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 | 640,509,420 | 8,223 | filimonov | In general: in common case data does not land in a single part/partition.
If your part has already a size of 150Gb it will never be merged currently (and it would be too heavy operation to merge 150Gb part with every incoming block).
So the proposed solution can work in some very special use cases (with small d... | 2020-06-08T10:11:26 | 2020-06-08T10:11:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,552,043 | 11,392 | den-crane | > Has life become unbearable because of this?
No. It's even not my issue. I opened it because I though it's a bug. If it was intended then I am OK. | 2020-06-08T11:43:11 | 2020-06-08T11:43:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,587,890 | 11,459 | vzakaznikov | > @vzakaznikov does the above seem like it would work with a `LiveView` (considering number of rows?)
@jakesyl, I am not sure. You will have to give it a try. From the first glance it seems like what you need is MV. Live views do not store any data and just keep partial query result in memory. So it really depends o... | 2020-06-08T13:00:41 | 2020-06-08T13:00:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,710,453 | 11,519 | filimonov | There was one manual conflict resolution (related to db atomic).
Didn't recheck build locally, need CI (kafka) tests to pass, labels of affected PRs need to be fixed. | 2020-06-08T15:43:46 | 2020-06-08T15:43:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,128,647 | 9,545 | And42 | Merged within pull request: https://github.com/ClickHouse/ClickHouse/pull/11226 | 2020-06-06T22:54:42 | 2020-06-06T22:54:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,289,320 | 8,995 | alexey-milovidov | @akuzm memory leak is the top priority bug. It just should not exist. | 2020-06-07T22:29:07 | 2020-06-07T22:29:20 | {} | MEMBER |
ClickHouse | ClickHouse | 640,425,969 | 11,482 | filimonov | > pretty unreliable with Clickhouse running in a Docker container, so I rebuilt it with frame pointer support.
So it looks like not DateTime64 related. Can you please update the issue title and description correspondingly? | 2020-06-08T07:40:40 | 2020-06-08T07:40:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,479,409 | 11,500 | CurtizJ | You can set settings `min_rows_for_wide_part` and `min_bytes_for_wide_part` for any `MergeTree`-family table (by `ALTER MODIFY SETTING` query or globally in config). In that case, if size of inserted part in rows or in bytes less than corresponding threshold, part will will be written in compact format, otherwise in wi... | 2020-06-08T09:15:34 | 2020-06-08T09:16:05 | {} | MEMBER |
ClickHouse | ClickHouse | 640,255,279 | 11,468 | den-crane | You need to upgrade to 20.1.12.86.
Probably you are experiencing https://github.com/ClickHouse/ClickHouse/issues/9019 https://github.com/ClickHouse/ClickHouse/issues/8982 and then you need to reload data / re-create the table because a primary key index is corrupted. | 2020-06-07T17:52:24 | 2020-06-07T17:52:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,294,505 | 11,492 | alexey-milovidov | This pull request has high chance to be abandoned. | 2020-06-07T23:16:45 | 2020-06-07T23:16:45 | {} | MEMBER |
ClickHouse | ClickHouse | 640,502,355 | 8,223 | filimonov | > > Isn't it the same as FINAL? `select ... from table FINAL` ?
>
> On the query output, yes it's. But usage of FINAL limits the query being single threaded.
That situation changed here: https://github.com/ClickHouse/ClickHouse/pull/10463
You can try it in one of 20.5 testing releases.
| 2020-06-08T09:58:00 | 2020-06-08T09:59:06 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,529,002 | 10,286 | qoega | Just to make it clear. If we suppose the most paranoid setup(and have no bugs):
```
min_rows_to_sync_after_merge=1
min_rows_to_sync_before_removing_source=1
sync_on_insert=1
```
+ using RAID under fs.
We can't lose data after answering OK to user in case of sudden power failure or single disk failure?
| 2020-06-08T10:52:26 | 2020-06-08T10:52:26 | {} | MEMBER |
ClickHouse | ClickHouse | 640,544,196 | 8,377 | alesapin | Closed in favor of https://github.com/ClickHouse/ClickHouse/pull/11509 | 2020-06-08T11:26:25 | 2020-06-08T11:26:25 | {} | MEMBER |
ClickHouse | ClickHouse | 640,562,603 | 10,607 | qoega | For documentation:
https://github.com/ClickHouse/ClickHouse/pull/10607/files#diff-fef10815cbf8ba0462108a68efe8f4c4R12-R14 | 2020-06-08T12:07:16 | 2020-06-08T12:07:30 | {} | MEMBER |
ClickHouse | ClickHouse | 640,654,803 | 10,163 | Felixoid | #11514 | 2020-06-08T14:24:31 | 2020-06-08T14:24:31 | {} | MEMBER |
ClickHouse | ClickHouse | 640,110,377 | 11,293 | alexey-milovidov | It's harmful, because it may prevent server to startup for 120 seconds. | 2020-06-06T19:55:39 | 2020-06-06T19:55:39 | {} | MEMBER |
ClickHouse | ClickHouse | 640,120,341 | 11,487 | alexey-milovidov | Note that Arcadia (Yandex synchronization check) is using outdated version of Poco.
It's Ok to completely disable DNS cache for Arcadia. | 2020-06-06T21:27:34 | 2020-06-06T21:27:34 | {} | MEMBER |
ClickHouse | ClickHouse | 640,273,474 | 7,096 | alexey-milovidov | @azat Yes, it's also done: #11446 | 2020-06-07T20:15:24 | 2020-06-07T20:15:24 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 640,328,596 | 11,458 | alexey-milovidov | > Functional stateless tests (unbundled) — fail: 1, passed: 1873, skipped: 40
The new test cannot work for "unbundled" build, I've disabled it.
> Functional stateless tests (debug) — fail: 1, passed: 1904, skipped: 9
Flaky test is reported. | 2020-06-08T02:35:38 | 2020-06-08T02:35:38 | {} | MEMBER |
ClickHouse | ClickHouse | 640,350,321 | 11,464 | zzsmdfj | > @zzsmdfj the test error is `WARNING:mkdocs.structure.pages:Documentation file 'sql-reference/functions/type-conversion-functions.md' contains a link to 'sql-reference/data-types/lowcardinality.md' which is not found in the documentation files.`. Could you fix it by adding the lowcardinality.md which is missing in zh ... | 2020-06-08T03:58:08 | 2020-06-08T03:58:08 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,352,081 | 11,328 | shubham721 | Thanks @den-crane
I have added account_id in the sorting key. Now, This OOM issue is not coming up but query is still taking around 500ms.
I was just wondering that in the following documentation
https://clickhouse.tech/docs/en/engines/table-engines/mergetree-family/mergetree/#primary-keys-and-indexes-in-queries
... | 2020-06-08T04:05:31 | 2020-06-08T04:05:31 | {} | NONE |
ClickHouse | ClickHouse | 640,529,069 | 10,950 | Akazz | Quite a work has been done here thanks to you, @myrrc! And yet there is more to be done to make this implementation work. Thus we better suspend work on this PR until later. | 2020-06-08T10:52:36 | 2020-06-08T10:52:36 | {
"rocket": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,588,576 | 11,494 | Turum | datagrip, tabbix, also tried over linked server session from mssql. Each returns the same error. | 2020-06-08T13:02:01 | 2020-06-08T13:02:01 | {} | NONE |
ClickHouse | ClickHouse | 640,589,150 | 11,401 | akuzm | > The new test results just don't make sense.
> I'll run it again, with more iterations, less frequent profiler, etc.
jemalloc code paths have changed, they are somewhat more frequent in profiles, SoftPageFaults metric variability is the same for at least some queries.
By the way, the most unstable queries (cryp... | 2020-06-08T13:03:16 | 2020-06-08T13:03:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,599,234 | 10,050 | qoega | Can you try to write a test that follows described case? | 2020-06-08T13:21:48 | 2020-06-08T13:21:48 | {} | MEMBER |
ClickHouse | ClickHouse | 640,122,568 | 11,482 | bobrik | Looks like `--call-graph dwarf` is pretty unreliable with Clickhouse running in a Docker container, so I rebuilt it with frame pointer support. Missing zstd frames were a giveaway.
* Query with `optimize_read_in_order = 0` running in a loop:
)
Engine=MergeTree()
PARTITION BY tu... | 2020-06-07T14:42:59 | 2020-06-07T14:42:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,342,827 | 11,473 | den-crane | It's rather expected behavior.
`select *` is a mis-usage. Try to select only required columns.
try `set join_algorithm = 'auto', max_block_size=1000` | 2020-06-08T03:29:03 | 2020-06-08T03:29:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,419,131 | 11,475 | filimonov | Please help a bit with debugging that:
Run `SHOW TABLE STATUS LIKE 'clients'` on MySQL 5.7 and on MySQL 8, and compare the output.
What is the table engine on the MySQL side?
Is the table partitioned on MySQL side?
Do the update time reported by `SHOW TABLE STATUS` (on MySQL side) corresponds to the actual... | 2020-06-08T07:26:00 | 2020-06-08T07:26:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,614,127 | 9,621 | den-crane | @ralphM78 it seems that the issue is back. Try this exact version 20.4.1.3163 / or 20.4.2.9.
```
20.4.1.3163
echo 'select count() from metrics PREWHERE (Date >= '"'"2000-01-01"'"') AND (Date <= '"'"2000-01-01"'"') where (Path in('`clickhouse-client -q "select arrayMap(x -> toString(x), range(10000))"|sed 's/\[\... | 2020-06-08T13:47:40 | 2020-06-08T14:08:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,637,148 | 11,466 | sneakythr0ws | My colleague had same error with selects during another inserts(1000 per minute). | 2020-06-08T14:11:06 | 2020-06-08T14:11:06 | {} | NONE |
ClickHouse | ClickHouse | 640,294,122 | 11,493 | alexey-milovidov | "Arcadia" is broken because it contains copy-paste code. Will fix after merge. | 2020-06-07T23:13:11 | 2020-06-07T23:13:11 | {} | MEMBER |
ClickHouse | ClickHouse | 640,297,732 | 11,493 | abyss7 | > "Unbundled" build is broken accidentially by @abyss7
> Already fixed in master.
It looks more like it was already broken by @proller some time ago, when he added the logic for "unbundled" build and haven't even tested it with our docker image.
Also, unfortunately, we can't test the update to Dockerfile beforeh... | 2020-06-07T23:44:37 | 2020-06-07T23:44:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,457,924 | 11,490 | alesapin | Thank you for your report. Can you provide query with id `3a55e895-0716-4503-8baa-788ce119e7b8`? You can just `grep '3a55e895-0716-4503-8baa-788ce119e7b8' /var/log/clickhouse-server/clickhouse-server.log` or select it from `system.query_log`: `select query from system.query_log where query_id='3a55e895-0716-4503-8baa-7... | 2020-06-08T08:39:48 | 2020-06-08T08:39:48 | {} | MEMBER |
ClickHouse | ClickHouse | 640,513,732 | 9,621 | ralphM78 | Hello. I have tried to upgrade all nodes to clickhouse-server-20.4.4.18-2. Without any select queries everything looks good. As soon as the first select requests start, load increases and the processlist is growing. After blocking all select queries everything went to normal. "perf top" shows
```
50.26% ParalInput... | 2020-06-08T10:20:10 | 2020-06-08T10:20:10 | {} | NONE |
ClickHouse | ClickHouse | 640,510,056 | 1,287 | etah000 | I have finished a patch to support distributed ddl(currently, only delete/updatet/drop partition operations have been tested) on cross replication clusters。 [distributed-alter-query](https://github.com/etah000/ClickHouse/tree/distributed-alter-query)。For distributed ddl on crosss replication clusters, the syntax should... | 2020-06-08T10:12:38 | 2020-06-08T10:20:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,531,126 | 8,223 | filimonov | Closing that since the story/concept can't work for the general case, and the use case is actually about deduplication, so let's use #5125 to track that. | 2020-06-08T10:57:00 | 2020-06-08T10:57:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,543,254 | 11,479 | 4ertus2 | If it's really fixed in 20.4 `set experimental_use_processors = 1` could be a workaround for 20.3. | 2020-06-08T11:24:25 | 2020-06-08T11:24:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,572,025 | 11,494 | Turum | here it is.
```
2020.06.08 15:22:34.715031 [ 2379 ] {d381b90d-4688-4c0d-8280-0fabd1817b4e} <Error> executeQuery: Code: 24, e.displayText() = DB::Exception: Cannot write to ostream at offset 2097152: While executing TabSeparatedRowOutputFormat (version 20.3.8.53) (from 10.2.146.189
:60125) (in query: SELECT *^M FRO... | 2020-06-08T12:27:53 | 2020-06-08T12:27:53 | {} | NONE |
ClickHouse | ClickHouse | 640,126,854 | 11,293 | azat | >It's harmful, because it may prevent server to startup for 120 seconds.
And not only this, but eventually server will not start
>It will be fixed automatically when we will change the engine of system database to Atomic #7512
Any ETA?
>We have loaded system database and started up tables. Some tables start... | 2020-06-06T22:35:02 | 2020-06-06T22:35:02 | {} | MEMBER |
ClickHouse | ClickHouse | 640,245,022 | 11,458 | alexey-milovidov | > Need to replace support for colors in replxx, so WHITE will actually mean "intense".
Done. | 2020-06-07T16:34:36 | 2020-06-07T16:34:36 | {} | MEMBER |
ClickHouse | ClickHouse | 640,352,934 | 11,464 | zzsmdfj | by the way, i add lowCardinality.md file. | 2020-06-08T04:08:39 | 2020-06-08T04:08:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,383,089 | 8,085 | LiemLin | @TisonKun how did you fix this one? | 2020-06-08T05:45:41 | 2020-06-08T05:45:41 | {} | NONE |
ClickHouse | ClickHouse | 640,466,355 | 11,392 | alesapin | Has life become unbearable because of this? | 2020-06-08T08:54:16 | 2020-06-08T08:54:16 | {} | MEMBER |
ClickHouse | ClickHouse | 640,550,829 | 11,468 | den-crane | >Is there any way to recreate the table
No
Have you upgrade CH ? Has it solve the issue? | 2020-06-08T11:40:47 | 2020-06-08T11:40:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,561,355 | 11,510 | den-crane | try to add parameters
clickhouse-client --input_format_parallel_parsing=0 --max_insert_block_size=1000000 | 2020-06-08T12:04:34 | 2020-06-08T12:04:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,132,072 | 11,458 | alexey-milovidov | `00993_system_parts_race_condition_drop_zookeeper` - reported. | 2020-06-06T23:33:52 | 2020-06-06T23:33:52 | {} | MEMBER |
ClickHouse | ClickHouse | 640,177,585 | 11,391 | azat | >Also, it's won't work, if we implement any other authentication options. So it may be a temporary solution for now.
Actually the same thing is used in `StorageReplicatedMergeTree::sendRequestToLeaderReplica` | 2020-06-07T08:33:24 | 2020-06-07T08:33:24 | {} | MEMBER |
ClickHouse | ClickHouse | 640,223,410 | 11,463 | den-crane | There is no way to achieve it using MV. MV is an insert trigger. It never reads a source table. | 2020-06-07T13:59:41 | 2020-06-07T14:00:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,223,907 | 11,468 | den-crane | What CH version do you use?
Can you provide `show create table my_table`?
Do you have a sample data to reproduce such behavior? | 2020-06-07T14:03:27 | 2020-06-07T14:03:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,249,863 | 11,468 | asurovenko | I use **20.1.3.7** version.
And query for creating table(some fields hidden):
`CREATE TABLE my_table (`date` DateTime, `value` String, `value2` LowCardinality(String), `value3` Int32) ENGINE = ReplicatedMergeTree('/clickhouse/tables/{shard}/dev/my_table', '{replica}') PARTITION BY date ORDER BY date SETTINGS index_g... | 2020-06-07T17:11:14 | 2020-06-07T17:11:14 | {} | NONE |
ClickHouse | ClickHouse | 640,590,097 | 11,510 | Turum | It works, many thanks. | 2020-06-08T13:05:10 | 2020-06-08T13:05:10 | {} | NONE |
ClickHouse | ClickHouse | 640,673,933 | 11,494 | Turum | But this is if I understand correctly the CH client parameter, I get the same error when I'm trying to execute the same select over mssql linked server like this.
```
exec ('SELECT *
FROM odbc(''DSN=QExport;Uid=clickhouse;Pwd=clickhouse;Timeout=0;'',''dbo'', ''v_TodayOrders_ODBC'')
') at [CLICKHOUSE_25]
```
So ... | 2020-06-08T14:42:51 | 2020-06-08T14:42:51 | {} | NONE |
ClickHouse | ClickHouse | 640,224,981 | 11,244 | den-crane | MV ` default.test_aggregated_hour_mv TO default.test_aggregated_hour` knows nothing about the data stored or engines of the source tables `test` , `test_aggregated`.
Here an example with Null engine.
test and test_aggregated store nothing!!!
```
/* ------------ENGINE = Null; -------- */
CREATE TABLE default.t... | 2020-06-07T14:12:13 | 2020-06-07T14:12:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,310,661 | 11,499 | alexey-milovidov | Please teach me how to do it with SystemTap. | 2020-06-08T01:08:40 | 2020-06-08T01:08:40 | {} | MEMBER |
ClickHouse | ClickHouse | 640,394,168 | 8,085 | tisonkun | IIUC something related to https://github.com/ClickHouse/ClickHouse/pull/8600 . But if you pull the latest master it should work. | 2020-06-08T06:21:23 | 2020-06-08T06:21:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,227,095 | 11,409 | den-crane | Can you provide the table DDL and the output of the query debug `SET send_logs_level = 'debug'`
| 2020-06-07T14:27:45 | 2020-06-07T14:27:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,256,953 | 11,491 | alexey-milovidov | It's a bug in gcc. You can update it to 9.2.1, 9.3 or 10, or to clang 10. | 2020-06-07T18:05:45 | 2020-06-07T18:05:45 | {} | MEMBER |
ClickHouse | ClickHouse | 640,272,889 | 7,096 | azat | Any plans on doing the same for `DB::Exception`? | 2020-06-07T20:10:52 | 2020-06-07T20:10:52 | {} | MEMBER |
ClickHouse | ClickHouse | 640,289,765 | 11,495 | vzakaznikov | This PR is to address https://github.com/ClickHouse/ClickHouse/issues/10460 and comments regarding usage of live view in https://github.com/ClickHouse/ClickHouse/issues/11459. | 2020-06-07T22:32:52 | 2020-06-07T22:32:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,381,121 | 11,462 | TianyiLu | Referring to the document I finally solve the problem. It is the certificate issuer that make the handshake failure happens. The common name, CN of server.crt should be **localhost** when testing at local. So I tried the tutorial to generate new certicate that CN=localhost and it works.
Note that the CN of the CA c... | 2020-06-08T05:39:10 | 2020-06-08T05:39:10 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 640,359,272 | 11,331 | shubham721 | Thanks @den-crane
In your query, you have use argmax based on rto_count = 0, But we can't distinguish b/w the case of update event where this is default(zero) and the case event comes again where the value is zero and we want this in updated record.
With replacingmergetree and collapsingmergetree, problem I am facin... | 2020-06-08T04:28:39 | 2020-06-08T05:45:27 | {} | NONE |
ClickHouse | ClickHouse | 640,427,227 | 8,085 | LiemLin |
@TisonKun
**Hello, I used the latest master, but the following error occurred：**
[2567/4144] Building CXX object programs/client/CMakeFiles/clickhouse-client-lib.dir/Client.cpp.o
FAILED: programs/client/CMakeFiles/clickhouse-client-lib.dir/Client.cpp.o
/usr/bin/clang++ -DBOOST_SYSTEM_NO_DEPRECATED -DPOCO_ENABL... | 2020-06-08T07:43:01 | 2020-06-08T07:43:01 | {} | NONE |
ClickHouse | ClickHouse | 640,472,512 | 8,597 | alesapin | This PR's implementation seems a bit confusing both from the implementation side and user experience. I think that we should implement it in a different way. Add two new types of queries `ALTER TABLE aaa MODIFY COLUMN bbb DROP TTL` and `ALTER TABLE aaa DROP TTL`. I'll close this PR, because there is no progress for a l... | 2020-06-08T09:04:53 | 2020-06-08T09:04:53 | {} | MEMBER |
ClickHouse | ClickHouse | 640,520,795 | 10,923 | Enmk | Closes #10672 | 2020-06-08T10:34:48 | 2020-06-08T10:34:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,539,207 | 10,697 | CurtizJ | For reviewers: `MergeTreeDataPartWriterOnDisk` is old `IMergeTreeDataPartWriter`. Git correctly shows it as rename in 9a2855f, but in ui diff is shown incorrectly. So, there is nom need to review these changes. | 2020-06-08T11:15:14 | 2020-06-08T11:15:14 | {} | MEMBER |
ClickHouse | ClickHouse | 640,575,534 | 11,494 | den-crane | It looks like a timeout in your client application. Not a CH issue.
What do you use as a client? | 2020-06-08T12:35:33 | 2020-06-08T12:35:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,577,477 | 10,542 | siradjev | Hi, can this fix be introduced to 20.3 LTS too?
Currently, due to #10483 it's not possible to troubleshoot any query with dictionary usage.
Alternatively, can #10483 itself be fixed in 20.3? | 2020-06-08T12:39:45 | 2020-06-08T12:39:45 | {} | NONE |
ClickHouse | ClickHouse | 640,577,997 | 11,300 | alexey-milovidov | Integration tests of ODBC interaction are broken in this pull request.
Most likely due to the fact that we check the scenario of segfault inside clickhouse-odbc-bridge.
But as clickhouse-odbc-bridge is a special process that is expected to crash, we should disable sentry for it. | 2020-06-08T12:40:50 | 2020-06-08T12:40:50 | {} | MEMBER |
ClickHouse | ClickHouse | 640,497,873 | 10,163 | Felixoid | Looks solved to me. Is it?
**upd**: No, doesn't seem to work with kafka
```
CREATE TABLE analytics.kafka_string
(
`string` String
)
ENGINE = Kafka
SETTINGS kafka_broker_list = 'kafka.local:9092', kafka_topic_list = 'events-valid', kafka_group_name = 'ch-test-mikhail-1', kafka_format = 'JSONAsString', kafk... | 2020-06-08T09:49:22 | 2020-06-08T13:10:26 | {
"heart": 2
} | MEMBER |
ClickHouse | ClickHouse | 640,109,576 | 11,482 | bobrik | I was able to replicate your findings with synthetic data, but unfortunately they don't seem to apply to production data in a straightforward way.
```
SELECT count() as count,
countIf(service = 'nginx-fl') target_count
FROM jaeger_index_v2
WHERE timestamp >= now() - 3600
AND timestamp <= now()
... | 2020-06-06T19:48:46 | 2020-06-06T19:48:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,555,414 | 11,494 | den-crane | It does not look like CH error. Can you copy/paste error+trace from CH logs? | 2020-06-08T11:50:54 | 2020-06-08T11:50:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,585,749 | 11,401 | akuzm | The new test results just don't make sense.
I'll run it again, with more iterations, less frequent profiler, etc. | 2020-06-08T12:56:30 | 2020-06-08T12:56:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,715,806 | 11,520 | filimonov | There was one manual conflict resolution (related to db atomic).
Didn't recheck build locally, need CI (kafka) tests to pass, labels of affected PRs need to be fixed. | 2020-06-08T15:52:47 | 2020-06-08T15:52:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,225,590 | 11,328 | den-crane | try to use `having`
```
SELECT
return_ref_id,
return_id,
tenant_id,
max(account_id) AS max_account_id,
...
GROUP BY (tenant_id, return_ref_id, return_id)
having account_id='ACC14002975663275913'
```
try to execute the same query in single thread `set max_threads=1` | 2020-06-07T14:16:51 | 2020-06-07T14:16:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,553,407 | 11,444 | Slach | @blinkov link fixed on https://github.com/ClickHouse/ClickHouse/pull/11444/commits/62707935307688a6ff0f321674118d4ae6c8f31d | 2020-06-08T11:46:16 | 2020-06-08T11:46:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,688,023 | 11,484 | alesapin | 01079_parallel_alter_modify_zookeeper:
1. https://clickhouse-test-reports.s3.yandex.net/11458/ce4f528b7358cde7d2a3da7b620079f7f5d30348/functional_stateless_tests_(debug).html#fail1
most interesting moments in log:
```
2020.06.08 03:12:21.112047 [ 245 ] {} <Trace> test_fvg67y.concurrent_alter_mt_1: Executing log e... | 2020-06-08T15:05:00 | 2020-06-08T15:17:32 | {} | MEMBER |
ClickHouse | ClickHouse | 640,281,931 | 11,494 | Turum | Here it is,
```
default> insert into default.orders_history
SELECT *
FROM odbc('DSN=QExport;Uid=clickhouse;Pwd=clickhouse;BatchSize=10000;Timeout=0;','dbo', 'v_TodayOrders_ODBC')
[2020-06-08 00:21:29] Read timed out
```
| 2020-06-07T21:22:44 | 2020-06-07T21:22:44 | {} | NONE |
ClickHouse | ClickHouse | 640,297,624 | 11,400 | vitlibar | The following should work:
```
CREATE ROW POLICY filter1 ON default.table1 FOR SELECT USING account_id = 120 TO ALL
```
or
```
CREATE ROW POLICY filter1 ON default.table1 FOR SELECT USING account_id = 120 TO ALL EXCEPT admin;
CREATE ROW POLICY filter2 ON default.table1 FOR SELECT USING 1 TO admin;
```
... | 2020-06-07T23:43:44 | 2020-06-07T23:43:44 | {} | MEMBER |
ClickHouse | ClickHouse | 640,608,930 | 9,621 | ralphM78 | @den-crane What exactly do you need ? | 2020-06-08T13:38:49 | 2020-06-08T13:38:49 | {} | NONE |
ClickHouse | ClickHouse | 640,665,025 | 11,455 | KochetovNicolai | No backport because there are no affected releases. | 2020-06-08T14:33:05 | 2020-06-08T14:33:05 | {} | MEMBER |
ClickHouse | ClickHouse | 640,123,632 | 11,398 | den-crane | No info. No repro. Closing. | 2020-06-06T22:00:52 | 2020-06-06T22:01:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,359,060 | 11,458 | alexey-milovidov | I've noticed a bug when I edit some text with parentheses but I cannot reproduce it more... | 2020-06-08T04:28:03 | 2020-06-08T04:28:03 | {} | MEMBER |
ClickHouse | ClickHouse | 640,495,518 | 11,490 | millin | Sorry for misinformation.
According log it was not during ALTER query itself, but during auto-introspection after ALTER query.
Client - DataGrip/clickhouse-jdbc 0.2.4.
```
2020.06.07 11:17:48.799301 [ 544 ] {} <Trace> DynamicQueryHandler: Request URI: /?extremes=0&query_id=3a55e895-0716-4503-8baa-788ce119e7b8&data... | 2020-06-08T09:45:02 | 2020-06-08T09:45:02 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,122,003 | 11,391 | azat | >Do you plan to continue this PR anyway, or will postpone it?
Sure, I will continue (marked as draft to avoid attracting extra attention until it will be reworked/fixed)
Right now I have two options:
- just introduce options to make it backward compatible (simple)
- introduce proxy/impersonate users (as suggest... | 2020-06-06T21:43:58 | 2020-06-06T21:43:58 | {} | MEMBER |
ClickHouse | ClickHouse | 640,226,193 | 11,332 | den-crane | ```
--- ORDER BY (m_date,country)
CREATE MATERIALIZED VIEW default.Items_MV
ENGINE = AggregatingMergeTree() PARTITION BY m_date ORDER BY (m_date,country)
AS SELECT
m_date,
country,
countState() AS total_played
FROM
default.Items
GROUP BY
m_date,
country
┌─────m_date─┬─country─┬─countMerge(t... | 2020-06-07T14:21:11 | 2020-06-07T14:21:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,294,075 | 11,493 | alexey-milovidov | "Unbundled" build is broken accidentially by @abyss7
Already fixed in master. | 2020-06-07T23:12:45 | 2020-06-07T23:12:45 | {} | MEMBER |
ClickHouse | ClickHouse | 640,494,226 | 11,468 | asurovenko | Is there any way to recreate the table if there is small amount of empty disk place? My table size is 900Gb and I have aboun 100Gb free space. | 2020-06-08T09:42:35 | 2020-06-08T09:42:35 | {} | NONE |
ClickHouse | ClickHouse | 640,527,191 | 11,488 | 4ertus2 | Should `select *` work as `select * from system.one` or we need disable both `select *` and `select xyz.*`? | 2020-06-08T10:48:30 | 2020-06-08T10:48:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,602,868 | 11,331 | den-crane | >In your query, you have use argmax based on rto_count = 0,
You can use some fake default Null or 65536 (rto_count UInt32 default 65536)
But I agree that it's an ugly solution. A much better way is Replacingmergetree.
>With replacingmergetree and collapsingmergetree, problem
>I am facing for update queries. I... | 2020-06-08T13:28:18 | 2020-06-08T13:28:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,673,404 | 11,514 | xmariachi | Here, JSONAsString is applied to the full table. Is it possible to define a JSONEachRow format, yet define a field to be a JSON that should be parsed and set as a String (basically define field as JSONAsString where the JSON object will be serialized into a String). | 2020-06-08T14:41:54 | 2020-06-08T14:41:54 | {} | NONE |
ClickHouse | ClickHouse | 640,678,363 | 11,494 | den-crane | @Turum you did not show any error from CH with timeout exception.
CH errors starts with "DB:: ....".
like `DB::NetException: Timeout exceeded while reading from socket`
This error `DB::Exception: Cannot write to ostream at offset` actually means that a client has disappeared and CH server unable to send data (i... | 2020-06-08T14:50:29 | 2020-06-08T14:52:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,397,145 | 8,131 | filimonov | I think it's not actual anymore, since we've switched to replxx? | 2020-06-08T06:30:31 | 2020-06-08T06:30:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,423,193 | 11,223 | qoega | Something fixed that problem on master. Currently runs OK | 2020-06-08T07:34:55 | 2020-06-08T07:34:55 | {} | MEMBER |
ClickHouse | ClickHouse | 640,450,885 | 11,504 | dependabot-preview[bot] | Looks like certifi is up-to-date now, so this is no longer needed. | 2020-06-08T08:27:01 | 2020-06-08T08:27:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,590,505 | 11,494 | den-crane | tabbix has a default timeout setting 5 seconds. Try to change to 5000 under (gear) settings.
jdbc has a default timeout setting 30 seconds. Try to add JDBC parameter `socket_timeout = 5000` | 2020-06-08T13:05:55 | 2020-06-08T13:05:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 640,660,401 | 11,454 | KochetovNicolai | BTW, it's strange that pr was merged without review. | 2020-06-08T14:29:22 | 2020-06-08T14:29:22 | {} | MEMBER |
ClickHouse | ClickHouse | 644,962,242 | 11,668 | bharatnc | Can't seem to find the reason why `00092_obfuscator` fails consistently with all the tests. Any inputs? Cc: @alexey-milovidov @blinkov | 2020-06-16T19:16:32 | 2020-06-16T19:16:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 645,120,601 | 11,723 | byx313 | > > Our environment is centos6,so 19.9.5.36 is our newest version
>
> Are you sure? I think newer versions should work as well https://repo.yandex.ru/clickhouse/rpm/stable/x86_64/
I think we have made the mistake.
We get rpm from [https://packagecloud.io/Altinity/clickhouse](url),and in this page the newest for ... | 2020-06-17T03:10:42 | 2020-06-17T03:10:42 | {} | NONE |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.