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 | 573,306,892 | 8,571 | alexey-milovidov | Yes, give me a few days... | 2020-01-11T11:09:57 | 2020-01-11T11:09:57 | {} | MEMBER |
ClickHouse | ClickHouse | 573,747,593 | 6,681 | alexey-milovidov | Ok.
Let's integrate this patch... Could you please send a PR? | 2020-01-13T16:24:36 | 2020-01-13T16:24:36 | {} | MEMBER |
ClickHouse | ClickHouse | 573,787,118 | 8,636 | alexey-milovidov | What query do you execute? What is the structure of your tables and configuration of the cluster? | 2020-01-13T17:51:11 | 2020-01-13T17:51:11 | {} | MEMBER |
ClickHouse | ClickHouse | 573,820,946 | 8,638 | KochetovNicolai | The same error for all builds in master.
Should be fixed now. | 2020-01-13T19:07:52 | 2020-01-13T19:07:52 | {} | MEMBER |
ClickHouse | ClickHouse | 574,047,159 | 4,901 | fillest | >
>
> If even restart doesn't help, it means that ZooKeeper is not accessible at all.
Oops, I've just noticed `No metadata in ZooKeeper` after the restart. I definitely could fully reset Zk cluster (it was some time ago and it's dev environment and we share Zk with Kafka cluster - which I was re-creating at some... | 2020-01-14T07:45:49 | 2020-01-14T07:45:49 | {} | NONE |
ClickHouse | ClickHouse | 573,157,893 | 8,618 | alexey-milovidov | We also provide pre-build binaries.
Example: https://clickhouse-builds.s3.yandex.net/0/fd40b25107f68942be6878c777d9d985b9185746/report.html | 2020-01-10T18:44:29 | 2020-01-10T18:44:29 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 573,291,663 | 8,618 | haitang002 | > You don't have to build all targets.
> To build clickhouse, just type `ninja clickhouse`.
ninja clickhouse.
Thank you for your reply，I have successfully built clicokhouse on the aarch64 platform using this command. I hope clickhouse has better support for the aarch64 platform，Thank you again !! | 2020-01-11T07:34:59 | 2020-01-11T07:34:59 | {} | NONE |
ClickHouse | ClickHouse | 573,325,632 | 8,566 | alexey-milovidov | `zkutil.multi_async` is flacky test.
| 2020-01-11T15:19:25 | 2020-01-11T15:19:25 | {} | MEMBER |
ClickHouse | ClickHouse | 573,333,718 | 7,081 | alexey-milovidov | > For different indexes: set different default values for granularity may also be helpful?
Usually granularity 1 is enough and bigger value is considered as fine-tuning. | 2020-01-11T16:55:47 | 2020-01-11T16:55:47 | {} | MEMBER |
ClickHouse | ClickHouse | 573,436,893 | 6,892 | alexey-milovidov | Tsan report is relevant and it looks familiar to me.
Did you integrate all the patches? | 2020-01-12T17:18:30 | 2020-01-12T17:18:30 | {} | MEMBER |
ClickHouse | ClickHouse | 573,539,657 | 8,628 | DarkWanderer | Duplicate of #7019, sorry | 2020-01-13T07:32:20 | 2020-01-13T07:32:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,909,690 | 8,624 | azat | >Reinventing aerospike (non distributed) ? :)
rocksdb? | 2020-01-13T22:47:23 | 2020-01-13T22:47:23 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 573,957,936 | 8,513 | nauu | @alex-krash, The column types as follows:
```
┌─name─┬─type─────────────┬─default_type─┬─default_expression─┬─comment─┬─codec_expression─┬─ttl_expression─┐
│ id │ Int32 │ │ │ │ │ │
│ name │ Nullable(String) │ │ ... | 2020-01-14T01:40:21 | 2020-01-14T01:40:21 | {} | MEMBER |
ClickHouse | ClickHouse | 574,040,340 | 8,627 | 865357384 | OK ，Thank you very much for answering my question ！！！ | 2020-01-14T07:24:40 | 2020-01-14T07:24:40 | {} | NONE |
ClickHouse | ClickHouse | 574,042,872 | 8,513 | alex-krash | @nauu, so the situation is the following:
1. JDBC bridge discover data types from your external database, and map each row in sequence `Int32, Nullable(String)`
2. On ClickHouse side (in table DDL), you specified your expectations about this data - `Int16, Nullable(String)`.
3. When reading the data, content is br... | 2020-01-14T07:32:11 | 2020-01-14T07:32:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,080,563 | 8,644 | alexey-milovidov | But we already merged it: https://github.com/ClickHouse/ClickHouse/blob/master/dbms/src/Storages/LiveView/LiveViewBlockInputStream.h#L188 | 2020-01-14T09:22:02 | 2020-01-14T09:22:02 | {} | MEMBER |
ClickHouse | ClickHouse | 573,622,278 | 6,168 | YiuRULE | @nikitamikhaylov I normally fix the problem brought by @filimonov, nothing else expected on my side ? :p | 2020-01-13T11:36:40 | 2020-01-13T11:36:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,080,973 | 8,644 | KochetovNicolai | I've done the same in https://github.com/ClickHouse/ClickHouse/commit/bcc0cd8e47022e7e16ad90341da2eea137129270
Do not understand why there is no conflict | 2020-01-14T09:22:56 | 2020-01-14T09:22:56 | {} | MEMBER |
ClickHouse | ClickHouse | 560,025,636 | 7,799 | orloffv | in PREWHERE section
```Method createColumn() is not implemented for data type Function((String, String) -> String)```
```
SELECT
arrayJoin AS kv_key
FROM system.query_log
ARRAY JOIN ProfileEvents.Names AS arrayJoin
PREWHERE has(arrayMap(key -> key, ProfileEvents.Names), 'Query')
WHERE arrayJoin = 'Query'
... | 2019-11-30T22:14:53 | 2020-01-14T10:11:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,301,172 | 8,331 | Vxider | Hi, Thanks for the outstanding work of `LIVE VIEW`, it has guided me when I implement `WINDOW VIEW`.
`WINDOW VIEW` is aimed at implementing the stream processing of ClickHouse. It is different from `LIVE VIEW`.
## Design
In terms of design, `WINDOW VIEW` and `LIVE VIEW` are different from **how to aggregate d... | 2020-01-14T18:07:09 | 2020-01-14T18:07:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,120,361 | 5,779 | temoon | Thanks for fix of tables, next time - fix database ;)
```
CREATE DATABASE `.test`
```
This XML syntax won't work for my `.test` database:
```
<?xml version="1.0"?>
<yandex>
<users>
<default>
<databases>
<.test>
<table name='.table'>
... | 2020-01-10T17:04:47 | 2020-01-10T17:04:47 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,317,658 | 8,151 | alexey-milovidov | If we want to go forward, we have to answer the following question:
- is it the issue with libunwind (it cannot decode unwind tables but it should)
- or with clang (clang generates wrong unwind tables)?
To check it, we can build with clang and omit-frame-pointer, but link with gcc's libunwind or original "nongnu" ... | 2020-01-11T13:36:56 | 2020-01-11T13:36:56 | {} | MEMBER |
ClickHouse | ClickHouse | 573,750,878 | 4,901 | alexey-milovidov | Another question is why you are using 19.4.2 if newer versions (19.14 ... 19.17) are available. | 2020-01-13T16:30:56 | 2020-01-13T16:30:56 | {} | MEMBER |
ClickHouse | ClickHouse | 573,810,404 | 8,626 | excitoon | Reproducing with TSan did not give anything. | 2020-01-13T18:43:31 | 2020-01-13T18:43:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,043,110 | 8,634 | cliicy | hi ,
thanks for your information.
I load 59986052 rows data and compared the compressed with uncompressed. there are indeed very different.
compressed info:
uncompressed info:
. | 2020-01-14T18:47:40 | 2020-01-14T18:47:49 | {} | MEMBER |
ClickHouse | ClickHouse | 574,130,634 | 6,681 | qMBQx8GH | @alexey-milovidov , hi!
I'm done with a PR.
Not sure how to point on it thou.
Here is the link: https://github.com/ClickHouse/ClickHouse/pull/8647 | 2020-01-14T11:27:25 | 2020-01-14T11:27:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,148,237 | 8,647 | qMBQx8GH | > Ok. And we also need a test (simple .sql test is enough, see `dbms/tests/queries`) to check that the setting works.
done | 2020-01-14T12:19:37 | 2020-01-14T12:19:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,219,674 | 7,646 | filimonov | @ivanprostran did you check system.metrics BackgroundSchedulePoolTask ? Are you really close the the top? | 2020-01-14T15:10:21 | 2020-01-14T15:10:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,306,309 | 8,651 | alexey-milovidov | The whole list is partly outdated, because now we mark all easy tasks with tags:
https://github.com/ClickHouse/ClickHouse/issues?q=is%3Aopen+is%3Aissue+label%3A%22easy+task%22
But I'm not sure that all tasks from this list are present as issues. | 2020-01-14T18:19:57 | 2020-01-14T18:19:57 | {} | MEMBER |
ClickHouse | ClickHouse | 573,118,150 | 8,475 | alexey-milovidov | `edgeLengthM` - tech debt. | 2020-01-10T16:59:14 | 2020-01-10T16:59:14 | {} | MEMBER |
ClickHouse | ClickHouse | 573,300,578 | 8,487 | alexey-milovidov | Thank you!
This is way old code, I also found a simple way to overcome this issue: #8616 | 2020-01-11T09:41:53 | 2020-01-11T09:41:53 | {} | MEMBER |
ClickHouse | ClickHouse | 573,300,713 | 8,493 | alexey-milovidov | Thank you! I also applied some changes to your code, please review: #8617 | 2020-01-11T09:43:39 | 2020-01-11T09:43:39 | {} | MEMBER |
ClickHouse | ClickHouse | 573,333,836 | 8,493 | alexey-milovidov | @kiransunkari This pull request is merged and all your commits are in master along with my commits. | 2020-01-11T16:57:17 | 2020-01-11T16:57:17 | {} | MEMBER |
ClickHouse | ClickHouse | 573,401,083 | 8,101 | alexey-milovidov | Does it reproduce if you build with bundled libraries (where Poco already contains this patch)? | 2020-01-12T10:25:32 | 2020-01-12T10:25:32 | {} | MEMBER |
ClickHouse | ClickHouse | 573,698,451 | 8,602 | alexey-milovidov | > possible than we still have a trouble if CH has not been shutdown properly
It's not that bad :) | 2020-01-13T14:47:52 | 2020-01-13T14:47:52 | {} | MEMBER |
ClickHouse | ClickHouse | 573,745,876 | 7,477 | alexey-milovidov | @Vinatorul Hi! How is it going?
I will be happy to hear some news... | 2020-01-13T16:21:07 | 2020-01-13T16:22:49 | {} | MEMBER |
ClickHouse | ClickHouse | 573,851,706 | 8,641 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/pull/937#issuecomment-314255963 | 2020-01-13T20:20:16 | 2020-01-13T20:20:16 | {
"laugh": 1
} | MEMBER |
ClickHouse | ClickHouse | 573,943,325 | 8,626 | excitoon | It turned out that that is simply deduplication of `ReplicatedMergeTree`. | 2020-01-14T00:41:21 | 2020-01-14T00:41:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,309,995 | 8,472 | alexey-milovidov | This PR adds more issues. | 2020-01-14T18:29:29 | 2020-01-14T18:29:29 | {} | MEMBER |
ClickHouse | ClickHouse | 574,323,128 | 5,826 | alexey-milovidov | Current status: I have added query_id to these messages but not the query itself.
I think it's enough. | 2020-01-14T19:01:56 | 2020-01-14T19:01:56 | {} | MEMBER |
ClickHouse | ClickHouse | 573,117,782 | 8,416 | abyss7 | Will add tests in a separate PR. | 2020-01-10T16:58:18 | 2020-01-10T16:58:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,189,093 | 8,614 | alesapin | test are ok | 2020-01-10T20:15:42 | 2020-01-10T20:15:42 | {} | MEMBER |
ClickHouse | ClickHouse | 573,316,354 | 8,620 | alexey-milovidov | Awesome!
BTW, do you have any ideas how we can automate this?
I tried to do it with `aspell` but it requires manual review of results. | 2020-01-11T13:19:40 | 2020-01-11T13:19:40 | {} | MEMBER |
ClickHouse | ClickHouse | 573,330,679 | 7,081 | zhang2014 | For different indexes: set different default values for granularity may also be helpful? | 2020-01-11T16:18:42 | 2020-01-11T16:18:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,345,284 | 8,581 | excitoon | Problem does not reproduce on 19.16.9.37. | 2020-01-11T19:07:26 | 2020-01-11T19:07:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,474,648 | 6,462 | toddyocum | I'm also seeing not dropped partitions.
ClickHouse server version 19.14.7 revision 54425.
Here is the Engine part of the output from the 'show create table' command for the table in question
CREATE TABLE default.annotated_flow_local (`timestamp` DateTime,
...snip....
ENGINE = MergeTree() PARTITION BY toStar... | 2020-01-13T00:18:39 | 2020-01-13T00:19:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,564,957 | 6,681 | qMBQx8GH | Hi, all!
I've tried to make suggested fix, but it didn't work for me.
Now I the following lines in the server log:
```
2020.01.13 10:47:40.777896 [ 1 ] {} <Information> DatabaseOrdinary (system): Starting up tables.
2020.01.13 10:47:40.896256 [ 1 ] {} <Error> Application: Caught exception while loading metadata: C... | 2020-01-13T08:59:42 | 2020-01-13T08:59:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,739,609 | 8,634 | den-crane | they are just file's(streams) headers.
Try to load more rows (million) and compare compressed VS uncompressed again | 2020-01-13T16:08:32 | 2020-01-13T16:08:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,850,961 | 8,638 | Kondzolko | > The same error for all builds in master.
> Should be fixed now.
Thank you. | 2020-01-13T20:18:17 | 2020-01-13T20:18:17 | {} | NONE |
ClickHouse | ClickHouse | 574,178,494 | 7,080 | alexey-milovidov | There is no timeline for this task. But it is easy to implement.
The only important thing is to check that replica is not active, and while removing - remove some nodes first in a multi transaction with check of `active` node - to avoid possible race condition. | 2020-01-14T13:39:35 | 2020-01-14T13:39:35 | {} | MEMBER |
ClickHouse | ClickHouse | 574,318,858 | 4,747 | alexey-milovidov | It was implemented through virtual column: https://github.com/ClickHouse/ClickHouse/pull/7624 | 2020-01-14T18:51:51 | 2020-01-14T18:51:51 | {} | MEMBER |
ClickHouse | ClickHouse | 573,065,312 | 8,608 | den-crane | OK. Column aliases are tricky. Good.
```
select 0 as PGID, PGID2 from
(select sg.PGID AS PGID2 from tmp_data.sy_game sg)
``` | 2020-01-10T14:50:54 | 2020-01-10T14:50:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,158,040 | 8,618 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/issues/8027#issuecomment-566671201 | 2020-01-10T18:44:57 | 2020-01-10T18:44:57 | {} | MEMBER |
ClickHouse | ClickHouse | 573,680,980 | 8,101 | martenlindblad | It does not reproduce, with prebundled libaries it works. | 2020-01-13T14:11:34 | 2020-01-13T14:11:34 | {} | NONE |
ClickHouse | ClickHouse | 573,690,155 | 8,559 | alexey-milovidov | Could you please take a look to `ClickHouse build check — 0/14 builds are OK`? | 2020-01-13T14:30:31 | 2020-01-13T14:30:38 | {} | MEMBER |
ClickHouse | ClickHouse | 573,719,356 | 4,901 | fillest | We probably hit the same problem (not sure if conditions are different). We're on 19.16.2.2.
The relevant messages in system.text_log seem to be (I added some comments and cut some lines):
```
Code: 33, e.displayText() = DB::Exception: Cannot read all data. Bytes read: 0. Bytes expected: 4., Stack trace:
0. 0x55c24... | 2020-01-13T15:29:05 | 2020-01-13T15:29:05 | {} | NONE |
ClickHouse | ClickHouse | 573,761,616 | 8,635 | den-crane | Attach with non-existing ZK path works without errors because it needed to start CH without ZK (in readonly mode) https://github.com/ClickHouse/ClickHouse/pull/4691 | 2020-01-13T16:53:03 | 2020-01-13T16:53:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,308,938 | 8,651 | alexey-milovidov | ba905e19bf | 2020-01-14T18:26:57 | 2020-01-14T18:26:57 | {} | MEMBER |
ClickHouse | ClickHouse | 447,836,695 | 79 | Crazylearner30 | Trying to load data into the table and its throwing error which is not in specific
Table:
CREATE TABLE i2.company_temp (
a UInt32,
b String,
c String,
.....
......
) ENGINE = MergeTree PARTITION BY state_province_code
ORDER BY
domain SETTINGS index_granularity = 8192
Insert statement:
cli... | 2018-12-17T12:57:03 | 2020-01-13T09:27:20 | {} | NONE |
ClickHouse | ClickHouse | 573,745,704 | 8,627 | meob | I reply from my personal experience...
We generally used ETL tools... but with ClickHouse is very easy to access MySQL tables:
https://clickhouse.yandex/docs/en/operations/table_engines/mysql/
If the application logic is complex You can use MySQL views.
We generally have a complex OLTP DB design in MySQL, we c... | 2020-01-13T16:20:45 | 2020-01-13T16:20:45 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,750,184 | 4,901 | alexey-milovidov | You can debug it with some zookeeper client (`zkCli.sh` for example) to answer these questions:
- what ZooKeeper servers in a cluster are online and accept connections;
- is ZooKeeper cluster configured correctly - so all ZK servers belong to the same cluster;
- what is the latency of ZK connection and operations.
... | 2020-01-13T16:29:34 | 2020-01-13T16:29:49 | {} | MEMBER |
ClickHouse | ClickHouse | 574,155,637 | 8,331 | vzakaznikov | @Vxider, I am wondering how it is fundamentally different from LV with an exception of having a target table (materialized). Right now using LV with a sub-query a window can be created. For example, using an LV a real-time moving average over the last 10 values can be calculated as
```sql
CREATE LIVE VIEW ma10 AS
S... | 2020-01-14T12:40:57 | 2020-01-14T12:40:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,312,713 | 8,646 | alexey-milovidov | `arrayReduce` function with `uniqState` can also be used to build the value of `AggregateFunction(uniq, String)` type. | 2020-01-14T18:36:24 | 2020-01-14T18:36:24 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 574,324,655 | 5,893 | alexey-milovidov | @kamalov-ruslan @xPoSx @demo-99 | 2020-01-14T19:05:54 | 2020-01-14T19:05:54 | {} | MEMBER |
ClickHouse | ClickHouse | 574,325,049 | 5,895 | alexey-milovidov | @kamalov-ruslan @xPoSx @demo-99 | 2020-01-14T19:06:58 | 2020-01-14T19:06:58 | {} | MEMBER |
ClickHouse | ClickHouse | 573,688,373 | 8,101 | alexey-milovidov | We use multiple patches for Poco. Actually, using "stock" Poco version is not safe with ClickHouse. | 2020-01-13T14:26:49 | 2020-01-13T14:26:49 | {} | MEMBER |
ClickHouse | ClickHouse | 574,025,367 | 6,681 | qMBQx8GH | > Ok.
>
> Let's integrate this patch... Could you please send a PR?
Yes I'll send one. I'll try my best to finish it today | 2020-01-14T06:32:54 | 2020-01-14T06:32:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,323,503 | 5,862 | alexey-milovidov | `gz` and `br` has been implemented.
Now other generic compression methods can be added easily. | 2020-01-14T19:02:52 | 2020-01-14T19:02:52 | {} | MEMBER |
ClickHouse | ClickHouse | 573,174,174 | 8,612 | alesapin | Checked on gentoo under root user in docker. Works. | 2020-01-10T19:30:57 | 2020-01-10T19:30:57 | {} | MEMBER |
ClickHouse | ClickHouse | 573,795,839 | 8,581 | alexey-milovidov | It should be fixed, indeed. | 2020-01-13T18:11:38 | 2020-01-13T18:11:38 | {} | MEMBER |
ClickHouse | ClickHouse | 573,945,868 | 7,646 | den-crane | users.xml (select * from system.settings)
```
<?xml version="1.0"?>
<yandex>
<profiles>
<default>
<background_schedule_pool_size>25</background_schedule_pool_size>
```
also users
```
<max_insert_block_size>10485760</max_insert_block_size>
<max_... | 2020-01-14T00:51:56 | 2020-01-14T00:51:56 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,082,549 | 8,644 | Vxider | Thanks，I'm sorry for that. | 2020-01-14T09:26:23 | 2020-01-14T09:26:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,163,152 | 8,489 | KochetovNicolai | Good. And the same change for HDFS. | 2020-01-14T13:02:14 | 2020-01-14T13:02:14 | {} | MEMBER |
ClickHouse | ClickHouse | 573,062,225 | 8,607 | den-crane | >The most used columns in the queries we have put in order by.
It's not quite right.
All table -- all columns will be sorted in order of `ORDER BY`
You have to define `ORDER BY` to achieve maximum compression by puting similar values in columns as close as possible. The same time `ORDER BY` -- creates an PRIMARY ... | 2020-01-10T14:43:02 | 2020-01-10T14:45:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,323,386 | 8,444 | alexey-milovidov | It doesn't check for correct encoding in decode functions (at least in `_tb64xd` function).
`_tb64xd` (tails processing routine) doesn't contain checks despite the fact that other routines have checks. I give up for now. | 2020-01-11T14:51:22 | 2020-01-11T14:59:46 | {} | MEMBER |
ClickHouse | ClickHouse | 573,325,535 | 8,581 | alexey-milovidov | @tavplubix There was an issue in some version from master (around 31th December) that .sql files can be created in a wrong place. Do you remember anything? | 2020-01-11T15:18:00 | 2020-01-11T15:18:07 | {} | MEMBER |
ClickHouse | ClickHouse | 573,669,184 | 5,318 | BayoNet | Merged in #8401, #8451 | 2020-01-13T13:43:28 | 2020-01-13T13:43:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,981,102 | 6,721 | mybluecrab | I used clang build ok on aarch64, but when I run the clickhouse-server , is locked. gdb info
#0 Poco::MutexImpl::lockImpl (this=0xffffe5703988) at /usr/include/Poco/Mutex_POSIX.h:59
#1 0x0000fffff67f3c44 in Poco::FastMutex::lock (this=0xffffe5703988)
at /usr/include/Poco/Mutex.h:238
#2 Poco::ScopedLock<Poc... | 2020-01-14T03:21:25 | 2020-01-14T03:21:25 | {} | NONE |
ClickHouse | ClickHouse | 574,100,544 | 8,634 | alexey-milovidov | Yes. Technically, when you choose codec NONE, data is still placed into compressed container but not compressed inside it. ClickHouse will report the size of this container as "data_compressed_bytes" regardless to the fact that data is not compressed inside it. | 2020-01-14T10:11:09 | 2020-01-14T10:11:09 | {} | MEMBER |
ClickHouse | ClickHouse | 573,624,804 | 8,632 | 4ertus2 | partial_merge_join is in progress and not fully optimized yet. But it's expected that MergeJoin would be slower than HashJoin in general. The main case it should be faster is sorted inputs, where we could avoid sort phase in merge join. But it needs special support and not realized yet.
The main reason for partial_m... | 2020-01-13T11:44:15 | 2020-01-13T11:46:00 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,642,370 | 8,581 | tavplubix | I remember #8176, in case of restart after incomplete drop .sql file could be moved to current dir, but it looks not related | 2020-01-13T12:35:22 | 2020-01-13T12:35:22 | {} | MEMBER |
ClickHouse | ClickHouse | 573,743,501 | 8,627 | den-crane | Maybe.
It depends.
MySQL is much faster on "select by primary_key" and supports updates (CH `alter update` is much slower. ) | 2020-01-13T16:16:29 | 2020-01-13T16:16:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,765,857 | 8,635 | den-crane | I don't understand what problem are trying to solve. You don't have a knowledge is it new shard/replica or previously existent?
ZK does not store data only metadata (so to restore replica you need another live replica with data).
You can check zk_path existence using `select ... from system.zookeeper`.
| 2020-01-13T17:02:20 | 2020-01-13T17:02:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,238,076 | 7,249 | nvartolomei | #8449 | 2020-01-14T15:46:12 | 2020-01-14T15:46:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,050,824 | 5,690 | barrkel | It shouldn't happen at runtime; it should happen at analysis time, at the same time that type checking is being performed. This shouldn't happen per row processed; it should be when the parsed SQL is analyzed.
SQL-92 would interpret the literal 1000.0000 as effectively typed `decimal(8,4)`.
Section 5.3 (13) in ht... | 2020-01-10T14:12:44 | 2020-01-10T14:15:18 | {} | NONE |
ClickHouse | ClickHouse | 573,301,573 | 8,571 | oandrew | @alexey-milovidov could you take a look at this when you have a chance? Thanks | 2020-01-11T09:54:04 | 2020-01-11T09:54:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,564,863 | 7,646 | ghuname | @filimonov Can you please answer to @ivanprostran last question? | 2020-01-13T08:59:24 | 2020-01-13T08:59:24 | {} | NONE |
ClickHouse | ClickHouse | 573,600,412 | 8,513 | alex-krash | @nauu , thanks for your report. Could you, please, provide output of the following command:
```
DESC ( SELECT
id,
name
FROM jdbc('jdbc:postgresql://ip:5432/dolphinscheduler?user=test&password=test', '', 't_ds_process_definition') )
```
We should see that it gives the same column types as you used for t... | 2020-01-13T10:38:16 | 2020-01-13T10:38:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,905,086 | 8,624 | filimonov | Reinventing aerospike (non distributed) ? :)
See #2718 #5421 | 2020-01-13T22:34:23 | 2020-01-13T22:34:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,301,854 | 6,275 | bianxintong | can we have more details on the cause and progress on possible fixes? we're experiencing same issue. | 2020-01-11T09:57:56 | 2020-01-11T09:57:56 | {} | NONE |
ClickHouse | ClickHouse | 573,727,807 | 6,681 | qMBQx8GH | I've "fixed" all IParser::Pos instantions and only then found the code to fix this error.
In the file DatabaseOnDisk.cpp in function parseQueryFromMetadata where is tryParseQuery call w/o max_query_size and (just added) max_parser_depth parameters.
| 2020-01-13T15:45:26 | 2020-01-13T15:45:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 573,748,667 | 4,901 | alexey-milovidov | If even restart doesn't help, it means that ZooKeeper is not accessible at all. | 2020-01-13T16:26:37 | 2020-01-13T16:26:37 | {} | MEMBER |
ClickHouse | ClickHouse | 573,793,083 | 8,581 | vzakaznikov | Testing shows that 20.1.1.2107 still shows this issue. So is it expected to be fixed?
| 2020-01-13T18:05:21 | 2020-01-13T18:05:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,098,529 | 8,645 | 4ertus2 | https://en.wikipedia.org/wiki/Rounding
Round half to even
This variant of the round-to-nearest method is also called convergent rounding, statistician's rounding, Dutch rounding, Gaussian rounding, odd–even rounding,[6] or bankers' rounding. | 2020-01-14T10:05:56 | 2020-01-14T10:05:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,229,636 | 8,646 | den-crane | You can use Null table and MV
```
create table events_unique (
date Date,
group_id String,
client_id String,
event_type String,
product_id String,
value AggregateFunction(uniq, String)
) ENGINE = MergeTree(date, (group_id, client_id, event_type, product_id, date), 8192);
create table even... | 2020-01-14T15:29:42 | 2020-01-14T15:30:27 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 574,302,255 | 5,601 | oandrew | @Sugaroverdose most likely with the next PR once the current one is merged. Probably using libserdes would be the easiest. The main issue would be to figure out how to share state across different instances of InputFormat | 2020-01-14T18:09:45 | 2020-01-14T18:09:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 565,567,159 | 8,205 | alesapin | It can be implemented much faster and also seems useful! | 2019-12-13T19:02:38 | 2019-12-13T19:02:57 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.