CREATE TABLE Repl
(
`A` Integer,
`S` String
)
ENGINE = ReplicatedMergeTree('/clickhouse/tables/testxxx', 'changeme')
ORDER BY A
SETTINGS replicated_deduplication_window_seconds = 3, replicated_deduplication_window = 5555555;
insert into Repl values(1,'');
select sleep(3) from numbers(20)... | 2020-09-18T16:15:30 | 2020-09-18T16:15:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 695,141,752 | 13,953 | bohutang | @dev1ant
We can limit the IO throughput by throttle parameter for the backup. | 2020-09-19T00:59:57 | 2020-09-19T00:59:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,436,274 | 13,574 | xjewer | @alesapin can you elaborate, why the issue is closed with the same comment, that @nvartolomei was trying thoroughly addressing. Seems like we have a bit of misunderstanding in the proposed implementation.
Unarguably, parts mutations is not something convenient to propagate fingerprints, but it does handle stopping m... | 2020-09-17T19:01:26 | 2020-09-17T19:03:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,700,942 | 14,847 | zhang2014 | 👍 | 2020-09-18T07:15:49 | 2020-09-18T07:15:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,843,692 | 14,535 | zhang2014 | `AST fuzzer — Assertion `(n >= (static_cast<ssize_t>(pad_left_) ? -1 : 0)) && (n <= static_cast<ssize_t>(this->size()))' failed`
test failure unrelated this pr. | 2020-09-18T12:40:27 | 2020-09-18T12:40:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 694,864,072 | 14,921 | alexey-milovidov | All queries with `materialize` was sped up and now faster than gcc. | 2020-09-18T13:21:13 | 2020-09-18T13:21:13 | {} | MEMBER |
ClickHouse | ClickHouse | 695,013,445 | 14,684 | azat | >Ok. I also worried because DDLWorker is typical pain point - it does not scale for large number of queries for several other reasons. I doubt if parallel execution will solve it.
Sure you cannot execute ON CLUSTER queries with the same RPS as w/o it (due to zookeeper and so on), but usually these are ALTER queries ... | 2020-09-18T18:13:32 | 2020-09-18T18:13:32 | {} | MEMBER |
ClickHouse | ClickHouse | 695,142,716 | 14,979 | bohutang | For your question:
A1:
right, ClickHouse OrderBy is the only physical sort order on the disk.
But we can use MaterializedView to make many different tables with different OrderBy as MySQL index does.
Here some info about MaterializedView:
https://clickhouse.tech/docs/en/sql-reference/statements/create/view/#mate... | 2020-09-19T01:07:08 | 2020-09-19T01:08:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 695,173,406 | 4,070 | bohutang | Hi,
It's a bit late, but we can now use MaterializeMySQL engine to replicate MySQL datas.
```
CREATE DATABASE ckdb ENGINE = MaterializeMySQL('172.17.0.2:3306', 'ckdb', 'root', '123');
``` | 2020-09-19T06:38:13 | 2020-09-19T06:38:13 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 696,793,454 | 15,123 | den-crane | It's how copier works nowadays. For idempotent atomic re-tries.
It creates a xxx_piece_x table for each partition of a destination table. And after success insert into xxx_piece_x it attaches data from xxx_piece_x to the newtable. | 2020-09-22T15:24:00 | 2020-09-22T15:27:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 696,826,709 | 14,940 | akuzm | Yandex synchronization check was flappy https://nda.ya.ru/t/eiP2bb893Y4JK9
Stress test broken in master.
Other tests have passed before. | 2020-09-22T16:17:16 | 2020-09-22T16:17:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 696,869,696 | 14,265 | filimonov | Stress test failure is unrelated, reported here: https://github.com/ClickHouse/ClickHouse/issues/14414#issuecomment-696864958 | 2020-09-22T17:34:28 | 2020-09-22T17:34:28 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.