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 | 520,964,725 | 5,230 | filimonov | @Johnny-Three
What version do you have?
can you help to create well reproducable test case? | 2019-08-13T18:59:24 | 2019-08-13T18:59:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,029,226 | 6,471 | 4ertus2 | What if set optimize_read_in_order=0 ? | 2019-08-13T22:13:35 | 2019-08-13T22:13:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,117,675 | 6,445 | gusevartemasd | No one used ODBC driver with non replicated tables?
| 2019-08-14T06:23:12 | 2019-08-14T06:23:12 | {} | NONE |
ClickHouse | ClickHouse | 521,173,547 | 6,488 | alex-zaitsev | There is no way to make a single table readonly in current ClickHouse. Flexible RBAC should be available later this year (Q4). | 2019-08-14T09:29:42 | 2019-08-14T09:29:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,197,276 | 6,488 | alex-zaitsev | There is no master node in ClickHouse. I do not understand what you are talking about. Replication consistency is controlled by ZooKeeper, that also checks checksums and so on. | 2019-08-14T10:49:06 | 2019-08-14T10:49:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,337,359 | 6,448 | alexey-milovidov | > But since no one remembers what is included in all and extra, I prefer to state warning switches explicitly. But it is my personal preference, of course.
Your call.
We can find them in gcc docs. (But it's Ok to leave some duplicates...) | 2019-08-14T17:18:10 | 2019-08-14T17:18:10 | {} | MEMBER |
ClickHouse | ClickHouse | 520,345,234 | 6,439 | alesapin | I will debug broken tests. In addition to enabling the setting, we have to take `granularity_bytes` from zookeeper when creating replicated tables. I'll make an additional pull request. | 2019-08-12T08:53:07 | 2019-08-12T08:53:07 | {} | MEMBER |
ClickHouse | ClickHouse | 520,373,106 | 6,012 | KochetovNicolai | Part with output formats seems ok.
Probably, it is worth to add general settings class `RowOutputFormatParams` the same as `RowInputFormatParams`. Then next time you won't need to change every constructor in order to add new setting. | 2019-08-12T10:31:16 | 2019-08-12T10:31:16 | {} | MEMBER |
ClickHouse | ClickHouse | 520,897,692 | 5,450 | tavplubix | Will be assigned to @CurtizJ | 2019-08-13T16:01:12 | 2019-08-13T16:01:12 | {} | MEMBER |
ClickHouse | ClickHouse | 521,094,271 | 5,230 | inolddays | @filimonov thanks for the reply. I use version 19.7.5.27 and it happens every day now in our production environment. here attached the SQL and the log. As we observed, it often happens when queries executed in concurrency scenario.I am not sure if the info is enough to make a judgment? if not, plz tell me how to help.... | 2019-08-14T04:08:35 | 2019-08-14T07:09:12 | {} | NONE |
ClickHouse | ClickHouse | 521,180,373 | 5,331 | BayoNet | Examples are good but they don't answer my questions. For example, can I pass only the constants to the query, or I can pass, for example, column name like `SELECT {col_name} from table`? | 2019-08-14T09:50:21 | 2019-08-14T09:50:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,329,769 | 6,484 | den-crane | >Distributed(cluster - 1
-1 ? It should be a cluster name.
For example `test_shard_localhost` : `ENGINE = Distributed(test_shard_localhost, default, dummy)`
check for available clusters `select distinct cluster from system.clusters`
| 2019-08-14T16:57:31 | 2019-08-14T16:58:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,781,309 | 6,464 | ghost | My cluster consists of 10 shards.
I tried several times to eliminate the bottleneck that occurs at the disk-read.
The distributed query references the 10 shards.
```
data_dist --+--> data1.test.com/data_local
|
+--> data2.test.com/data_local
|
+--> dat... | 2019-08-13T10:31:48 | 2019-08-13T10:31:48 | {} | NONE |
ClickHouse | ClickHouse | 520,923,677 | 6,363 | alexey-milovidov | We have decided to remove AsynchronousMetrics from metric_log. | 2019-08-13T17:12:04 | 2019-08-13T17:12:04 | {} | MEMBER |
ClickHouse | ClickHouse | 520,940,748 | 6,457 | bessonov87 | > If you use LZ4 you should upgrade all servers all together from 19.6 to 19.7+ (you can do it one by one and you'll get some number of errors `Checksums of parts don't `. But you need to update all as soon as possible.)
>
> If you just need to add more shards then install at new servers the same CH version as old s... | 2019-08-13T17:58:18 | 2019-08-13T17:58:35 | {} | NONE |
ClickHouse | ClickHouse | 521,368,550 | 6,457 | bessonov87 | > So these errors Timeout: connect timed out: 10.0.0.42:9009 Attempt to read after eof
are only problem now? No more issues? Is it right?
No. The problem is we can not add more replicas. Everything work perfect until we are trying to add new replica. Yesterday I tried again. I installed CH 19.11.5.28 (same version ... | 2019-08-14T18:42:23 | 2019-08-14T18:42:23 | {} | NONE |
ClickHouse | ClickHouse | 520,317,072 | 6,393 | Chocorean | Thanks a lot. Clickhouse is running again, I reduced the pool size. | 2019-08-12T07:02:01 | 2019-08-12T07:02:01 | {} | NONE |
ClickHouse | ClickHouse | 520,351,510 | 6,148 | blinkov | @zhang2014 so we just close this one? | 2019-08-12T09:15:06 | 2019-08-12T09:15:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,360,167 | 5,966 | alesapin | Do not release, need to fix one bug. | 2019-08-12T09:44:15 | 2019-08-12T09:44:15 | {} | MEMBER |
ClickHouse | ClickHouse | 520,704,281 | 6,168 | blinkov | @YiuRULE queue of PRs to be reviewed is pretty long at the moment, someone will get to it eventually | 2019-08-13T06:19:06 | 2019-08-13T06:19:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,732,594 | 6,454 | akonyaev90 | orc_proto.pb.cc ? | 2019-08-13T08:03:08 | 2019-08-13T08:03:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,898,075 | 5,387 | tavplubix | Will be assigned to @akuzm | 2019-08-13T16:02:00 | 2019-08-13T16:02:00 | {} | MEMBER |
ClickHouse | ClickHouse | 520,961,418 | 6,474 | den-crane | sudo dmesg|tail -300
check for OOM messages like
```
Jun 10 13:34:25 s-reserve kernel: [827998.184637] Killed process 24184 (clickhouse-serv)
total-vm:19133724kB, anon-rss:9965712kB, file-rss:0kB, shmem-rss:0kB
```
Do you use CODEC(Delta for Date columns? or columns with Int16 / Int8 ?
https://github.co... | 2019-08-13T18:50:50 | 2019-08-13T18:56:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,965,752 | 6,475 | alexey-milovidov | > ClickHouse build requires gcc-8 and g++-8
No, we have switched to gcc-9 about a few weeks ago. | 2019-08-13T19:02:20 | 2019-08-13T19:02:20 | {} | MEMBER |
ClickHouse | ClickHouse | 521,115,791 | 6,468 | avant1 | I see. Should a note be added to the docs? | 2019-08-14T06:14:52 | 2019-08-14T06:14:52 | {} | NONE |
ClickHouse | ClickHouse | 521,359,665 | 6,497 | InsidersByte | Actually, there is a bigger bug here than the import exporting erroring.
If the length after filtering out the defaults is valid then it will succeed to import the data but create the incorrect number of rows.
```sql
INSERT INTO foobar FORMAT Values
(['1', ''], [0, 1])
;
select *
from foobar
INTO OUTFI... | 2019-08-14T18:18:43 | 2019-08-14T18:18:43 | {} | NONE |
ClickHouse | ClickHouse | 520,973,726 | 6,477 | dimarub2000 | I am thinking about adding an inclusive version, and probably improving semantics with QuantileExactBase class. | 2019-08-13T19:26:19 | 2019-08-13T19:26:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,992,984 | 6,475 | kcking | Got it, thanks! | 2019-08-13T20:19:19 | 2019-08-13T20:19:19 | {} | NONE |
ClickHouse | ClickHouse | 521,004,131 | 6,366 | alesapin | Performance flap. | 2019-08-13T20:51:36 | 2019-08-13T20:51:36 | {} | MEMBER |
ClickHouse | ClickHouse | 520,292,775 | 5,438 | alexey-milovidov | Do you want to add an integration test? | 2019-08-12T03:56:53 | 2019-08-12T03:56:53 | {} | MEMBER |
ClickHouse | ClickHouse | 520,366,541 | 6,246 | CurtizJ | I think that 3 issues you reported have different causes.
Segfault from this issue at line
```
4. /usr/bin/clickhouse-server(DB::ExternalLoader::LoadingDispatcher::reloadOutdated()+0x23a) [0x675e71a]
```
has been already fixed in #6082.
Second one from #6244 at line
```
3. /usr/bin/clickhouse-server(DB::Sp... | 2019-08-12T10:06:12 | 2019-08-12T10:06:12 | {} | MEMBER |
ClickHouse | ClickHouse | 520,366,910 | 6,244 | CurtizJ | Fixed in #6404. | 2019-08-12T10:07:37 | 2019-08-12T10:07:37 | {} | MEMBER |
ClickHouse | ClickHouse | 520,392,504 | 6,430 | josepowera | Just thinking of how to narrow this... We have created some kind of pseudo POC for unit test https://github.com/yandex/ClickHouse/blob/master/dbms/tests/integration/test_storage_kafka/test.py (sorry, we work in java, doing our best here in text editor, not python ide) and haven't found any unit test with multiple MVs f... | 2019-08-12T11:47:10 | 2019-08-12T11:47:10 | {} | NONE |
ClickHouse | ClickHouse | 520,579,557 | 6,448 | kreuzerkrieg | I can go over and remove all those included in `extra` and `all`
But since no one remembers what is included in `all` and `extra`, I prefer to state warning switches explicitly. But it is my personal preference, of course.
Your call.
| 2019-08-12T20:20:58 | 2019-08-12T20:20:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,959,692 | 6,457 | den-crane | Check system.replication_queue
num_tries, last_exception
Probably some replicas now indefinitely are looking for already removed parts. It's fixable by partition attach/detach. But it should be done very accurate and need to understand what happened with every missing part using info from logs/part_log. | 2019-08-13T18:46:03 | 2019-08-13T18:46:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,027,274 | 6,457 | bessonov87 | We are now inserting about 1.7 million records per minute. Most of the data is inserted into ~30 replicated tables. Inserts are going in batches, every 2 seconds.
> Check system.replication_queue
system.replication_queue is almost always empty on all 4 replicas.
Also I ran this query:
`SELECT database, ... | 2019-08-13T22:05:56 | 2019-08-13T22:05:56 | {} | NONE |
ClickHouse | ClickHouse | 521,272,239 | 6,459 | alexey-milovidov | We don't have Mac OS build in our CI - it means that it is not continuously supported. This task is to add it. | 2019-08-14T14:35:09 | 2019-08-14T14:35:09 | {} | MEMBER |
ClickHouse | ClickHouse | 520,399,231 | 6,437 | 4ertus2 | There's ASOF JOIN that could help you. The idea is
1. make ```SELECT * FROM your_table AS t1 ASOF JOIN your_table AS t2 ON t1.T >= t2.T WHERE t2.Tag != ''```
2. then make resulting column depending on comparison of t1.T and t2.T: ```if(t1T - t2.T > _some_, t1.Tag, t2.Tag) AS resultTag```.
But there're restrictions... | 2019-08-12T12:11:19 | 2019-08-12T12:11:19 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,875,671 | 6,457 | den-crane | >19.11.5.28
Did you check CH versions at all replicas?
In 19.7 lz4 library was updated. Now 19.7+ is incompatible with 19.6- and produces `hash of uncompressed files doesn't match` | 2019-08-13T15:08:06 | 2019-08-13T15:08:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,927,947 | 6,471 | den-crane | ? Column types of
transfers_from value , transfer_from_bin
transfers_to value, transfer_to_bin
PARTITION BY / ORDER BY of: transfers_from, transfers_to
| 2019-08-13T17:24:08 | 2019-08-13T17:24:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,203,519 | 6,488 | astudnev | At least 2 use cases for this:
1) i do update of clickhouse software for one of server and i do not want that at any circumstance the new software can corrupt database. So i would configure the updated servers as read only. I already met with the situation when updating a server corrupted a table on existing table: ... | 2019-08-14T11:12:09 | 2019-08-14T11:12:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,272,643 | 6,488 | alex-zaitsev | What you describe is non-sense. If you stop the replica, your system becomes inconsistent. | 2019-08-14T14:36:06 | 2019-08-14T14:36:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,483,010 | 6,235 | blinkov | @BayoNet ping | 2019-08-12T15:51:15 | 2019-08-12T15:51:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,574,839 | 6,448 | alexey-milovidov | Ok, but we don't need to list warnings that are already present in `-Wextra`. | 2019-08-12T20:06:30 | 2019-08-12T20:06:30 | {} | MEMBER |
ClickHouse | ClickHouse | 520,576,865 | 6,446 | alexey-milovidov | Ok.
PS. For reference: http://sqlfiddle.com/#!9/0a28f18/361 | 2019-08-12T20:12:51 | 2019-08-12T20:12:51 | {} | MEMBER |
ClickHouse | ClickHouse | 520,676,327 | 6,168 | YiuRULE | Does we get some news about this PR on Yandex side ? | 2019-08-13T03:38:16 | 2019-08-13T03:38:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,829,488 | 5,925 | alex-krash | @alexey-milovidov , I've implemented everything in this function, as we discussed. Tried to make [test](https://github.com/yandex/ClickHouse/blob/ea9cf3a62f42f0ed48efe7baac2be46bfaa7ae5e/dbms/tests/queries/0_stateless/00957_neighbour.sql) as detailed, as possible. | 2019-08-13T13:14:49 | 2019-08-13T13:14:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,931,247 | 6,457 | den-crane | If you use LZ4 you should upgrade all servers all together from 19.6 to 19.7+ (you can do it one by one and you'll get some number of errors `Checksums of parts don't `. But you need to update all as soon as possible.)
If you just need to add more shards then install at new servers the same CH version as old servers... | 2019-08-13T17:32:53 | 2019-08-13T17:32:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,140,328 | 6,445 | alex-zaitsev | @gusevartemasd , ODBC driver is widely used. What you see does not seem to be related to the driver, but looks like your application. 'param_values_history' is your table, how do you insert into it from the driver? | 2019-08-14T07:47:44 | 2019-08-14T07:47:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,227,216 | 6,299 | CurtizJ | I disabled read in order optimization with joins as it works incorrectly now, because after join UnionBlockInputStream is called. | 2019-08-14T12:34:16 | 2019-08-14T12:34:16 | {} | MEMBER |
ClickHouse | ClickHouse | 521,228,606 | 6,413 | akuzm | From the first principles, the deletion should go as follows:
1) atomically mark the table as deleted in a single source of truth, Zookeeper in this case,
2) report that the deletion query has succeeded,
3) do whatever local garbage collection is needed, in a way resilient to restarts etc.
This is an often-used ... | 2019-08-14T12:38:43 | 2019-08-14T12:39:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,371,311 | 6,448 | alexey-milovidov | > if you are ok with these, get your guys ready for float comparison failures
We are Ok with exact (bitwise) float comparisons in most of cases.
If you don't see something suspicious, we can simply omit this warning.
| 2019-08-14T18:49:34 | 2019-08-14T18:49:34 | {} | MEMBER |
ClickHouse | ClickHouse | 520,314,317 | 5,230 | inolddays | hi, @filimonov is there any plan on this bug? we run into the same problem. have no idea when will it happen and how to resolve it. | 2019-08-12T06:47:45 | 2019-08-12T06:47:45 | {} | NONE |
ClickHouse | ClickHouse | 520,368,068 | 6,325 | kreuzerkrieg | > We need to add as many warnings as possible.
Here you go https://github.com/yandex/ClickHouse/pull/6448
| 2019-08-12T10:12:00 | 2019-08-12T10:12:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,802,165 | 6,105 | Imomoi | @alexey-milovidov On 19.11 it looks OK. No more crashes for now. (Around 2 weeks in release). | 2019-08-13T11:47:26 | 2019-08-13T11:47:26 | {} | NONE |
ClickHouse | ClickHouse | 520,917,975 | 6,470 | alexey-milovidov | Reference: http://sqlfiddle.com/#!9/0a28f18/363 | 2019-08-13T16:56:12 | 2019-08-13T16:56:12 | {} | MEMBER |
ClickHouse | ClickHouse | 520,990,340 | 6,474 | iameugenejo | The raw core report has the following section:
```
ProcStatus:
Name: clickhouse-serv
Umask: 0027
State: D (disk sleep)
```
| 2019-08-13T20:11:39 | 2019-08-13T20:12:07 | {} | NONE |
ClickHouse | ClickHouse | 521,016,444 | 6,437 | Eiilin-Maur | Thanks for the advice.
Tried to implement first part. It turns out the following:
| 2019-08-13T21:29:04 | 2019-08-13T21:29:04 | {} | NONE |
ClickHouse | ClickHouse | 520,400,900 | 6,437 | 4ertus2 | Another idea you may try is to round your T column to 2 minutes intervals and JOIN table to itself with ANY LEFT JOIN on rounded T values. It would be faster then ASOF JOIN variant, but would contain some rounding error. | 2019-08-12T12:16:58 | 2019-08-12T12:16:58 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,459,498 | 6,439 | alesapin | Fix for https://github.com/yandex/ClickHouse/issues/6394: https://github.com/yandex/ClickHouse/pull/6452 | 2019-08-12T14:55:21 | 2019-08-12T14:55:40 | {} | MEMBER |
ClickHouse | ClickHouse | 520,728,956 | 6,454 | akonyaev90 | Is it possible to generate protobuf on the fly? - Alexey, what file are you talking about? | 2019-08-13T07:51:22 | 2019-08-13T07:51:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,885,877 | 6,457 | bessonov87 | > > 19.11.5.28
>
> Did you check CH versions at all replicas?
> In 19.7 lz4 library was updated. Now 19.7+ is incompatible with 19.6- and produces `hash of uncompressed files doesn't match`
@den-crane Thanks for the answer.
If that's the cause of the issue, what is the proper way to upgrade to 19.7+?
Or should... | 2019-08-13T15:32:14 | 2019-08-13T15:33:00 | {} | NONE |
ClickHouse | ClickHouse | 521,015,563 | 6,474 | iameugenejo | We updated to this new version a little less than 3 months ago after using 18.16.1 for a while. This has never happened before, and it's been happening for the entire duration of time we've been using this new version.
Should I attempt to upgrade the clickhouse to the latest release, or should I downgrade to 18.16.1... | 2019-08-13T21:26:17 | 2019-08-13T21:26:17 | {} | NONE |
ClickHouse | ClickHouse | 521,170,095 | 6,488 | astudnev | What you mean profile? It is about just a table, not user... or tables have profiles? | 2019-08-14T09:19:36 | 2019-08-14T09:19:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,668,981 | 6,148 | zhang2014 | @blinkov Yes, because its contents have been merged : ) | 2019-08-13T02:55:54 | 2019-08-13T02:55:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,875,171 | 6,462 | simPod | It’s altered but like month ago. Still had to delete them manually. So I created the table from scratch with TTL set on creation. It’s not dropping old parts anyway 🤔
> On Aug 13, 2019, at 5:05 PM, Denis Zhuravlev <notifications@github.com> wrote:
>
> Is it new table or altered (to add TTL) ?
>
> TTL works only for... | 2019-08-13T15:06:54 | 2019-08-13T15:06:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,917,801 | 6,470 | alexey-milovidov | > Added alias user() for compatibility with MySQL (I supposed this is important for us)
Yes. And we can make it case-insensitive. | 2019-08-13T16:55:45 | 2019-08-13T16:55:45 | {} | MEMBER |
ClickHouse | ClickHouse | 520,984,887 | 6,468 | alexey-milovidov | This is known issue but it cannot be fixed in zero-cost fashion. Most likely won't fix. | 2019-08-13T19:56:52 | 2019-08-13T19:56:52 | {} | MEMBER |
ClickHouse | ClickHouse | 521,014,247 | 6,480 | den-crane | note for CH devs: 19.11.2.7 works under WSL. | 2019-08-13T21:21:56 | 2019-08-13T21:22:27 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,211,841 | 6,472 | astudnev | > Have you recreated this table on new version?
no. I downgraded to 19.9 and recreated the table to fix it | 2019-08-14T11:42:24 | 2019-08-14T11:42:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,240,394 | 6,459 | weiqxu | the newest code can't build in MAC OS . Will it support MAC any more? | 2019-08-14T13:14:11 | 2019-08-14T13:14:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,635,908 | 6,323 | alexey-milovidov | This is totally Ok. It only looks strange. | 2019-08-12T23:50:16 | 2019-08-12T23:50:16 | {} | MEMBER |
ClickHouse | ClickHouse | 521,025,880 | 6,437 | 4ertus2 | 1. You need newer version for ASOF JOIN with ON syntax. Older ones support USING syntax only.
2. ASOF needs at least one equality for now. You need ```t1.some = t2.some AND t1.Activity_Start >= t2.Activity_Start``` or ```using(some, Activiti_Start)``` | 2019-08-13T22:00:53 | 2019-08-13T22:02:07 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,280,895 | 6,372 | alexey-milovidov | That's great! Please keep in touch. | 2019-08-12T01:51:21 | 2019-08-12T01:51:21 | {} | MEMBER |
ClickHouse | ClickHouse | 520,345,195 | 5,587 | vichoho | This issue still happened and CH always tried to merge the abnormal part and caused the merge process never stopped.
Does anyone know the root cause? | 2019-08-12T08:52:57 | 2019-08-12T08:52:57 | {} | NONE |
ClickHouse | ClickHouse | 520,845,672 | 6,454 | akonyaev90 | now orc_proto building with proto-command (I added command in arrow-cmake cmakeFile)
typos corrected | 2019-08-13T13:57:23 | 2019-08-13T13:57:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,174,980 | 5,331 | alex-zaitsev | @BayoNet
Look at examples:
https://github.com/yandex/ClickHouse/blob/master/dbms/tests/queries/0_stateless/00954_client_prepared_statements.sh
https://github.com/yandex/ClickHouse/blob/master/dbms/tests/queries/0_stateless/00956_http_prepared_statements.sh | 2019-08-14T09:33:50 | 2019-08-14T09:33:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,178,256 | 6,488 | astudnev | Problem is NOT in user access. The problem is that clickhouse nodes read data for replication from arbitrary remote nodes, and sometimes it breaks the original table. I want to prevent master node to read anything from untrusted nodes. This is not user level security, it is security of server | 2019-08-14T09:43:48 | 2019-08-14T09:43:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,208,015 | 6,488 | astudnev | The name is confusing. This is not read-only this is write-only | 2019-08-14T11:28:37 | 2019-08-14T11:28:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,211,487 | 6,472 | alesapin | Have you recreated this table on new version? | 2019-08-14T11:41:07 | 2019-08-14T11:41:07 | {} | MEMBER |
ClickHouse | ClickHouse | 521,333,998 | 1,522 | doom369 | Getting the same error:
```
2019.08.14 16:31:16.739119 [ 43 ] {} <Error> ServerErrorHandler: Code: 101, e.displayText() = DB::NetException: Unexpected packet from client, Stack trace:
0. /usr/bin/clickhouse-server(StackTrace::StackTrace()+0x30) [0x7f58f30]
1. /usr/bin/clickhouse-server(DB::Exception::Exception(... | 2019-08-14T17:08:57 | 2019-08-14T17:08:57 | {} | NONE |
ClickHouse | ClickHouse | 521,341,075 | 6,448 | kreuzerkrieg | > > But since no one remembers what is included in all and extra, I prefer to state warning switches explicitly. But it is my personal preference, of course.
> > Your call.
>
> We can find them in gcc docs. (But it's Ok to leave some duplicates...)
just removed all redundant stuff
if you are ok with these, get ... | 2019-08-14T17:27:37 | 2019-08-14T17:27:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,767,236 | 5,818 | millin | @abyss7 I think it is easy to fix. Can you check? | 2019-08-13T09:45:31 | 2019-08-13T09:45:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,931,393 | 6,471 | astudnev | CREATE TABLE IF NOT EXISTS ethereum.transfers_from_storage ON CLUSTER all_sharded
(
tx_date Date,
tx_time DateTime,
blockchain_id UInt32,
block UInt32,
tx_hash_bin String,
tx_from_bin String,
tx_to_bin String,
transfer_from_bin String,
transfer_to_bin String,
currency_id UInt32,
... | 2019-08-13T17:33:16 | 2019-08-13T17:33:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,006,673 | 5,885 | LukeCzmoch | Hi,
Sorry for the late response but I’m on vacation right now. It was a part of the equation to calculate those percentiles or simply I made a mistake but I don’t suppose.
I believe this could give you an idea what I meant:
http://www.utdallas.edu/~serfling/3332/ComputingPercentiles.pdf
Regards,
Luke
Fro... | 2019-08-13T20:58:54 | 2019-08-13T20:58:54 | {} | NONE |
ClickHouse | ClickHouse | 521,122,892 | 6,471 | astudnev | > What if set optimize_read_in_order=0 ?
I already downgraded the servers, can not test. | 2019-08-14T06:45:31 | 2019-08-14T06:45:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,153,460 | 6,445 | gusevartemasd | > @gusevartemasd , ODBC driver is widely used. What you see does not seem to be related to the driver, but looks like your application. 'param_values_history' is your table, how do you insert into it from the driver?
My collector send data to ODBC clickhouse driver and driver send it to clickhouse-server( I do not... | 2019-08-14T08:29:34 | 2019-08-14T10:48:19 | {} | NONE |
ClickHouse | ClickHouse | 520,478,969 | 2,764 | 4ertus2 | It's possible to use IN for SEMI JOIN and NOT IN for ANTI JOIN now. But they have restrictions on Nullable types. Right SQL emulation is via EXISTS / NOT EXISTS. NOT EXISTS is not equal to NOT IN for Nullables.
It's also possible to emulate ANTI JOIN with ```LEFT JOIN ... WHERE x IS NULL```. | 2019-08-12T15:41:49 | 2019-08-12T15:55:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,565,847 | 6,454 | alexey-milovidov | Is it possible to generate protobuf on the fly? (auto-generated file does not look good). | 2019-08-12T19:39:12 | 2019-08-12T19:39:12 | {} | MEMBER |
ClickHouse | ClickHouse | 520,804,371 | 6,105 | alexey-milovidov | Thank you! | 2019-08-13T11:55:30 | 2019-08-13T11:55:30 | {} | MEMBER |
ClickHouse | ClickHouse | 521,044,431 | 6,457 | den-crane | So these errors `Timeout: connect timed out: 10.0.0.42:9009` `Attempt to read after eof`
are only problem now? No more issues? Is it right?
It could be real 1s timeout. Try
```
cat /etc/clickhouse-server/users.d/profile_default.xml
<yandex>
<profiles>
<default>
<http_connection_timeo... | 2019-08-13T23:18:11 | 2019-08-13T23:18:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,214,295 | 6,466 | alexey-milovidov | It's not ready. | 2019-08-14T11:51:01 | 2019-08-14T11:51:01 | {} | MEMBER |
ClickHouse | ClickHouse | 521,324,668 | 1,534 | doom369 | @blinkov getting the same error, few hundreds of records in 1 batch once per minute:
```
2019.08.14 16:31:16.739119 [ 43 ] {} <Error> ServerErrorHandler: Code: 101, e.displayText() = DB::NetException: Unexpected packet from client, Stack trace:
0. /usr/bin/clickhouse-server(StackTrace::StackTrace()+0x30) [0x7f58... | 2019-08-14T16:43:46 | 2019-08-14T16:59:03 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 520,874,174 | 6,462 | den-crane | Is it new table or altered (to add TTL) ?
TTL works only for new parts after alter. | 2019-08-13T15:04:31 | 2019-08-13T15:04:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 520,892,307 | 5,097 | tavplubix | Will be assigned to @Akazz | 2019-08-13T15:48:14 | 2019-08-13T15:48:14 | {} | MEMBER |
ClickHouse | ClickHouse | 520,976,104 | 6,474 | iameugenejo | I see this:
```
[Mon May 20 21:17:52 2019] Out of memory: Kill process 7820 (clickhouse-serv) score 975 or sacrifice child
[Mon May 20 21:17:52 2019] Killed process 7820 (clickhouse-serv) total-vm:392840976kB, anon-rss:127940644kB, file-rss:0kB, shmem-rss:0kB
```
but this isn't related. The crash happened this m... | 2019-08-13T19:33:02 | 2019-08-13T19:57:43 | {} | NONE |
ClickHouse | ClickHouse | 521,169,681 | 6,488 | alex-zaitsev | Use readonly profile. | 2019-08-14T09:18:21 | 2019-08-14T09:18:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,172,123 | 5,331 | BayoNet | @sanych73 hello.
I'm trying to add the documentation for this PR, and I need your help. Can you describe the syntax of format string for a query? Where the placeholders can be? What are the rules for placeholders naming? What kinds of placeholders can be. And so on. | 2019-08-14T09:25:23 | 2019-08-14T09:25:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 521,332,424 | 6,486 | alexey-milovidov | What exactly does it fix? | 2019-08-14T17:04:36 | 2019-08-14T17:04:36 | {} | MEMBER |
ClickHouse | ClickHouse | 508,123,560 | 5,850 | alexey-milovidov | Could you please also add a test? | 2019-07-03T14:42:55 | 2019-07-03T14:42:55 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.