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 | 463,599,204 | 4,342 | nezed | Having the same issue on 19.3.3
Currently you can use `CAST(a, 'String')` as workaround
```sql
SELECT a FROM t1 UNION ALL SELECT CAST(a, 'String') FROM t2;
``` | 2019-02-14T11:50:07 | 2019-02-14T11:50:07 | {
"heart": 3,
"hooray": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,015,000 | 4,212 | alesapin | Abort here https://pastebin.com/TZTyMCcP. | 2019-02-15T11:38:32 | 2019-02-15T11:38:32 | {} | MEMBER |
ClickHouse | ClickHouse | 464,048,863 | 11 | alexey-milovidov | > First we will try to add dynamic code generation for compositions of simple functions.
FYI, dynamic code generation has been implemented. | 2019-02-15T13:22:21 | 2019-02-15T13:22:21 | {
"+1": 5
} | MEMBER |
ClickHouse | ClickHouse | 464,171,883 | 4,402 | alexey-milovidov | Reproduced. | 2019-02-15T19:33:13 | 2019-02-15T19:33:13 | {} | MEMBER |
ClickHouse | ClickHouse | 464,329,721 | 4,411 | kkrakovych | Hi @filimonov ,
Thank you for the link. Here is a workaround for Decimal values:
```
select val_decimal
, case when val_decimal is null then 1 else val_decimal end as weird_value
, multiIf(val_decimal is null, null, 100 / weird_value) as result
from test_table;
``` | 2019-02-16T10:04:55 | 2019-02-16T10:04:55 | {
"+1": 1,
"heart": 1,
"hooray": 1
} | NONE |
ClickHouse | ClickHouse | 465,229,601 | 4,442 | josepowera | We have seen similar kafka problem for "CSV" format after moving from 18.16.1 to 19.3.4
```
2019.02.19 00:58:12.444114 [ 33 ] {} <Error> void DB::StorageKafka::streamThread(): Code: 27, e.displayText() = DB::Exception: Cannot parse input: expected , before: \0: (at row 2)
Could not print diagnostic info because ... | 2019-02-19T17:26:16 | 2019-02-19T17:26:16 | {} | NONE |
ClickHouse | ClickHouse | 463,571,236 | 4,391 | blinkov | Basically initial query comes to one server and it becomes responsible for communicating with other servers to provide the requested result. Partial aggregation is done on (by default) one replica of each shard and then this initial server merges these partial results to produce the final result.
Do you have any mor... | 2019-02-14T10:14:01 | 2019-02-14T10:14:01 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,868,637 | 4,391 | ghost | Thank you for your answer!
I have more specific questions. I wonder the details of distributed queries.
When I send the query `select count(id) from users_dist group by id`, the initial server send the query `select count(id) from users group by id` to the other servers.
What are the partial results of the oth... | 2019-02-15T01:16:32 | 2019-02-15T01:16:32 | {} | NONE |
ClickHouse | ClickHouse | 463,954,510 | 4,400 | mkabilov | this PR has nothing to do with the deprecated create table syntax.
There's a reference to `columns` parameter ( https://github.com/yandex/ClickHouse/blame/master/docs/en/operations/table_engines/summingmergetree.md#L27 ), which is not present in the table create syntax ( https://github.com/yandex/ClickHouse/blame/ma... | 2019-02-15T08:41:47 | 2019-02-15T08:41:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,955,444 | 4,400 | blinkov | oh, ok | 2019-02-15T08:45:12 | 2019-02-15T08:45:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,966,339 | 4,406 | nezed | Can you please provide more info about how you are starting container and which docker network configuration is used? | 2019-02-15T09:22:50 | 2019-02-15T09:22:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,157,952 | 4,282 | kmatt | re https://github.com/yandex/ClickHouse/issues/4282#issuecomment-464122556
Dropping of Kafka engine tables is also working as normal on v19.3.3 | 2019-02-15T18:49:51 | 2019-02-15T18:49:51 | {} | NONE |
ClickHouse | ClickHouse | 464,180,559 | 4,393 | arctica | @simPod since the old sorting key is a prefix of the new sorting key, and there being no data in the just added column, the data at the moment of table modification is both sorted by the old and the new sorting key and hence wont need any modification.
I also needed a moment to understand the documention in this reg... | 2019-02-15T20:00:42 | 2019-02-15T20:00:42 | {} | NONE |
ClickHouse | ClickHouse | 464,359,758 | 4,419 | arctica | Actually I think the proposed feature would also obsolete #4413 because one could have a table with just a few columns in the primary/sorting key and use the AGGREGATE BY query to sort the rows however one desires. And because the columns are not part of the primary/sort key, one is free to add or remove columns as one... | 2019-02-16T16:19:49 | 2019-02-16T16:19:49 | {} | NONE |
ClickHouse | ClickHouse | 463,282,889 | 3,852 | c3mb0 | Just to chime in on the subject, we have a similar problem that could be improved through this functionality. Running conditional queries on rapidly mutating rows is kind of a hassle at the moment. As far as I understand there are a few options (consider a `ReplacingMergeTree` with `id` as pk):
1 - Use HAVING - `SEL... | 2019-02-13T17:07:59 | 2019-02-13T18:14:18 | {} | NONE |
ClickHouse | ClickHouse | 463,571,313 | 4,362 | andatt | Using the 18.10.3 docker container from https://hub.docker.com/r/yandex/clickhouse-server/tags
this is what I get:
```
(srf_framework_venv) andrew@andrew-Inspiron-7560:~/$ docker exec -it -u root some-clickhouse-server_18103 bash
root@569ff78b093b:/# clickhouse --client
ClickHouse client version 18.10.3.
Conn... | 2019-02-14T10:14:13 | 2019-02-14T10:15:55 | {} | NONE |
ClickHouse | ClickHouse | 464,332,347 | 4,413 | arctica | @den-crane Sorry but I don't see how the primary key helps with the usecase of dropping columns that are not the last one in the table. Your example just modifies the last column which is fine with ORDER BY restrictions. Try dropping the country column in your example.
> Code: 44. DB::Exception: Received from localh... | 2019-02-16T10:19:38 | 2019-02-16T10:19:38 | {} | NONE |
ClickHouse | ClickHouse | 464,774,336 | 4,431 | filimonov | What exactly do you mean by "parser brakes"? Can you provide some sample? | 2019-02-18T15:26:52 | 2019-02-18T15:26:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 465,108,151 | 4,443 | filimonov | What do you mean by 'unable to use'? (Related to #4421)
```
SELECT toDateTime(now(), 'Etc/GMT+5')
┌─toDateTime(now(), 'Etc/GMT+5')─┐
│ 2019-02-19 07:19:14 │
└────────────────────────────────┘
1 rows in set. Elapsed: 0.026 sec.
SELECT toDateTime(now(), 'Etc/GMT-5')
┌─toDateTime(now(), 'Etc... | 2019-02-19T12:21:33 | 2019-02-19T12:21:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,650,806 | 4,399 | alesapin | Maybe we can add integration test? Fill one replica queue with trash, and after that test `is_lost` logic. | 2019-02-14T14:39:44 | 2019-02-14T14:40:08 | {} | MEMBER |
ClickHouse | ClickHouse | 463,974,764 | 3,388 | lapkofear | It's pretty tricky issue due to absence of any errors except for logs like:
<table_name> (Replicated OutputStream): Block with like: <block_id> already exists; ignoring it. | 2019-02-15T09:51:02 | 2019-02-15T09:51:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,388,948 | 4,420 | filimonov | We also have similar case with base64 encoded string. In some cases the compression rate for base64-encoded string is even worse than 3/4 which can be gained by simple storing as binary.
Some simple test (lz4 default compression)
1 mln rows of strings in base64:
data_compressed_bytes: 11807899
data_uncompress... | 2019-02-16T21:45:53 | 2019-02-17T20:11:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,654,294 | 2,418 | KochetovNicolai | > It appears to be going through the columns in order and requiring that the first ones are all the numeric columns, then checks the remaining ones are categorical.
You're right. When you call `modelEvaluate` function for CatBoost model, you should firstly specify numeric features, than categorical (even if you used... | 2019-02-18T09:28:14 | 2019-02-18T09:28:14 | {} | MEMBER |
ClickHouse | ClickHouse | 464,800,979 | 4,414 | alesapin | > What exactly is a granule? Is it a row?
Granule is a batch of rows of fixed size which addresses with primary key. Term make sense only for MergeTree* engine family. It can be set with setting `index_granularity=N`, default value is 8192 rows per batch. So if you use default value, you will have index per each 819... | 2019-02-18T16:32:26 | 2019-02-18T16:32:26 | {} | MEMBER |
ClickHouse | ClickHouse | 465,164,709 | 4,423 | den-crane | Do you use Docker ?
Do you use LowCardinality? | 2019-02-19T15:02:05 | 2019-02-19T15:02:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,315,692 | 3,497 | Jack012a | which version of the release includes this bug fix?
I am not able to find the update information through
https://github.com/yandex/ClickHouse/blob/master/CHANGELOG.md.
Thank you
On Tue, Feb 12, 2019 at 6:06 PM alexey-milovidov <notifications@github.com>
wrote:
> The "File not found" exception at server startup is fi... | 2019-02-13T18:39:09 | 2019-02-13T18:39:09 | {} | NONE |
ClickHouse | ClickHouse | 463,328,379 | 3,497 | alexey-milovidov | It's in the upcoming release 19.3.3. We're preparing changelog right now. Stay tuned! | 2019-02-13T19:13:47 | 2019-02-13T19:13:47 | {} | MEMBER |
ClickHouse | ClickHouse | 463,946,856 | 4,400 | blinkov | Deprecated create table syntax is described separately below recommended one. | 2019-02-15T08:12:56 | 2019-02-15T08:12:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,019,665 | 4,406 | filimonov | That error just means that ClickHouse can't bind to IPv6 address, as it's disabled, but Clickhouse should work properly using IPv4 only.
If you want to get rid of that error - you should adjust that file in docker:
[/etc/clickhouse-server/config.d/docker_related_config.xml](https://github.com/yandex/ClickHouse/blob... | 2019-02-15T11:51:31 | 2019-02-15T11:51:31 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 465,081,618 | 4,431 | geldot | We've been seeing this with all releases after 19.1 (so we're stuck pinned to that release until fixed). Using the HTTP interface and TSV import here.
Syntax errors arise at arbitrary points in the stream after the first few hundred rows with no Buffer, straight insert to MergeTree. | 2019-02-19T10:45:44 | 2019-02-19T10:46:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,632,869 | 4,392 | 4ertus2 | We need queries leads to crash to fix it. Could you attach server debug logs with queries that executes just before crash? It's a query analyze phase so we need logs of last few moments of server's activity.
As workaround you may disable push down predicate 'enable_optimize_predicate_expression = 0'. | 2019-02-14T13:48:40 | 2019-02-14T13:48:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,361,303 | 4,420 | arctica | A compression codec would automatically uncompress the binary data into hex before computation can run on it, that would be inefficient in terms of computation (but save disk space).
A data type that automatically unhex()s a string into binary would make more sense (similar to how DateTime takes a string and encodes... | 2019-02-16T16:39:54 | 2019-02-16T16:39:54 | {} | NONE |
ClickHouse | ClickHouse | 464,646,716 | 3,605 | TH-HA | Hello
I just want to share a few thougths on this subject as i 'm looking to implement the same thing.
I can imagine 3 evolution to reach this goal :
1 : ALTER PARTITION MOVE PART TO [NODE.]PATH
Allow to move a partition to a different path . Of course move is safe : ie. Copy the file ( let the old one s... | 2019-02-18T09:06:48 | 2019-02-18T09:06:48 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 464,786,609 | 4,434 | proller | https://github.com/yandex/ClickHouse/pull/4438 | 2019-02-18T15:55:27 | 2019-02-18T15:55:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 465,061,766 | 4,040 | perez1987 | +1 | 2019-02-19T09:47:20 | 2019-02-19T09:47:20 | {} | NONE |
ClickHouse | ClickHouse | 465,310,098 | 3,759 | proller | https://github.com/yandex/ClickHouse/commit/e20c77e4c6f9d8326c387fd0a2833ed959bf1658 | 2019-02-19T21:07:15 | 2019-02-19T21:07:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,623,903 | 4,390 | alexey-milovidov | Yes, we are preparing to add secondary indices in a table.
> Additional context
I've found workarround
```
CREATE TABLE `some` (date Date, `index` UInt16) ENGINE = Memory()
```
This (backticks around identifier name) is the proper workaround.
> Probably I've missed new and unannounced ability to create cus... | 2019-02-14T13:19:41 | 2019-02-14T13:19:41 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 463,669,767 | 4,362 | nezed | Was fixed somewhere between `18.12.13` and `18.14.13`
So the only solution is to upgrade
| 2019-02-14T15:29:12 | 2019-02-14T15:29:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,724,688 | 4,393 | nezed | [Check this note in docs](https://clickhouse.yandex/docs/en/query_language/alter/#manipulations-with-key-expressions)
And have a closer look at this test:
https://github.com/yandex/ClickHouse/blob/cec49357da502c511e006ea8b5d8fbffeb521478/dbms/tests/queries/0_stateless/00754_alter_modify_order_by.sql#L11-L25
Espe... | 2019-02-14T17:52:35 | 2019-02-14T17:52:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,020,080 | 3,605 | miha-g | We have a very similar scenario as @thinksinside-LQ has described. We would greatly benefit if "older" partitions of a table could be stored on a different location (HDD partition). Currently we use Postgres and just move "old" table partitions to different tablespace, which is located on HDD partition.
Is there a r... | 2019-02-15T11:52:37 | 2019-02-15T11:52:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,242,609 | 4,413 | den-crane | Maybe you've reinvented primary key
https://clickhouse.yandex/docs/en/operations/table_engines/mergetree/#selecting-the-primary-key
```
CREATE TABLE metrics
(
date Date,
country String,
domain String,
browser String,
hits UInt64,
bandwidth UInt64
)
ENGINE = MergeTree
PRIMARY KEY ... | 2019-02-15T23:17:13 | 2019-02-15T23:19:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,453,978 | 4,391 | ghost | In the document,
> For example, for a query with GROUP BY, data will be aggregated on remote servers, and the intermediate states of aggregate functions will be sent to the requestor server. Then data will be further aggregated.
> An intermediate state of the aggregation (for uniq, this is the hash table for calcu... | 2019-02-17T12:58:37 | 2019-02-17T13:25:38 | {} | NONE |
ClickHouse | ClickHouse | 465,080,854 | 4,414 | arctica | Thank you for the explanation. Maybe a small piece of text could be added to the documentation like "(a granule is one block of primary key containing `index_granularity` rows)?
I see now how this index can be properly used. It only makes sense when the value being filtered for is very sparse or one needs very fine ... | 2019-02-19T10:43:11 | 2019-02-19T10:43:11 | {} | NONE |
ClickHouse | ClickHouse | 465,273,314 | 4,446 | alexey-milovidov | 00838_system_tables_drop_table_race - TODO | 2019-02-19T19:21:07 | 2019-02-19T19:21:07 | {} | MEMBER |
ClickHouse | ClickHouse | 463,603,091 | 4,342 | alex-zaitsev | Sure cast works. One way or another.
It is also interesting that behaviour of arrays is not consistent here. I.e. element of Array(LowCardinality(String)) is String. array of LowCardinality(String) is array of String and so on.
Common supertype for String and LowCardinality(String) should help.
```
SELECT
... | 2019-02-14T12:04:41 | 2019-02-14T12:04:55 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,636,336 | 4,395 | BayoNet | Лейбл не ставил специально, поскольку русская документация заведомо не соберётся. Лейбл поставлю на следующий PR, когда волью перевод. | 2019-02-14T13:58:14 | 2019-02-14T13:58:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,695,283 | 4,398 | alesapin | `00838_system_tables_drop_table_race` seems like flap. | 2019-02-14T16:32:26 | 2019-02-14T16:32:26 | {} | MEMBER |
ClickHouse | ClickHouse | 463,727,953 | 4,393 | simPod | Good to know! This should be IMO added to the docs in some form.
But still, this sentence is wrong I think, it makes no sense:
> data parts need not be changed (they remain sorted by the new sorting key expression).
Shouldn't it be like this?
```diff
- data parts need not be changed (they remain sorted by th... | 2019-02-14T18:01:30 | 2019-02-14T18:01:30 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,101,930 | 4,411 | filimonov | Please check that: #1562 #2385 #1768 #1816 | 2019-02-15T16:00:09 | 2019-02-15T16:01:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,156,685 | 4,282 | kmatt | Initial testing with v19.3.3 (54415) is allowing partition drops to occur as they did with v18. | 2019-02-15T18:45:36 | 2019-02-15T18:45:36 | {} | NONE |
ClickHouse | ClickHouse | 463,646,456 | 4,401 | 4ertus2 | #4392 | 2019-02-14T14:27:06 | 2019-02-14T14:27:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,671,121 | 4,362 | andatt | Thanks. That explains it! | 2019-02-14T15:32:33 | 2019-02-14T15:32:33 | {} | NONE |
ClickHouse | ClickHouse | 464,302,559 | 4,406 | filimonov | PS. With `listen_try` it rather should not be reported as Error. | 2019-02-16T07:06:32 | 2019-02-16T07:06:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,830,336 | 3,605 | miha-g | @filimonov Thank you for the clarification and examples. I applied your suggested approach to our proof of concept and it works great!
From the selecting perspective this is just as good as the future solution is expected to be.
From the inserting perspective, there is some extra work needed to take care of insertion... | 2019-02-18T18:04:15 | 2019-02-18T18:04:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,834,856 | 4,428 | den-crane | So you have 400 servers in your cluster, divide them into 20 sub-clusters.
Servers 1..20 will be members of sub_cluster_l1 and members of super-cluster cluster (with all 400 servers) the same time.
Put data of clients with names A.. into cluster_l1 and clients with names Z.. into sub_cluster_l20
Create 21 distribu... | 2019-02-18T18:22:08 | 2019-02-18T18:22:45 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 465,096,863 | 4,431 | geldot | Possibly related to #4442 in the Kafka engine, pointing toward a bug introduced in lower level streaming classes | 2019-02-19T11:39:44 | 2019-02-19T11:39:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 465,204,400 | 4,445 | alexey-milovidov | Docs check - known issue
Autotests - obsolete Jenkins tests.
Performance test - unrelated. | 2019-02-19T16:26:44 | 2019-02-19T16:32:32 | {} | MEMBER |
ClickHouse | ClickHouse | 463,294,220 | 3,852 | bgranvea | > Do you mean a new data type "SimpleAggregateFunction"?
yes, sorry I was not very clear. A new data type very similar to AggregateFunction. | 2019-02-13T17:39:32 | 2019-02-13T17:39:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,375,925 | 4,359 | alexey-milovidov | @banyrule Да, этот способ был бы более логичным и вызывал бы меньше вопросов. Буду рад, если у нас получится это когда-нибудь исправить! | 2019-02-13T21:19:19 | 2019-02-13T21:19:19 | {} | MEMBER |
ClickHouse | ClickHouse | 464,184,134 | 4,350 | arctica | I am also running into this problem because I am implementing replication and sharding outside of ClickHouse as we don't want to rely on Zookeeper. To read data, we need to use the remote() table function in order to aggregate data from several shards. For security purposes, it would be great if the user could run only... | 2019-02-15T20:12:53 | 2019-02-15T20:12:53 | {} | NONE |
ClickHouse | ClickHouse | 465,112,138 | 4,435 | filimonov | Currently it's not intuitive that you can't count rows with rowNumberInBlock in case of array join.
Probably it should be either changed, either documented.
```
SELECT rowNumberInBlock()
FROM system.one
ARRAY JOIN [1, 2, 3] AS x
┌─rowNumberInBlock()─┐
│ 0 │
│ 0 │
│ ... | 2019-02-19T12:35:39 | 2019-02-19T12:36:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,502,045 | 4,421 | filimonov | 1) Most probably it's not the best idea, as in that case you need to care of daylight saving and historical changes manually
2) But if you really need it and understand all possible complications, tzdata allows you to reference those absolute offsets like that:
```select toDateTime(now(), 'Etc/GMT+8');```
3) please... | 2019-02-17T20:03:11 | 2019-02-17T20:05:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,631,605 | 3,605 | miha-g | @filimonov if I understand you correctly, this would leave us with 2 separate databases and 2 separate tables, each in its own database. So in order to use this solution, we would also have to make a `UNION ALL` between the 2 tables in different databases in order to get the whole dataset. Please correct me, if I misun... | 2019-02-18T08:21:10 | 2019-02-18T08:21:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,832,074 | 3,421 | ztlpn | Curiously, since 18.16 the problem, while still present, is much less frequent and in my tests manifests itself only under asan. | 2019-02-18T18:10:53 | 2019-02-18T18:10:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 465,133,631 | 4,439 | alexey-milovidov | One test has failed. | 2019-02-19T13:47:04 | 2019-02-19T13:47:04 | {} | MEMBER |
ClickHouse | ClickHouse | 465,215,828 | 4,414 | alexey-milovidov | > As I now understand it, the data skipping index is tied to the primary key. E.g. If I have index_granularity=8192 and GRANULARITY=1, then each 8192 rows, the index contains say the minmax for the Nth primary key.
Correct.
> Is there an advantage to tieing the data skipping index to the primary key or would it m... | 2019-02-19T16:51:33 | 2019-02-19T16:51:33 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 463,654,160 | 4,399 | alexey-milovidov | > Fill one replica queue with trash
It can be done only with manual creating of nodes, because by itself, replica is moving log entries to its queue and processes them almost immediately. | 2019-02-14T14:48:40 | 2019-02-14T14:48:40 | {} | MEMBER |
ClickHouse | ClickHouse | 464,138,807 | 4,409 | alexey-milovidov | > interpreting individual Fields as arrays of arrays.
`Field` can be `Array`. `Array` is `std::vector<Field>`. `Array` elements can also be `Array` that means we have multidimensional array. | 2019-02-15T17:49:01 | 2019-02-15T17:49:26 | {} | MEMBER |
ClickHouse | ClickHouse | 464,772,168 | 4,423 | dr1nf3rn0 |
Reproduced on 19.1.6. Full stack:
```
2019.02.13 09:18:02.339898 [ 66 ] {} <Error> HTTPHandler: Code: 87, e.displayText() = DB::ErrnoException: Cannot seek through file /var/lib/clickhouse/data/DS_BMS_TEST/CC_STORE_DIM/f0478438e7458a6ce363150e36495c3d_0_0_0/BRANCH_NAME.bin, errno: 22, strerror: Invalid argument: ... | 2019-02-18T15:21:43 | 2019-02-18T15:22:37 | {} | NONE |
ClickHouse | ClickHouse | 464,985,026 | 4,431 | alniks | > INSERT INTO events (source_id, contact_id, created_at, email, ip, latitude, longitude, type, url, user_agent, user_id, event_date) format CSV
8, 98765, '2015-07-02 06:00:02', 'email@email.com', '', 0, 0, 1, '', '', 2,'2015-07-02'
... 5k similar lines
8, 104177, '2015-07-02 14:18:17', 'another@email.com', '10.10.1... | 2019-02-19T05:08:41 | 2019-02-19T05:10:59 | {} | NONE |
ClickHouse | ClickHouse | 465,291,806 | 4,047 | evgeny-l | @alesapin
Hi! Seems like I'm experiencing similar issue with my ClickHouse setup.
ClickHouse error log got spammed with messages like:
019.02.19 20:09:33.781020 [ 2 ] {} <Error> events.bid_events (StorageReplicatedMergeTree): DB::StorageReplicatedMergeTree::queueTask()::<lambda(DB::StorageReplicatedMergeTree::L... | 2019-02-19T20:12:57 | 2019-02-19T20:12:57 | {
"+1": 3
} | NONE |
ClickHouse | ClickHouse | 463,482,575 | 3,885 | zombiemonkey | https://github.com/yandex/ClickHouse/issues/4389 | 2019-02-14T04:14:21 | 2019-02-14T04:14:21 | {} | NONE |
ClickHouse | ClickHouse | 463,984,870 | 4,405 | alexey-milovidov | Performance test has failed.
- [x] Never move all requested columns to PREWHERE. | 2019-02-15T10:19:28 | 2019-02-15T19:27:08 | {} | MEMBER |
ClickHouse | ClickHouse | 464,760,323 | 4,436 | filimonov | ```sql
select
groupid,
argMax(value, timestamp)
FROM table
where timestamp > now() - INTERVAL 1 DAY
group by groupid
``` | 2019-02-18T14:53:34 | 2019-02-18T14:53:34 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 465,060,548 | 4,040 | filimonov | Full stacktrace:
```
Received exception from server (version 19.3.3):
Code: 10. DB::Exception: Received from localhost:9000, ::1. DB::Exception: Not found column comment_expression in block. There are only columns: name, type, default_type, default_expression. Stack trace:
0. clickhouse-server(StackTrace::StackTr... | 2019-02-19T09:43:45 | 2019-02-19T09:43:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 465,080,448 | 4,431 | M1ha-Shvn | Same problem. I've tried 19.3.3 and 19.3.4. On 19.1.8 it works fine
I use CollapsingMergeTree without Buffer. | 2019-02-19T10:41:55 | 2019-02-19T10:42:38 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 463,907,656 | 721 | kkrgwbj | I also encountered the same problem, changing listern_host to 0.0.0.0 also failed.clickhouse version is 18
@alexey-milovidov @bamx23 | 2019-02-15T05:01:59 | 2019-02-15T05:01:59 | {} | NONE |
ClickHouse | ClickHouse | 463,964,109 | 4,409 | kzon | @alexey-milovidov I need your advice. It looks like there is no methods for interpreting individual `Field`s as arrays of arrays. How we can do it? Do we need new GatherUtils function? Or I missed something about memory layout of `Field`? | 2019-02-15T09:15:20 | 2019-02-15T09:15:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,302,377 | 3,605 | filimonov | > Or is there currently any manual workarounds to achieve this functionality?
You can mount / symlink database folder in `/var/lib/clickhouse/data` to another storage and move data by detach/attach parts from fast storage database to slow storage database (of course table definitions should match) | 2019-02-16T07:04:06 | 2019-02-16T07:04:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,430,803 | 4,393 | simPod | @arctica yes, that makes sense! Something like
```diff
- data parts need not be changed (they remain sorted by the new sorting key expression).
+ existing data parts do not need to be changed. The new sorting key is actually extended old sorting key and therefore the data are already sorted by the new sorting expr... | 2019-02-17T08:56:46 | 2019-02-17T08:56:46 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,670,651 | 3,605 | filimonov | > @filimonov if I understand you correctly, this would leave us with 2 separate databases and 2 separate tables, each in its own database.
Actually, it can be in one database, enough that table directories will point to different storages.
So the simplest way to archive it currently:
```sql
create table mydb.my... | 2019-02-18T10:13:10 | 2019-02-18T10:13:10 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,769,116 | 4,435 | alexey-milovidov | The order of execution is unspecified. | 2019-02-18T15:14:21 | 2019-02-18T15:14:21 | {} | MEMBER |
ClickHouse | ClickHouse | 464,837,904 | 4,439 | kvinty | [pdqsort.log](https://gist.github.com/kvinty/5b7a0c4be9cd561889c31705deaa2946)
[radix.log](https://gist.github.com/kvinty/514682f90312192d2941bfbc10a864b6) | 2019-02-18T18:34:33 | 2019-02-18T18:34:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,855,659 | 4,040 | mcarbonneaux | i've the same when try to connect to 18.14.19 from 19.1.6...
```
Received exception from server (version 19.1.6):
Code: 10. DB::Exception: Received from localhost:9000, 127.0.0.1. DB::Exception: Not found column comment_expression in block. There are only columns: name, type, default_type, default_expression.
```... | 2019-02-18T19:47:14 | 2019-02-18T19:47:49 | {} | NONE |
ClickHouse | ClickHouse | 465,113,446 | 4,435 | filimonov | With regular joins rowNumberInBlock works correctly:
```
SELECT
number AS x,
y,
rowNumberInBlock()
FROM numbers(3)
ALL LEFT JOIN
(
SELECT
toUInt64(number % 4) AS x,
number AS y
FROM numbers(6)
) USING (x)
┌─x─┬─y─┬─rowNumberInBlock()─┐
│ 0 │ 0 │ ... | 2019-02-19T12:40:15 | 2019-02-19T12:40:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 465,273,811 | 4,040 | alexey-milovidov | The fix will be in the next release. | 2019-02-19T19:22:29 | 2019-02-19T19:22:29 | {
"+1": 3
} | MEMBER |
ClickHouse | ClickHouse | 465,274,296 | 4,442 | pdeva | we are seeing a similar message. clickhouse 19.3.4
The error msg should atleast output the table/kafka topic this relates to, so we know where to even begin debugging.
```
2019.02.19 19:22:05.594152 [ 31 ] {} <Error> void DB::StorageKafka::streamThread(): Code: 27, e.displayText() = DB::Exception: Cannot parse in... | 2019-02-19T19:23:44 | 2019-02-19T19:23:44 | {} | NONE |
ClickHouse | ClickHouse | 463,469,712 | 3,885 | zombiemonkey | I actually think this may be justification for parameterized views instead. That would be a very powerful feature. | 2019-02-14T03:09:28 | 2019-02-14T03:09:28 | {} | NONE |
ClickHouse | ClickHouse | 463,523,038 | 4,362 | etah000 | also works on 18.14.13. | 2019-02-14T07:33:35 | 2019-02-14T07:33:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,555,490 | 4,390 | nezed | Probably I've missed new and unannounced ability to create custom indexes
https://clickhouse.yandex/docs/en/operations/table_engines/mergetree/#data-skipping-indices | 2019-02-14T09:26:39 | 2019-02-14T09:28:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,678,153 | 3,611 | den-crane | I can reproduce it with 100% probability.
ClickHouse server version 18.14.18 revision 54409.
```
not sure is it important or not -- time=5.73 ms latency to zk.
ping zoo1
.
64 bytes from 10......2: icmp_seq=1 ttl=62 time=5.73 ms
64 bytes from 10......2: icmp_seq=2 ttl=62 time=5.12 ms
```
on 2 servers:
CREATE... | 2019-02-14T15:50:00 | 2019-02-14T15:57:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 463,986,206 | 4,207 | alexey-milovidov | Scheduled for the next few days. | 2019-02-15T10:22:58 | 2019-02-15T10:22:58 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 464,573,487 | 4,430 | zhang2014 | ```
ClickHouse :) SELECT COUNT() FROM test SETTINGS max_execution_bytes_speed = 30971520;
SELECT COUNT()
FROM test
SETTINGS max_execution_bytes_speed = 30971520
┌──COUNT()─┐
│ 10000000 │
└──────────┘
1 rows in set. Elapsed: 0.329 sec. Processed 10.00 million rows, 10.00 MB (30.37 million rows/s., 30.37 M... | 2019-02-18T03:52:00 | 2019-02-18T03:52:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 464,750,542 | 3,033 | fessmage | Got this too - when i define config with only tcp_port_secure on single server - its fine. But when i added cluster config with zookeeper section - server wont start, with `<Error> Application: Not found: tcp_port` in log. its good that this error can be mitigated by define tcp_port and closing it with firewall, but qu... | 2019-02-18T14:27:59 | 2019-02-18T14:27:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 465,110,504 | 4,443 | Slind14 | ahh, sorry my mistake. The query had multiple "toDate" and I forgot one of them. | 2019-02-19T12:29:36 | 2019-02-19T12:29:36 | {} | NONE |
ClickHouse | ClickHouse | 463,465,807 | 3,885 | zombiemonkey | Unfortunately this fix in 19.3.3 has broken some key features we've been using where we calculate deltas on 30-40 million series of realtime data (billions of rows per day). We had some very complex logic we were able to encapsulate in views with predicate push down. In our case we are not dependent on the result of th... | 2019-02-14T02:50:23 | 2019-02-14T02:52:47 | {} | NONE |
ClickHouse | ClickHouse | 463,648,123 | 4,401 | alexey-milovidov | A test? | 2019-02-14T14:31:57 | 2019-02-14T14:31:57 | {} | MEMBER |
ClickHouse | ClickHouse | 464,122,556 | 4,282 | kmatt | In preparation for the test, I experienced a similar hang of a DROP TABLE of a Kafka engine type on v18.16.1. Attaching the trace for this condition here, captured at 10 minutes into the drop process. Not the consumer materialized views were dropped first, so there should be no connections to the Kafka engines.
[tra... | 2019-02-15T16:58:32 | 2019-02-15T17:06:00 | {} | NONE |
ClickHouse | ClickHouse | 464,363,036 | 1,826 | zotov-vs | > В последние недели ODBC драйвер был сильно доработан, стоит попробовать свежие версии.
сервер Clickhouse 19.1.6
драйвер odbc 1.0.0.20190201
(Also you can: write a loop with copy assignment (but sometimes compiler cannot replace it with a call to memcpy and it will work slower; or you can get rid of extra copying)). | 2019-02-19T13:45:57 | 2019-02-19T13:45:57 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.