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 | 364,298,697 | 320 | zhang2014 | @alexey-milovidov can we close this issues? | 2018-02-09T00:45:36 | 2018-02-09T00:45:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 364,375,923 | 1,879 | KochetovNicolai | autotest | 2018-02-09T09:05:50 | 2018-02-09T09:05:50 | {} | MEMBER |
ClickHouse | ClickHouse | 364,500,419 | 838 | alexey-milovidov | We decided the following:
For every column in a table it will be possible to specify a list of compression codecs stacked together.
Example:
```
x UInt64 CODEC(Delta, LZ4(1))
```
- it means, that Delta codec will be applied, then LZ4 with parameter 1.
If codec is specified for a column, it will have highest ... | 2018-02-09T17:24:54 | 2018-02-09T17:24:54 | {
"+1": 7
} | MEMBER |
ClickHouse | ClickHouse | 364,863,631 | 1,882 | zhang2014 | Is there documentation on how to add a performance test and see it ? | 2018-02-12T09:06:10 | 2018-02-12T09:06:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 364,963,522 | 1,883 | lamberken | :+1: | 2018-02-12T15:47:48 | 2018-02-12T15:47:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 364,994,224 | 1,882 | alexey-milovidov | How to add performance test?
Look at examples in `tests/performance`.
You can add your test case in .xml file, possibly in separate directory.
Performance test is a query or multiple queries, or a query with macro-substitutions,
along with run mode (run infinite query, run query in a loop),
stop conditions (ex... | 2018-02-12T17:16:21 | 2018-02-12T17:16:21 | {} | MEMBER |
ClickHouse | ClickHouse | 365,062,498 | 1,889 | alexey-milovidov | Yes, the performance is very different. Due to some implementation details, array `[[],[2]]` is not treated as a constant, and less efficient code path is selected.
Although this is not a bug, we should improve performance in this case. | 2018-02-12T21:08:43 | 2018-02-12T21:08:51 | {} | MEMBER |
ClickHouse | ClickHouse | 365,253,246 | 1,894 | ztlpn | Hi! I am not sure I understand what you are trying to achieve (queries 1 and 2 produce different output), can you try to state it more clearly? Some examples of input and desired output would be helpful.
Why is simple `SELECT [f(Dim1)] FROM ...` insufficient? | 2018-02-13T12:31:22 | 2018-02-13T12:31:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 366,232,861 | 601 | ztlpn | Yes, it seems fixed. Thanks for pointing out! | 2018-02-16T13:19:25 | 2018-02-16T13:19:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,392,549 | 863 | alexey-milovidov | Basic idea is to separate InterpreterSelectQuery and InterpreterUnionQuery; and ASTSelectQuery and ASTUnionQuery. I plan to do it only after https://github.com/yandex/ClickHouse/pull/1935 | 2018-02-21T16:51:36 | 2018-02-21T16:51:36 | {} | MEMBER |
ClickHouse | ClickHouse | 367,748,591 | 1,118 | alexey-milovidov | > @alexey-milovidov It would be also useful to include some recommendation about index granularity for the queries. E.g. when the most common and frequent query touches only few rows, but had to scan much more rows for that, we can suggest: "If you want this query to be faster you can change your table index granularit... | 2018-02-22T17:02:10 | 2018-02-22T17:02:10 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 367,752,170 | 1,844 | alexey-milovidov | `IntHash32` is definitely bad choice for HyperLogLog.
In fact, current implementation of HLL comes into ClickHouse from another project, where we have to manage large amount of very tiny (tens of bytes) HLLs for antifraud calculations. In that project, precision doesn't really matter. And we have to use the same has... | 2018-02-22T17:13:14 | 2018-02-22T17:13:52 | {} | MEMBER |
ClickHouse | ClickHouse | 367,759,276 | 1,870 | alexey-milovidov | https://github.com/arkhipov is going to do this. | 2018-02-22T17:35:42 | 2018-02-22T17:35:42 | {} | MEMBER |
ClickHouse | ClickHouse | 364,301,244 | 117 | zhang2014 | Oh, sorry, I forgot about it.I will support it later. @alexey-milovidov | 2018-02-09T00:59:09 | 2018-02-09T00:59:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 364,323,943 | 1,874 | lugt | Thx. #1880 | 2018-02-09T03:21:42 | 2018-02-09T03:33:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 364,911,660 | 1,883 | ztlpn | It depends on your query load. A good first step is to determine which resource (CPU, network bandwith or disk IO) is the bottleneck. You can do it using standard unix tools (e.g. dstat). | 2018-02-12T12:41:17 | 2018-02-12T12:41:17 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 366,336,769 | 1,918 | alexey-milovidov | ```
#define DBMS_MIN_REVISION_WITH_TIME_ZONE_PARAMETER_IN_DATETIME_DATA_TYPE 54311
```
That is slightly incorrect and compatibility isn't working for 1.1.54318. | 2018-02-16T19:35:28 | 2018-02-16T19:35:28 | {} | MEMBER |
ClickHouse | ClickHouse | 366,555,169 | 838 | Sindbag | I will take that. | 2018-02-18T22:47:00 | 2018-02-18T22:47:00 | {} | NONE |
ClickHouse | ClickHouse | 367,740,486 | 1,945 | alexey-milovidov | B. | 2018-02-22T16:37:39 | 2018-02-22T16:37:39 | {} | MEMBER |
ClickHouse | ClickHouse | 367,762,112 | 1,929 | alexey-milovidov | What OS and what version of ClickHouse do you use?
Do you have any ulimit on memory usage?
Could you please provide an output of
`sudo strace -f -p $(pidof clickhouse-server)`
with failed mmap call and with all previous mmap calls? | 2018-02-22T17:44:49 | 2018-02-22T17:44:49 | {} | MEMBER |
ClickHouse | ClickHouse | 364,300,823 | 1,840 | zhang2014 | @ludv1x I mean, rely on this refactor `InterpreterShowCreateQuery` , like `InterpreterShowTablesQuery`. But, as @alexey-milovidov says , their behavior is not consistent | 2018-02-09T00:56:42 | 2018-02-09T00:56:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 365,351,697 | 1,882 | alexey-milovidov | Looks like you compile with debug mode, because I see different numbers:
```
:) SELECT count() FROM system.numbers WHERE NOT ignore(lengthUTF8(materialize('Hello, world')))
SELECT count()
FROM system.numbers
WHERE NOT ignore(lengthUTF8(materialize('Hello, world')))
Ok.
Query was cancelled.
0 rows in se... | 2018-02-13T18:03:30 | 2018-02-13T18:03:30 | {} | MEMBER |
ClickHouse | ClickHouse | 367,360,682 | 1,910 | Slach | why not merged? | 2018-02-21T15:20:38 | 2018-02-21T15:20:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,745,759 | 1,944 | alexey-milovidov | As you see on the screenshot, that threads use 0.0% of CPU time and they are sleeping (S).
The green color in htop means "this is thread, not process" and it is completely irrelevant to status or CPU usage. | 2018-02-22T16:53:25 | 2018-02-22T16:53:25 | {} | MEMBER |
ClickHouse | ClickHouse | 367,748,323 | 1,118 | alexey-milovidov | > is there any plan to implement this feature ?
We started working on this plan: https://github.com/yandex/ClickHouse/issues/1118#issuecomment-323190806 | 2018-02-22T17:01:20 | 2018-02-22T17:01:20 | {} | MEMBER |
ClickHouse | ClickHouse | 364,654,752 | 180 | alexey-milovidov | Thanks! | 2018-02-10T13:37:06 | 2018-02-10T13:37:06 | {} | MEMBER |
ClickHouse | ClickHouse | 365,062,601 | 1,889 | alexey-milovidov | Fixed in master. | 2018-02-12T21:09:07 | 2018-02-12T21:09:07 | {} | MEMBER |
ClickHouse | ClickHouse | 367,528,200 | 1,911 | alexey-milovidov | Ok. Although it looks useless now. | 2018-02-22T00:41:57 | 2018-02-22T00:41:57 | {} | MEMBER |
ClickHouse | ClickHouse | 364,653,928 | 1,882 | alexey-milovidov | Something is wrong: https://travis-ci.org/yandex/ClickHouse/jobs/339794280#L6819 | 2018-02-10T13:29:22 | 2018-02-10T13:29:22 | {} | MEMBER |
ClickHouse | ClickHouse | 364,660,447 | 1,882 | zhang2014 | OK, I may need some time,because I can't use the computer tomorrow | 2018-02-10T14:54:13 | 2018-02-10T14:54:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 364,963,447 | 1,886 | gerasim13 | i found a bug, so dont merge this branch yet | 2018-02-12T15:47:36 | 2018-02-12T15:47:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 366,663,422 | 1,918 | bocharov | @alexey-milovidov thanks, we're going to backport the patch and try again. | 2018-02-19T11:33:40 | 2018-02-19T11:33:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,339,747 | 1,914 | KochetovNicolai | https://github.com/yandex/ClickHouse/pull/1938 | 2018-02-21T14:16:51 | 2018-02-21T14:16:51 | {} | MEMBER |
ClickHouse | ClickHouse | 367,217,597 | 863 | alexey-milovidov | It is blocked by other tasks.
Direct implementation on top of existing code of `InterpreterSelectQuery` and `ExpressionAnalyzer` is possible, but most likely will introduce a huge mess (probably not - you should try).
First we want to rewrite current implementation of UNION ALL. It is not necessary for this task,... | 2018-02-21T05:24:56 | 2018-02-21T05:24:56 | {} | MEMBER |
ClickHouse | ClickHouse | 367,525,088 | 1,297 | alexey-milovidov | Fixed long time ago. | 2018-02-22T00:25:55 | 2018-02-22T00:25:55 | {} | MEMBER |
ClickHouse | ClickHouse | 363,644,920 | 1,849 | lamberken | @ludv1x, detailed information
### Hardware
five machines as five shards, each of them is CPU E5-2650 v3 @ 2.30GHz, raid5
### Table
local table : datacenter.trajectory_trackless_complex_shard
distribute table : datacenter.trajectory_trackless_complex, shard: rand()
rows: `select count() from datacenter.trajector... | 2018-02-07T03:27:40 | 2018-02-09T06:10:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 364,492,365 | 127 | zhang2014 | I tried to reappear with the above statement, But I can't reproduce this problem. Maybe this problem has been fixed? Can we close this Issue? @alexey-milovidov | 2018-02-09T16:57:16 | 2018-02-09T16:57:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 364,503,566 | 1,881 | alexey-milovidov | Almost perfect :)
Just few points... | 2018-02-09T17:36:06 | 2018-02-09T17:36:06 | {} | MEMBER |
ClickHouse | ClickHouse | 364,998,083 | 1,882 | alexey-milovidov | In performance test, you should have a representative set of cases for different string sizes and for different ratio of non-ASCII characters. Also it's better to get realistic cases from real dataset... but first we will use only synthetic test cases.
The queries can look like this:
```
SELECT count() FROM syst... | 2018-02-12T17:29:00 | 2018-02-12T18:05:40 | {} | MEMBER |
ClickHouse | ClickHouse | 365,183,079 | 1,882 | zhang2014 | change before:
``` code javascript
Running: functions_length
[
{
"hostname": "bogon",
"main_metric": "avg_rows_per_second",
"num_cores": 4,
"num_threads": 8,
"parameters": {
"string": ["materialize('')", "materialize('Hello, world')", "toString(number)", "reinterpretAsString(number... | 2018-02-13T08:16:05 | 2018-02-13T08:16:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 365,572,911 | 1,894 | fireya | thank you very much for clarification | 2018-02-14T11:18:22 | 2018-02-14T11:18:22 | {} | NONE |
ClickHouse | ClickHouse | 365,829,604 | 1,907 | zhang2014 | Just fix the part of @MikhailKalashnikov in #590 | 2018-02-15T05:30:23 | 2018-02-15T05:30:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 366,152,900 | 601 | zhang2014 | @ludv1x I saw that it was solved in the master.Can we close this issue? | 2018-02-16T06:12:31 | 2018-02-16T06:12:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,169,537 | 1,928 | alexey-milovidov | Two tests fail. | 2018-02-21T00:23:30 | 2018-02-21T00:23:30 | {} | MEMBER |
ClickHouse | ClickHouse | 367,717,816 | 1,118 | bocharov | @alexey-milovidov It would be also useful to include some recommendation about index granularity for the queries. E.g. when the most common and frequent query touches only few rows, but had to scan much more rows for that, we can suggest: "If you want this query to be faster you can change your table index granularity ... | 2018-02-22T15:31:16 | 2018-02-22T15:31:16 | {
"+1": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 364,486,881 | 1,881 | zhang2014 | @alexey-milovidov Please help review the changes | 2018-02-09T16:39:05 | 2018-02-09T16:39:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 365,114,507 | 1,771 | alexey-milovidov | @dvahram
We should continue. | 2018-02-13T00:52:57 | 2018-02-13T00:52:57 | {} | MEMBER |
ClickHouse | ClickHouse | 365,584,601 | 1,569 | ztlpn | The problem is probably the trailing line break. Re-create it with `echo -n 1 > /opt/clickhouse/data/af_0/events_partition/format_version.txt`.
I should fix it properly. | 2018-02-14T11:57:56 | 2018-02-14T11:58:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 365,646,421 | 1,633 | silviucpp | Any news on this ? | 2018-02-14T15:41:16 | 2018-02-14T15:41:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 366,200,163 | 1,908 | javisantana | what do you mean with `arrayCumReduce`? something like cumsum but more generic?:
```
:) select arrayCumReduce(x, y -> y - x, [1, 2, 3, 4]
[1, 1, 1, 1]
```
I was thinking about the reverse method:
```
:) select arrayDiff([100, 102, 103, 104])
[100, 1, 1, 1]
```
Other things I think would be useful are ... | 2018-02-16T10:35:52 | 2018-02-16T10:35:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 366,314,956 | 1,916 | alexey-milovidov | Ok. | 2018-02-16T18:14:02 | 2018-02-16T18:14:02 | {} | MEMBER |
ClickHouse | ClickHouse | 366,717,012 | 1,813 | mikhailov-jet | Removal of replace_running_query fixed this problem. | 2018-02-19T14:56:55 | 2018-02-19T14:58:22 | {} | NONE |
ClickHouse | ClickHouse | 364,300,409 | 320 | alexey-milovidov | Yes. Thanks!
(It was already closed automatically.) | 2018-02-09T00:54:26 | 2018-02-09T00:54:58 | {} | MEMBER |
ClickHouse | ClickHouse | 364,925,258 | 1,863 | ludv1x | Indeed, there is some inconsistency here. If you drop a MV (which size is 0), a real table should be dropped and checked also, but the size of the MV is only checked. | 2018-02-12T13:41:07 | 2018-02-12T13:41:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,401,397 | 1,910 | ludv1x | > why not merged?
I just opened a new PR https://github.com/yandex/ClickHouse/pull/1937 which has a more suitable branch name. | 2018-02-21T17:17:44 | 2018-02-21T17:17:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,678,134 | 1,945 | osbertngok | c.c. @sundy-li | 2018-02-22T13:19:35 | 2018-02-22T13:19:35 | {} | NONE |
ClickHouse | ClickHouse | 367,762,864 | 1,939 | alexey-milovidov | #1861 | 2018-02-22T17:47:19 | 2018-02-22T17:47:19 | {} | MEMBER |
ClickHouse | ClickHouse | 367,763,153 | 1,842 | alexey-milovidov | Fixed in master. | 2018-02-22T17:48:15 | 2018-02-22T17:48:15 | {} | MEMBER |
ClickHouse | ClickHouse | 364,615,515 | 1,881 | zhang2014 | It would be better if I could fix the problem, It helps me grow : ) | 2018-02-10T01:40:19 | 2018-02-10T01:40:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 364,687,657 | 1,882 | alexey-milovidov | You should to add a case in performance tests.
Look at `tests/performance`. | 2018-02-10T20:15:17 | 2018-02-10T20:15:17 | {} | MEMBER |
ClickHouse | ClickHouse | 365,045,829 | 1,882 | alexey-milovidov | Running performance test is as simple as
```
clickhouse performance-test .
```
in the directory with your test. | 2018-02-12T20:07:12 | 2018-02-12T20:07:12 | {} | MEMBER |
ClickHouse | ClickHouse | 366,350,901 | 1,918 | ztlpn | Unfortunately we have tagged a release candidate just before the fix was committed. In case there are no critical problems with it the next release candidate will not appear until after a ~week. | 2018-02-16T20:35:09 | 2018-02-16T20:35:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 366,451,886 | 1,908 | alexey-milovidov | Examples (along with `arrayReduce` function, that is already exist in ClickHouse):
```
arrayReduce('avg', [1, 2, 3]) = 2
arrayCumReduce('avg', [1, 2, 3]) = [1, 1.5, 2]
arrayReduce('uniq', [10, 11, 11, 12]) = 3
arrayCumReduce('uniq', [10, 11, 11, 12]) = [1, 2, 2, 3]
arrayReduce('corr', [1, 2, 3], [1, 0, 1]) ... | 2018-02-17T16:18:06 | 2018-02-17T16:18:22 | {} | MEMBER |
ClickHouse | ClickHouse | 366,505,275 | 1,908 | javisantana | Ok, I get it, makes sense.
I was thinking about (and sorry about the offtopic) working on some methods to allow easier geospatial filtering/analysis. I worked with CH in the past and worked really great (I wrote a [blogspot](https://carto.com/blog/inside/geospatial-processing-with-clickhouse/) about it which BTW i... | 2018-02-18T10:09:31 | 2018-02-18T10:09:31 | {
"+1": 4
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,736,573 | 1,943 | alexey-milovidov | :+1: | 2018-02-22T16:26:16 | 2018-02-22T16:26:16 | {} | MEMBER |
ClickHouse | ClickHouse | 365,933,538 | 1,908 | alexey-milovidov | Great! | 2018-02-15T13:54:07 | 2018-02-15T13:54:07 | {} | MEMBER |
ClickHouse | ClickHouse | 366,687,586 | 1,904 | ztlpn | Values format expects empty strings to be enclosed in single quotes, like this: `VALUES (...,'',...)`. If you want the CSV syntax for inserted data, you should use the CSV format (`INSERT INTO ... FORMAT CSV`). | 2018-02-19T13:04:40 | 2018-02-19T13:04:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,294,660 | 167 | sartor | Is there any news for updates/deletes? | 2018-02-21T11:17:08 | 2018-02-21T11:17:08 | {
"+1": 4
} | NONE |
ClickHouse | ClickHouse | 364,967,726 | 1,832 | ztlpn | This is related to the new parsing code for floats introduced in #1665 (note that empty strings for Int columns should still work). The [documentation](https://clickhouse.yandex/docs/en/formats/tabseparated.html) states that parsing of empty strings is implementation-specific, but in this case we probably should just r... | 2018-02-12T15:59:23 | 2018-02-12T15:59:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 365,114,871 | 1,710 | alexey-milovidov | First step: #1884.
(Currently it doesn't change the default behaviour, only introduces a new function) | 2018-02-13T00:55:00 | 2018-02-13T00:55:00 | {} | MEMBER |
ClickHouse | ClickHouse | 365,270,251 | 1,894 | fireya | I will try to explain from beginning.
I have multiple hierarchical dictionaries, for example:
Hierarchy1:
1
*2
*3
**4
Hierarchy2:
1
*2
3
*4
And some table with 100 millions of rows:
`CREATE table TestTable(Dim1 UInt64, Dim2 UInt64, Value UInt64)`
I need to transform value to all parents in hierarc... | 2018-02-13T13:43:56 | 2018-02-13T14:03:56 | {} | NONE |
ClickHouse | ClickHouse | 365,835,205 | 258 | zhang2014 | Can we close this issue? I saw that it was solved in the master. @alexey-milovidov | 2018-02-15T06:13:54 | 2018-02-15T06:13:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,755,007 | 1,939 | alexey-milovidov | `OPTIMIZE` query can only be executed on leader replica.
As it's difficult for user to figure out what replica is the leader right now,
when the query is received by another replica, it will be simply forwarded to leader,
using `default` user. | 2018-02-22T17:22:11 | 2018-02-22T17:22:11 | {} | MEMBER |
ClickHouse | ClickHouse | 350,528,188 | 127 | sirkon | I decided to hunt down this issue, reproduced steps and oops
This is old clickhouse version, it works and I did nothing. I am running Ubuntu 17.10. I remember though I installed libpoco yesterday, together wi... | 2017-12-10T06:27:02 | 2018-02-09T19:54:16 | {} | NONE |
ClickHouse | ClickHouse | 364,963,852 | 1,886 | alexey-milovidov | Please provide an example of usage. | 2018-02-12T15:48:46 | 2018-02-12T15:48:46 | {} | MEMBER |
ClickHouse | ClickHouse | 365,535,182 | 1,569 | ivankosianenko | Hi @harlinb
We are created format_version.txt
but I didn't help us:
<Error> Application: DB::Exception: Cannot create table from metadata file /opt/clickhouse/metadata/af_0//events_partition.sql, error: DB::Exception: Bad version file:
/opt/clickhouse//data/af_0/events_partition/format_version.txt, stack trace:
... | 2018-02-14T08:47:58 | 2018-02-14T08:47:58 | {} | NONE |
ClickHouse | ClickHouse | 365,559,025 | 797 | ishirav | +1 we encountered a similar problem after adding nullable fields.
(on version 1.1.54140) | 2018-02-14T10:21:07 | 2018-02-14T10:32:16 | {} | NONE |
ClickHouse | ClickHouse | 365,942,983 | 1,908 | alexey-milovidov | Looking forward for `arrayCumReduce` function :) | 2018-02-15T14:28:19 | 2018-02-15T14:28:19 | {} | MEMBER |
ClickHouse | ClickHouse | 366,334,295 | 1,914 | ztlpn | The bug first appeared in 54276 (seems to be related to #994). | 2018-02-16T19:25:18 | 2018-02-16T19:25:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,191,709 | 863 | zhang2014 | What is the current state of this issue? I plan to implement it. @alexey-milovidov | 2018-02-21T02:18:18 | 2018-02-21T02:18:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,513,506 | 1,933 | alexey-milovidov | Thanks! The idea is fine.
You should also adjust failed tests. | 2018-02-21T23:26:36 | 2018-02-21T23:26:46 | {} | MEMBER |
ClickHouse | ClickHouse | 367,684,564 | 1,933 | zhang2014 | please help review the changes @alexey-milovidov | 2018-02-22T13:43:41 | 2018-02-22T13:43:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,741,870 | 1,945 | alexey-milovidov | And it is slightly outdated.
ClickHouse has support to perform GROUP BY and ORDER BY even if intermediate data doesn't fit in RAM. You can find it in documentation as "GROUP BY in external memory" https://clickhouse.yandex/docs/en/query_language/queries/#group-by-in-external-memory and in "ORDER BY clause" https://c... | 2018-02-22T16:41:40 | 2018-02-22T16:44:32 | {} | MEMBER |
ClickHouse | ClickHouse | 364,386,072 | 728 | zhang2014 | @luc1ph3r Can we close this issue? Because #320 has been fixed. | 2018-02-09T09:47:51 | 2018-02-09T09:47:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 365,350,048 | 1,894 | alexey-milovidov | > can trasform function support arrays as U type? This is the best and simplest solution for me.
It is not so hard to implement.
> is there any other method i haven't tried?
You can combine array-of-arrays subscript and transform function: first transform to indices (1..n), then do array subscript.
> can pe... | 2018-02-13T17:57:57 | 2018-02-13T17:57:57 | {} | MEMBER |
ClickHouse | ClickHouse | 366,330,751 | 1,914 | ztlpn | Confirmed, thanks! | 2018-02-16T19:11:19 | 2018-02-16T19:11:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 366,341,412 | 1,918 | bobrik | @alexey-milovidov, when do you plan to cut the next stable release? I wonder if we should just wait for that and avoid extra restart round. | 2018-02-16T19:53:56 | 2018-02-16T19:53:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 366,848,367 | 1,059 | zhang2014 | This problem is the same as #320, which is caused by `rewriteExpressionList` in subQuery.I've fixed it in #1763. Can we close this issue? @proller | 2018-02-20T02:17:21 | 2018-02-20T02:17:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,513,869 | 1,933 | alexey-milovidov | Also it's good idea to add a test with isolated case specifically for this issue. | 2018-02-21T23:28:13 | 2018-02-21T23:28:13 | {} | MEMBER |
ClickHouse | ClickHouse | 365,115,203 | 1,888 | alexey-milovidov | The test should be conditionally disabled: https://travis-ci.org/yandex/ClickHouse/jobs/340684606 | 2018-02-13T00:56:53 | 2018-02-13T00:56:53 | {} | MEMBER |
ClickHouse | ClickHouse | 365,276,020 | 1,732 | zero-master | They promised the opensource it but kept at it. | 2018-02-13T14:05:41 | 2018-02-13T14:05:41 | {} | NONE |
ClickHouse | ClickHouse | 367,258,673 | 863 | zhang2014 | Thank for you reply. I will try to make a pr.
BWT, Could you please tell me more about the implementation details of the new UNION ALL? or can you tell me about the branch that you're working on? I only found `better-union-all` and `union-all-preparation`, but they were relatively old. | 2018-02-21T09:08:19 | 2018-02-21T09:08:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 367,758,291 | 1,931 | alexey-milovidov | 03c3b18 This will help. | 2018-02-22T17:32:35 | 2018-02-22T17:32:35 | {} | MEMBER |
ClickHouse | ClickHouse | 364,226,217 | 117 | alexey-milovidov | @zhang2014 has fixed everything except `SHOW CREATE TABLE`. | 2018-02-08T19:44:09 | 2018-02-08T19:44:09 | {} | MEMBER |
ClickHouse | ClickHouse | 364,615,275 | 180 | zhang2014 | Can we close this issue? @alexey-milovidov | 2018-02-10T01:37:36 | 2018-02-10T01:37:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 365,528,611 | 1,190 | alexey-milovidov | The crash was fixed long time ago but the server is still trying to read temporary files. | 2018-02-14T08:19:15 | 2018-02-14T08:19:15 | {} | MEMBER |
ClickHouse | ClickHouse | 365,942,574 | 1,908 | alexey-milovidov | I've tried this variant of inner loop:
```
size_t pos = 0;
for (size_t i = 0; i < offsets.size(); ++i)
{
Result sum{};
for (; pos < offsets[i]; ++pos)
{
sum += data[pos];
res_values[pos] = sum;
}
... | 2018-02-15T14:26:54 | 2018-02-15T14:26:54 | {} | MEMBER |
ClickHouse | ClickHouse | 366,337,326 | 1,918 | alexey-milovidov | To avoid this issue, you can postpone upgrade until the next version will be released.
Another possibility is to backport this patch a074b77 and rebuild ClickHouse. | 2018-02-16T19:37:37 | 2018-02-16T19:37:37 | {} | MEMBER |
ClickHouse | ClickHouse | 367,514,128 | 1,633 | alexey-milovidov | Give me few days please. | 2018-02-21T23:29:26 | 2018-02-21T23:29:26 | {} | MEMBER |
ClickHouse | ClickHouse | 367,524,918 | 703 | alexey-milovidov | This crash was fixed long time ago. | 2018-02-22T00:24:56 | 2018-02-22T00:24:56 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.