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 | 325,315,662 | 1,118 | dulong | Hi @alexey-milovidov , is there any plan to implement this feature ? | 2017-08-28T10:13:17 | 2017-08-28T10:13:17 | {} | NONE |
ClickHouse | ClickHouse | 325,423,165 | 1,158 | vadimtk | The full dataset is http://files.pushshift.io/reddit/comments/
It is 2.86 billion rows , so I can give only link to original files.
On Mon, Aug 28, 2017 at 10:36 AM, alexey-milovidov <notifications@github.com
> wrote:
> Yes, quantileTDigest should be more accurate than quantile.
> Your result seems to be very off. Co... | 2017-08-28T17:40:36 | 2017-08-28T17:40:36 | {} | NONE |
ClickHouse | ClickHouse | 325,434,106 | 1,158 | alexey-milovidov | Thank you for the link! It should be interesting dataset.
I have run download... As I see, it will take about two days. | 2017-08-28T18:17:12 | 2017-08-28T18:17:12 | {} | MEMBER |
ClickHouse | ClickHouse | 325,440,626 | 1,154 | alexey-milovidov | `toRelativeDayNum(toDateTime('2017-08-01 00:00:00'))` is a constant expression.
All constant expressions are computed only once. | 2017-08-28T18:37:42 | 2017-08-28T18:37:42 | {} | MEMBER |
ClickHouse | ClickHouse | 325,826,262 | 1,168 | vavrusa | I understand there are some other possible ways how to create a loop, for example using `Merge` engine, so perhaps the recommendation would be to make such query killable without having to restart the whole cluster. HTTP interface didn't seem to have many connections, but still rejected all queries due to the limit for... | 2017-08-29T22:47:09 | 2017-08-29T22:47:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 323,412,023 | 1,121 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-18T17:22:01 | 2017-08-18T17:22:01 | {} | NONE |
ClickHouse | ClickHouse | 324,243,517 | 1,132 | flowbehappy | It turns out is my problem on `clickhouse-client` alias script. Sorry about that. | 2017-08-23T07:19:07 | 2017-08-23T07:19:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,315,606 | 1,062 | silviucpp | Ok then, if the decision that Nullable array are not supported is final then I don't see any option.. | 2017-08-23T12:36:39 | 2017-08-23T12:36:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,606,323 | 1,144 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-24T11:14:02 | 2017-08-24T11:14:02 | {} | NONE |
ClickHouse | ClickHouse | 324,732,680 | 379 | silviucpp | @ztlpn thanks ! now it's clear. probably you can close the issue as well | 2017-08-24T19:24:55 | 2017-08-24T19:24:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 325,567,733 | 1,067 | silviucpp | @alexey-milovidov thanks for explanation, maybe this can be added in documentation because might be a good hint for others as well | 2017-08-29T06:28:22 | 2017-08-29T06:28:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 326,366,646 | 1,177 | sunsingerus | Figured out what was that - have MariaDB client lib, and they have no `libmysqlclient.a` file included into `MariaDB-devel` package any more. Interesting that there is this [commit](https://github.com/yandex/ClickHouse/pull/1181/commits/41a8c1412189594d5c6e2d8b0191ee27a3a79d85) right from today. Think we can close this... | 2017-08-31T17:29:13 | 2017-08-31T17:29:13 | {} | NONE |
ClickHouse | ClickHouse | 326,934,174 | 1,191 | jackpgao | Found the problem.
I have 3 node: A, B, C.
And I have a basic table named 'table' with MergeTree engine and a distributed table named 'table_all' with distributed engine .
In the test environment, I build a test table named 'xx_test' and 'xx_test_all' on only one node, such as A, and I miss the different eng... | 2017-09-04T11:00:41 | 2017-09-04T11:00:41 | {} | NONE |
ClickHouse | ClickHouse | 323,923,076 | 1,131 | vavrusa | It's not a primary key like you expect in OLTP systems, it's more like a sorting key as I said. You can read about how it works here https://medium.com/@f1yegor/clickhouse-primary-keys-2cf2a45d7324 | 2017-08-22T05:29:29 | 2017-08-22T05:29:29 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,150,656 | 1,094 | eyalroz | It was the revision you had up when I filed the report. Ok ok, it's: 547516c6369f58170c9674862d4bcf64d26835d1
CMakeCache.txt is [here](https://pastebin.ca/3858013); CMakeOutput.log is [here](https://pastebin.ca/3858014); CMakeError.log is [here](https://pastebin.ca/3858015).
| 2017-08-22T21:02:01 | 2017-08-22T21:02:01 | {} | NONE |
ClickHouse | ClickHouse | 324,644,004 | 1,143 | innerr | Thanks a lot! | 2017-08-24T14:01:29 | 2017-08-24T14:01:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,711,060 | 1,142 | ztlpn | Thanks! Confirmed. | 2017-08-24T18:00:47 | 2017-08-24T18:00:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,743,075 | 1,124 | ztlpn | There is no ReplacingAggregatingMergeTree. Consider how AggregatingMergeTree is implemented: it stores the _state_ of the aggregate functions (something that can be later converted to actual value) and the states for the same primary key are then _merged_ in the background. No information about which versions went into... | 2017-08-24T20:08:37 | 2017-08-24T20:08:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,894,090 | 1,153 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-25T11:37:02 | 2017-08-25T11:37:02 | {} | NONE |
ClickHouse | ClickHouse | 325,485,255 | 1,134 | alexey-milovidov | If you have 0 or 1 values, you could get expected behaviour with:
`max(x)` for or
`min(x)` for and
`sum(x) % 2` for xor. | 2017-08-28T21:19:25 | 2017-08-28T21:19:25 | {} | MEMBER |
ClickHouse | ClickHouse | 326,079,790 | 1,028 | alexey-milovidov | Yes, exactly. | 2017-08-30T18:32:56 | 2017-08-30T18:32:56 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 326,984,456 | 1,128 | ztlpn | Thanks, I was able to reproduce - the problem is that the INSERT query does not acquire lock for the .inner. table, so it could be dropped while the insertion was in progress. | 2017-09-04T15:09:12 | 2017-09-04T15:09:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 323,575,988 | 580 | AlekSi | Wow. Maybe that could be better documented? | 2017-08-20T10:17:14 | 2017-08-20T10:17:14 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 323,662,944 | 1,060 | hyd998877 | thank you | 2017-08-21T07:03:31 | 2017-08-21T07:03:31 | {} | NONE |
ClickHouse | ClickHouse | 324,607,287 | 1,092 | bobrik | @alexey-milovidov
We use cache for all 3 dynamic dictionaries:
```
<lifetime>
<min>86400</min>
<max>172800</max>
</lifetime>
```
```
<lifetime>
<min>1800</min>
<max>3600</max>
</lifetime>
```
```
<lifetime>
<min>1800</min>
<max>3600</max>
... | 2017-08-24T11:19:03 | 2017-08-24T11:19:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,894,414 | 1,153 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-25T11:39:02 | 2017-08-25T11:39:02 | {} | NONE |
ClickHouse | ClickHouse | 325,423,466 | 1,157 | alexey-milovidov | Ok.
Additional details:
1. ClickHouse could work without `clang` as it is only used for `compile` feature, that is disabled by default. If you build a package without clang, it will work fine unless you enable that option.
2. It is recommended to use at least clang 4. I doubt that your system is enough fresh t... | 2017-08-28T17:41:42 | 2017-08-28T17:41:42 | {} | MEMBER |
ClickHouse | ClickHouse | 326,967,786 | 1,193 | proller | what wersion you use?
your problem should be fixed after 1a8e22c37f54d18b2162b2048db5171bca4d8594 | 2017-09-04T13:48:11 | 2017-09-04T13:48:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,026,017 | 1,094 | ztlpn | Hi, @eyalroz ,
Which revision of ClickHouse are you building?
Could you share your cmake console output? | 2017-08-22T13:28:32 | 2017-08-22T13:28:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,110,497 | 1,037 | ztlpn | Sorry for misinformation - the fix was not included in the 1.1.54276 release. It is in master now and will be included in the next release. | 2017-08-22T18:20:47 | 2017-08-22T18:20:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,268,397 | 1,073 | kshvakov | Спасибо | 2017-08-23T09:04:59 | 2017-08-23T09:04:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,395,229 | 1,140 | ztlpn | Здравствуйте! По какой-то причине не может выполниться вызов statvfs(<путь до данных таблицы>).
Проверьте, всё ли в порядке с директорией `/var/lib/clickhouse/data/<database>/<table>`. Также проблема может возникнуть из-за особенностей файловой системы контейнера (пример: https://forums.docker.com/t/mounting-volume-... | 2017-08-23T16:48:03 | 2017-08-23T16:48:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,409,291 | 1,140 | vmanyushin | да, действительно, мой косяк..... | 2017-08-23T17:40:35 | 2017-08-23T17:40:35 | {} | NONE |
ClickHouse | ClickHouse | 324,615,254 | 379 | silviucpp | I'm a bit confused but having toStartOf* how I can group results for example every 3 months based on a DateTime column... | 2017-08-24T12:00:34 | 2017-08-24T12:00:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,837,800 | 1,152 | kszucs | Thanks @Slach! I'm using the native driver.
I've recently find out, that clickhouse has NULL, isNull, Nullable, coalesce, which is undocumented.
How about `LIMIT BY`? Cumulative functions other than `runningDifference`?
Are they stable, released?
Right now I'm trying to implement a couple of analytical functions... | 2017-08-25T06:48:45 | 2017-08-25T07:39:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 325,487,383 | 1,067 | alexey-milovidov | Yes, this is expected, because each batch of data for INSERT is split to batches by partitions. And partitions are months.
There are two possible workarounds:
- use even larger batches (example: 100k instead of 10k);
- group data by months before insertion.
This is not an issue in typical use cases when you:
-... | 2017-08-28T21:27:30 | 2017-08-28T21:27:30 | {} | MEMBER |
ClickHouse | ClickHouse | 326,958,423 | 845 | ludv1x | Исправится в ближайшем релизе.
Заодно появится запрос `SYSTEM DROP DNS CACHE`. | 2017-09-04T13:03:35 | 2017-09-04T13:03:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 323,411,305 | 1,121 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-18T17:19:02 | 2017-08-18T17:19:02 | {} | NONE |
ClickHouse | ClickHouse | 323,443,033 | 1,122 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-18T19:37:02 | 2017-08-18T19:37:02 | {} | NONE |
ClickHouse | ClickHouse | 323,577,200 | 1,125 | inkrement | ok, now I know why this happens. I already had the same problem (replacing/re-creating a materialised view on which another view is based does not work) #1112 ... In this former bug report @ludv1x mentioned that it is similar to #598, which is already closed. But it is just related and the "you have to add a engine"-c... | 2017-08-20T10:44:09 | 2017-08-20T10:44:09 | {} | NONE |
ClickHouse | ClickHouse | 324,397,653 | 1,141 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-23T16:57:02 | 2017-08-23T16:57:02 | {} | NONE |
ClickHouse | ClickHouse | 324,398,186 | 1,141 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-23T16:59:02 | 2017-08-23T16:59:02 | {} | NONE |
ClickHouse | ClickHouse | 325,414,214 | 1,160 | alexey-milovidov | There is a reason. Look at:
https://clickhouse.yandex/docs/en/operations/settings/settings.html#output-format-json-quote-64bit-integers
You could get desired behaviour by setting `output_format_json_quote_64bit_integers` to 0. | 2017-08-28T17:08:48 | 2017-08-28T17:09:05 | {} | MEMBER |
ClickHouse | ClickHouse | 326,043,852 | 1,170 | alexey-milovidov | How to build: https://github.com/yandex/ClickHouse/blob/master/docs/en/development/build.rst | 2017-08-30T16:22:39 | 2017-08-30T16:22:39 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 323,230,041 | 1,066 | vavrusa | There's one case where neither address nor user is logged:
```
2017.08.18 00:31:46.819654 [ 182 ] <Error> ServerErrorHandler: Code: 101, e.displayText() = DB::NetException: Unexpected packet from client, e.what() = DB::NetException, Stack trace:
```
It's possibly garbage or mistake, but it'd be much easier to d... | 2017-08-18T00:40:43 | 2017-08-18T00:40:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 323,921,591 | 1,131 | shailee-m | But why should it duplicate in the first place? Does primary key mean something else in this context?
Here, the primary key is date,id which is the same for both the entries. That shouldn't happen. | 2017-08-22T05:17:32 | 2017-08-22T05:17:32 | {} | NONE |
ClickHouse | ClickHouse | 323,923,114 | 1,131 | shailee-m | Alright! ReplacingMergeTree works. but I still have to optimise it for de-duplication. | 2017-08-22T05:29:50 | 2017-08-22T05:29:50 | {} | NONE |
ClickHouse | ClickHouse | 323,072,489 | 1,060 | ludv1x | It's correct behaviour.
You just forgot to take into account background merges of AggregatingMergeTree.
The query
```
SELECT
NUMBER_AREA,
uniqMerge(UNIQ_DATA) AS uniq
FROM uniq_tb
GROUP BY NUMBER_AREA
```
eventually (after all background merges) works as:
```
SELECT -- here is outer select of y... | 2017-08-17T13:24:44 | 2017-08-23T10:28:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,356,077 | 1,125 | ludv1x | > A simple workaround is to delete the view which is based on the "new" (in my case I had to delete all views based on comments_videos) one and recreate it.
Help me reproduce the problem.
Had you had any ordinary (or MATERIALIZED) VIEWs making SELECTs from `comments_videos` before you (re)created `comments_videos`... | 2017-08-23T14:39:21 | 2017-08-23T14:39:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,450,053 | 1,125 | inkrement | Just MATERIALIZED VIEWs and yes I executed some SELECTs before. I already posted the show create table output in the first post or do you need it from the original tables as well? | 2017-08-23T20:17:09 | 2017-08-23T20:17:09 | {} | NONE |
ClickHouse | ClickHouse | 324,825,266 | 1,152 | Slach | What help need you ?
you can use native python driver https://github.com/mymarilyn/clickhouse-driver to connect to clickhouse and execute queries
| 2017-08-25T05:09:57 | 2017-08-25T05:10:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 325,428,562 | 1,157 | alexey-milovidov | And third (and most important):
Clang from Debian/Ubuntu package is built with dependence of its own shared libraries (libllvm.so, libclang.so). This is not suitable for ClickHouse, because ClickHouse doesn't add these shared libraries into package. When you build clang by yourself according to the instruction (http... | 2017-08-28T17:59:32 | 2017-08-28T18:00:12 | {} | MEMBER |
ClickHouse | ClickHouse | 325,621,127 | 1,146 | KochetovNicolai | autotest | 2017-08-29T10:19:38 | 2017-08-29T10:19:38 | {} | MEMBER |
ClickHouse | ClickHouse | 326,523,338 | 1,183 | KochetovNicolai | autotest | 2017-09-01T08:33:32 | 2017-09-01T08:33:32 | {} | MEMBER |
ClickHouse | ClickHouse | 323,442,231 | 1,122 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-18T19:34:09 | 2017-08-18T19:34:09 | {} | NONE |
ClickHouse | ClickHouse | 323,920,891 | 1,131 | vavrusa | MergeTree doesn't deduplicate data, it just sorts the rows by the primary key expression and compacts the data parts in the background. Perhaps you might want to take look at the ReplacingMergeTree or AggregatingMergeTree ? | 2017-08-22T05:11:58 | 2017-08-22T05:11:58 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,242,749 | 1,132 | flowbehappy | Commit: fc409f61e53646d2e88a57287496b67de546bca2 on master.
macOS: 10.12.5.
Using the default Terminal. | 2017-08-23T07:15:22 | 2017-08-23T07:15:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 325,511,858 | 1,157 | hnakamur | Thanks for additional details. I understand. | 2017-08-28T23:34:12 | 2017-08-28T23:34:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 325,515,808 | 1,158 | vadimtk | @alexey-milovidov yes it is interesting dataset, and it is good for ClickHouse.
I would suggest to host it on Yandex.Disk by Yandex and make it available for all ClickHouse users, as there is no guarantee that original link will be available long-term | 2017-08-29T00:00:56 | 2017-08-29T00:00:56 | {} | NONE |
ClickHouse | ClickHouse | 325,642,482 | 1,161 | igor-lukyanov | Thank you, closing it then, sorry for the duplicate. | 2017-08-29T12:03:17 | 2017-08-29T12:03:17 | {} | NONE |
ClickHouse | ClickHouse | 325,645,893 | 1,160 | silviucpp | Thanks,
I'll close this then | 2017-08-29T12:18:48 | 2017-08-29T12:18:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 326,043,694 | 1,170 | alexey-milovidov | You have an old CPU without support of SSE 4.2 instruction set.
ClickHouse can work with your CPU, but pre-built packages require newer CPU.
To run ClickHouse on your machine, you have to build it manually. | 2017-08-30T16:21:59 | 2017-08-30T16:21:59 | {} | MEMBER |
ClickHouse | ClickHouse | 323,339,095 | 1,110 | drkop | So i don't understand why stable won't compile, but doing this:
ln -s libs/libmysqlxx/include/mysqlxx build/contrib/libzlib-ng/mysqlxx
got an a result. | 2017-08-18T12:21:35 | 2017-08-18T12:21:35 | {} | NONE |
ClickHouse | ClickHouse | 323,739,661 | 1,037 | ztlpn | @ludv1x fixed DateTime CSV parsing in #1098
The fix is available in 1.1.54276 release. | 2017-08-21T13:04:41 | 2017-08-21T13:04:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,631,017 | 1,141 | alexey-milovidov | You should also use `getConfigRef` in all places, where configuration is used and `Context` is available. | 2017-08-24T13:13:57 | 2017-08-24T13:13:57 | {} | MEMBER |
ClickHouse | ClickHouse | 325,117,725 | 51 | ramazanpolat | Hello @rasmus93,
The following workaround will also work for AVG, SUM and other aggregate functions.
```
SELECT *
FROM
(
SELECT count(*) AS count
FROM mytable
UNION ALL
SELECT toUInt64 (0)
)
LIMIT 1
``` | 2017-08-26T11:32:20 | 2017-08-26T11:32:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 326,305,026 | 1,178 | alexey-milovidov | If you use Replicated tables, they will deduplicate inserted blocks of data:
-- if exactly identical block of data (same rows in same order) was inserted twice, it will be effectively inserted once.
See:
https://clickhouse.yandex/docs/en/table_engines/replication.html?highlight=deduplicated
(after "Blocks of da... | 2017-08-31T14:02:47 | 2017-08-31T14:03:02 | {
"+1": 3
} | MEMBER |
ClickHouse | ClickHouse | 323,737,408 | 931 | ztlpn | Implemented and turned on by default in 1.1.54276. | 2017-08-21T12:55:00 | 2017-08-21T12:55:00 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 323,737,547 | 999 | ztlpn | Fixed in 1.1.54276. | 2017-08-21T12:55:37 | 2017-08-21T12:55:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,606,899 | 1,144 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-24T11:17:02 | 2017-08-24T11:17:02 | {} | NONE |
ClickHouse | ClickHouse | 324,939,197 | 1,154 | silviucpp | Thanks,
This was the problem | 2017-08-25T14:36:52 | 2017-08-25T14:36:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 325,422,170 | 1,158 | alexey-milovidov | Yes, quantileTDigest should be more accurate than quantile.
Your result seems to be very off. Could you provide full dataset of values for investigation?
PS. The behaviour is also depend on how quantile is defined on discrete sets.
Look here for different variants:
https://en.wikipedia.org/wiki/Quantile#Estimat... | 2017-08-28T17:36:41 | 2017-08-28T17:36:41 | {} | MEMBER |
ClickHouse | ClickHouse | 323,734,611 | 1,067 | silviucpp | Regarding insert performance I observed that in case you insert data in a ReplicatedMergeTree where EventDate it's spread on 3-4 years the speed is pretty good with batches of 10k records.. But if the same records I spread them in 20 years.. it's at least 100 times slower.. | 2017-08-21T12:42:16 | 2017-08-21T12:42:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,629,576 | 1,141 | alexey-milovidov | You can have many clickhouse-servers on the same host, that listen to different ports.
And the address with different port is not local - if you send a query to that address, the query will be received by another instance of clickhouse-server. | 2017-08-24T13:07:54 | 2017-08-24T13:07:54 | {} | MEMBER |
ClickHouse | ClickHouse | 324,935,738 | 1,154 | ztlpn | As the reference date of `toRelativeDayNum()` is unspecified, if you want to align your intervals to a particular date, you need to do it manually, e.g. by grouping by `intDiv(toRelativeDayNum(date) - toRelativeDayNum(toDateTime('2017-08-01 00:00:00')), 2)` | 2017-08-25T14:25:32 | 2017-08-25T14:25:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 325,172,465 | 1,157 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-27T02:02:02 | 2017-08-27T02:02:02 | {} | NONE |
ClickHouse | ClickHouse | 325,172,528 | 1,157 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-27T02:04:02 | 2017-08-27T02:04:02 | {} | NONE |
ClickHouse | ClickHouse | 325,440,079 | 1,161 | alexey-milovidov | This issue is fixed here: https://github.com/yandex/ClickHouse/pull/1117
Fix is scheduled for next release. | 2017-08-28T18:35:36 | 2017-08-28T18:35:36 | {} | MEMBER |
ClickHouse | ClickHouse | 326,087,088 | 1,028 | silviucpp | Thanks a lot :) waiting for your fix | 2017-08-30T18:59:55 | 2017-08-30T18:59:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 326,961,132 | 1,190 | KochetovNicolai | Thank you for report. | 2017-09-04T13:16:34 | 2017-09-04T13:16:34 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 326,983,617 | 1,193 | marklit | I made it past the above error message. I've raised #1195 for the mariadb-related issues I'm now facing. | 2017-09-04T15:04:53 | 2017-09-04T15:04:53 | {} | NONE |
ClickHouse | ClickHouse | 325,124,407 | 1,154 | silviucpp | @ztlpn one more question here : there is any way I can optimize the toRelativeDayNum(toDateTime('2017-08-01 00:00:00')) to be computed only once and not for every single row that's processed ? | 2017-08-26T12:51:04 | 2017-08-26T12:51:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 325,485,508 | 1,134 | alexey-milovidov | If you want bitwise aggregate functions - currently there are no such functions, but this is very easy to implement. | 2017-08-28T21:20:25 | 2017-08-28T21:20:34 | {} | MEMBER |
ClickHouse | ClickHouse | 323,576,205 | 580 | AlekSi | > You could also run client as clickhouse --client.
Not in the current Docker image:
```
$ docker exec -ti clickhouse_1 /usr/bin/clickhouse --client
ClickHouse client version 1.1.54276.
Connecting to localhost:9000.
Connected to ClickHouse server version 1.1.54276.
Poco::Exception: File not found: /nonexis... | 2017-08-20T10:22:07 | 2017-08-20T10:23:55 | {
"+1": 5
} | NONE |
ClickHouse | ClickHouse | 323,986,033 | 838 | AlekSi | It will be useful for https://github.com/Percona-Lab/PromHouse | 2017-08-22T10:27:48 | 2017-08-22T10:27:48 | {} | NONE |
ClickHouse | ClickHouse | 326,085,860 | 1,134 | alexey-milovidov | https://github.com/yandex/ClickHouse/commit/1531268281b4b63640dca4ba24274ef4f918d570 | 2017-08-30T18:55:15 | 2017-08-30T18:55:15 | {
"+1": 2
} | MEMBER |
ClickHouse | ClickHouse | 324,375,267 | 1,138 | excitoon | Tests have passed, except new ones. | 2017-08-23T15:37:56 | 2017-08-24T13:37:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,637,786 | 1,144 | hnakamur | Thanks for merging! | 2017-08-24T13:39:12 | 2017-08-24T13:39:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 325,893,990 | 1,134 | AlexMAS | Thanks for your answer. Yes, I have some `flags` column which contains integer values (powers of 2) and I need a bitwise OR function to aggregate these values. Currently I use mathematical expression. But it will be cool if you implement bitwise aggregate functions in one of further releases. So this issue is an improv... | 2017-08-30T06:22:54 | 2017-08-30T06:22:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 326,154,012 | 1,177 | proller | can you show cmake console output ? | 2017-08-31T00:18:25 | 2017-08-31T00:18:25 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,107,307 | 1,132 | ztlpn | That's very strange - I couldn't reproduce it. Can you tell which version of ClickHouse and macOS are you using? | 2017-08-22T18:09:00 | 2017-08-22T18:09:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,633,783 | 1,144 | alexey-milovidov | Thanks! | 2017-08-24T13:24:41 | 2017-08-24T13:24:41 | {} | MEMBER |
ClickHouse | ClickHouse | 324,721,430 | 379 | ztlpn | You can group by `intDiv(toRelativeMonthNum(date), 3)`. | 2017-08-24T18:40:41 | 2017-08-24T18:40:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 326,972,962 | 1,193 | marklit | I was using the stable branch. I'll give ``1a8e22c`` a try and report back. | 2017-09-04T14:12:45 | 2017-09-04T14:12:45 | {} | NONE |
ClickHouse | ClickHouse | 326,987,041 | 1,195 | proller | https://github.com/yandex/ClickHouse/pull/1185 | 2017-09-04T15:22:37 | 2017-09-04T15:22:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 323,875,947 | 1,126 | vavrusa | I think it depends on your data. You're trading off memory that can be otherwise used for page cache. Also the lookup time in the sparse index isn't constant, but `log(N)` I think, so the bigger the index, the more time it takes to search it. Scanning 8191 unnecessary rows is not that expensive (when reading from disk,... | 2017-08-21T23:12:17 | 2017-08-21T23:12:17 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,619,408 | 1,143 | ztlpn | Here is one way to do it:
First, find the translation unit where the code of the function in question ends up and the corresponding object file. In our case it is `build/dbms/src/Functions/CMakeFiles/clickhouse_functions.dir/registerFunctionLeast.cpp.o` (this is a bit unusual – recently the giant `FunctionsArithmeti... | 2017-08-24T12:21:53 | 2017-08-24T12:21:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 324,753,391 | 1,151 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-24T20:49:02 | 2017-08-24T20:49:02 | {} | NONE |
ClickHouse | ClickHouse | 324,754,152 | 1,151 | robot-metrika-test | Can one of the admins verify this patch? | 2017-08-24T20:52:02 | 2017-08-24T20:52:02 | {} | NONE |
ClickHouse | ClickHouse | 326,075,341 | 1,028 | silviucpp | @alexey-milovidov at least you can confirm that behaviour will be the one you described above: accumulate non-NULL values from array and return non-nullable Array of non-nullable type.
Thanks,
Silviu | 2017-08-30T18:16:50 | 2017-08-30T18:16:50 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.