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 | 567,218,448 | 8,278 | den-crane | > ENGINE = AggregatingMergeTree() ... ORDER BY (date_time, item_id)
Should be ORDER BY (date_time, item_id, **value**)
In your case value have been corrupted during merge -- reduced by `any` function over (date_time, item_id)
---
You don't need AggregatingMergeTree + countState -- it's overkill
Summing... | 2019-12-18T21:21:19 | 2019-12-18T21:22:59 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,429,639 | 8,271 | alesapin | Binary became smaller for about 200 MB.
On master:
```
$ du -b clickhouse
2799382288 clickhouse
```
In this PR:
```
$ du -b clickhouse
2560651696 clickhouse
``` | 2019-12-19T10:23:35 | 2019-12-19T10:23:35 | {} | MEMBER |
ClickHouse | ClickHouse | 567,542,820 | 7,969 | chukynax | That's the way i create table
also i was played with
kafka_num_consumers | 1 for 1 server, 1 for 3 servers, 2 for 2 servers and etc
CREATE TABLE php_queue on cluster magic (
eventime Datetime,
id UInt64,
message String
) ENGINE = Kafka SETTINGS kafka_broker_list = 'click1:9092,click2:90... | 2019-12-19T15:42:51 | 2019-12-19T15:43:25 | {} | NONE |
ClickHouse | ClickHouse | 566,671,585 | 8,027 | alexey-milovidov | We have no packages for AArch64, just a binary.
You can manually unpack a package for x86_64 (to create directory structure) and then replace all binaries to AArch64 ones. | 2019-12-17T17:43:55 | 2019-12-17T17:43:55 | {} | MEMBER |
ClickHouse | ClickHouse | 566,670,282 | 8,027 | alexey-milovidov | Pre-built binaries work perfectly on this server. Example:
```milovidov@arm-huawei:~$ wget https://clickhouse-builds.s3.yandex.net/0/f914e01d88e244f25a1abe70466514b8d2a4c1cf/1576543668_binary/clickhouse
--2019-12-17 17:20:59-- https://clickhouse-build... | 2019-12-17T17:40:35 | 2019-12-17T17:44:30 | {} | MEMBER |
ClickHouse | ClickHouse | 566,680,611 | 8,151 | alexey-milovidov | Next we can try to isolate clang bug and send it to maintainers.
(It can be difficult because usually code generation varies a lot when you change the code a little.) | 2019-12-17T18:06:45 | 2019-12-17T18:06:45 | {} | MEMBER |
ClickHouse | ClickHouse | 566,966,879 | 8,262 | filimonov | > it is possible to lose data
You should not lose data even in such circumstances.
The loading /staring order - indeed should be improved. | 2019-12-18T10:12:53 | 2019-12-18T10:12:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,106,958 | 7,477 | Vinatorul | I'm not sure that I am doing everything well. As I found out - I can't change `getThreadNumber` to use `CurrentThread`. I can get rid of it in most places `getThreadNumber`, but I am not sure in changes like
`DB::writeBinary(getThreadNumber(), out);` to `DB::writeBinary(CurrentThread::get().thread_number, out);` (at `... | 2019-12-18T16:27:05 | 2019-12-18T16:27:05 | {} | NONE |
ClickHouse | ClickHouse | 567,219,957 | 8,278 | den-crane | Though maybe you had another idea. Pivoted table with counts in array in a row.
It can be achieved using https://clickhouse.yandex/docs/en/operations/table_engines/summingmergetree/#nested-structures | 2019-12-18T21:25:41 | 2019-12-18T21:25:41 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,250,399 | 8,258 | alexey-milovidov | The idea to find unreferenced `.o` files to avoid compiling them has failed:
```
milovidov@example:~/ClickHouse/build_clang9$ for i in $(find contrib/llvm/ -name *.o); do nm $i | grep -q -F "$(nm dbms/CMakeFiles/dbms.dir/src/Interpreters/ExpressionJIT.cpp.o | grep 'N4llvm' | grep -P ' [UuV] ' | grep -oP '\w+$')" && e... | 2019-12-18T22:59:12 | 2019-12-18T22:59:12 | {} | MEMBER |
ClickHouse | ClickHouse | 566,565,688 | 8,237 | alesapin | `00600_replace_running_query` -- known failure? | 2019-12-17T14:31:14 | 2019-12-17T14:31:23 | {} | MEMBER |
ClickHouse | ClickHouse | 566,650,067 | 8,228 | alexey-milovidov | It looks like you have configured MySQL table actually.
```
grep -r -i mysql /etc/clickhouse-server/
grep -i mysql /etc/metrika.xml
grep -r -i mysql /var/lib/clickhouse/
``` | 2019-12-17T16:52:08 | 2019-12-17T16:52:08 | {} | MEMBER |
ClickHouse | ClickHouse | 566,800,906 | 8,257 | den-crane | Check /var/log/clickhouse-server/clickhouse-server.err.log to get the full SQL text and full error message.
```
curl 'http://localhost:8123/?query=select%201%20--hahah'
1
```
JDBC driver issue https://github.com/ClickHouse/clickhouse-jdbc/issues/208 | 2019-12-17T23:48:48 | 2019-12-17T23:52:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,400,619 | 8,285 | alexey-milovidov | Long Term Support versions are introduced for big enterprise customers who don't like to update often. LTS version is released twice a year with limited support during one year. Limited support includes backports of bugfixes (only fixes that were easy to backport or that were required by customer).
For normal custom... | 2019-12-19T09:05:07 | 2019-12-19T09:06:10 | {} | MEMBER |
ClickHouse | ClickHouse | 567,549,859 | 8,296 | den-crane | >Is it possible to introduce a special setting to get the old behaviour back
optimize_read_in_order = 0 ? | 2019-12-19T15:59:20 | 2019-12-19T15:59:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,657,053 | 8,250 | alesapin | `test_adaptive_granularity/test.py::test_version_update_two_nodes` -- flap
`00600_replace_running_query` -- stable fail. | 2019-12-17T17:08:15 | 2019-12-17T17:08:15 | {} | MEMBER |
ClickHouse | ClickHouse | 566,808,280 | 8,151 | amosbird | > Let's do it (we can specify this flag for a single source file in CMake) as a first solution.
Um it's not enough though. This is only one place of exception unwinding failure. (f.g. it also fails at https://github.com/yandex/ClickHouse/blob/master/dbms/src/Functions/IFunction.cpp#L300 when doing `select throwIf(1)... | 2019-12-18T00:17:34 | 2019-12-18T06:10:42 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 566,962,312 | 8,228 | ashkank66 | I created a file mysql_dictionary in hope of getting rid of the error, but no proper result, so I deleted it later | 2019-12-18T10:00:29 | 2019-12-18T10:00:29 | {} | NONE |
ClickHouse | ClickHouse | 567,006,110 | 8,268 | alexey-milovidov | How do you insert this data into ClickHouse?
Your field escaping is correct. But it may broke before insertion.
Looks like everything is working fine:
```
clickhouse-local --input-format JSONEachRow -S "id UInt64, obj_oid String, date DateTime, custom_data String" <<< '{"id": 480, "obj_oid": "d76a0c", "date": "20... | 2019-12-18T12:09:39 | 2019-12-18T12:09:39 | {} | MEMBER |
ClickHouse | ClickHouse | 567,032,443 | 8,268 | AutopilotNN | You made me think. The my code is incorrect! So correctly:
`echo "INSERT INTO $CH_Name.$CH_Table FORMAT JSONEachRow " | cat - calls_tmp.json | curl -u $CH_User:$CH_Pass $CH_Host --data-binary @-
` | 2019-12-18T13:32:32 | 2019-12-18T13:32:32 | {} | NONE |
ClickHouse | ClickHouse | 567,410,946 | 7,707 | stavrolia | Sorry, I was sure it isn't. Than everything is fine | 2019-12-19T09:33:23 | 2019-12-19T09:33:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,538,904 | 8,167 | alexey-milovidov | That's ok. | 2019-12-19T15:33:52 | 2019-12-19T15:33:52 | {} | MEMBER |
ClickHouse | ClickHouse | 566,402,088 | 8,027 | amosbird | @jiugem I don't have a machine so please help testing if #8243 works | 2019-12-17T06:30:49 | 2019-12-17T16:49:36 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 566,703,573 | 8,185 | KochetovNicolai | Conflicts with 19.16 | 2019-12-17T19:05:36 | 2019-12-17T19:05:36 | {} | MEMBER |
ClickHouse | ClickHouse | 566,811,297 | 8,242 | ghost | @alexey-milovidov
It's not an issue for this PR, but how do you think about https://github.com/ClickHouse/ClickHouse/pull/8022?
I made that PR two weeks ago, but nobody left any comment.
I wonder how to do that. :-) | 2019-12-18T00:29:13 | 2019-12-18T00:29:13 | {} | NONE |
ClickHouse | ClickHouse | 567,027,687 | 8,268 | AutopilotNN | /bin/sh
echo "INSERT INTO $CH_Name.$CH_Table FORMAT JSONEachRow $(cat $DIR/$CH_Name/calls_proper.json)" | curl --silent -u $CH_User:$CH_Pass $CH_Host --data-binary @-
| 2019-12-18T13:18:46 | 2019-12-18T13:18:46 | {} | NONE |
ClickHouse | ClickHouse | 567,079,116 | 8,237 | alesapin | `test_mysql_protocol/test.py::test_mysql_federated` -- flap | 2019-12-18T15:25:46 | 2019-12-18T15:25:46 | {} | MEMBER |
ClickHouse | ClickHouse | 567,384,718 | 7,799 | Slach | the workaround which ok for me is `SELECT ... SETTINGS optimize_move_to_prewhere=0;` | 2019-12-19T08:16:24 | 2019-12-19T08:16:24 | {
"laugh": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,406,986 | 7,969 | chukynax | i found another thing, i cannot add second and more queue's to one clickhouse, only one queue per server is okay or i get TOPIC's error
also i changed partition to 1 and replication factor to 3 and i get the same error, even though i have two separate queue's in one clickhouse server | 2019-12-19T09:22:56 | 2019-12-19T09:22:56 | {} | NONE |
ClickHouse | ClickHouse | 567,419,052 | 8,275 | alesapin | Fix is trivial, tests are OK (all failed are flaps). | 2019-12-19T09:54:32 | 2019-12-19T09:54:32 | {} | MEMBER |
ClickHouse | ClickHouse | 567,536,833 | 8,282 | KochetovNicolai | This message means that current replica was waiting while query is executed on leader. After some timeout is exceeded, and query wasn't executed, it had thrown this exception.
You need to find out what's happen on leader replica. You can find which replica is leader form `system.replicas` table, `is_leader` column. ... | 2019-12-19T15:28:55 | 2019-12-19T15:28:55 | {} | MEMBER |
ClickHouse | ClickHouse | 566,671,201 | 8,027 | alexey-milovidov | How to get pre-built binaries:
1. Open the list of commits: https://github.com/ClickHouse/ClickHouse/commits/master
2. Click on green check mark.
3. Click on the details link in `ClickHouse build check — 15/15 builds are OK`
4. Download `clickhouse` binary for AArch64 build. | 2019-12-17T17:42:58 | 2019-12-17T17:42:58 | {} | MEMBER |
ClickHouse | ClickHouse | 566,763,922 | 8,253 | nvartolomei | @akuzm failures not related to PR. | 2019-12-17T21:48:34 | 2019-12-17T21:48:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,935,179 | 8,247 | lutuantai95 | Thanks @den-crane . I tried it I successfully | 2019-12-18T08:45:15 | 2019-12-18T08:45:15 | {} | NONE |
ClickHouse | ClickHouse | 567,076,862 | 8,269 | alexey-milovidov | Because we will totally replace our thread number with OS thread number? | 2019-12-18T15:20:25 | 2019-12-18T15:20:25 | {} | MEMBER |
ClickHouse | ClickHouse | 567,112,857 | 8,258 | alexey-milovidov | The change in binary size is Ok:
```
milovidov@milovidov-Pro-P30:~/work/ClickHouse/dbms/programs/server$ ls -l ../../../build/dbms/programs/clickhouse
-rwxrwxr-x 1 milovidov milovidov 2006845648 дек 18 18:17 ../../../build/dbms/programs/clickhouse
milovidov@milovidov-Pro-P30:~/work/ClickHouse/dbms/programs/server... | 2019-12-18T16:40:53 | 2019-12-18T16:40:53 | {} | MEMBER |
ClickHouse | ClickHouse | 567,491,970 | 8,285 | alexey-milovidov | Yes, LTS versions are available for everyone.
They are located in the same repository, just replace `stable` to `lts`:
https://repo.yandex.ru/clickhouse/deb/lts/main/
https://repo.yandex.ru/clickhouse/rpm/lts/x86_64/ | 2019-12-19T13:35:09 | 2019-12-19T13:35:09 | {} | MEMBER |
ClickHouse | ClickHouse | 567,528,056 | 8,296 | Har01d | Possibly duplicates #6613 | 2019-12-19T15:10:30 | 2019-12-19T15:10:30 | {} | NONE |
ClickHouse | ClickHouse | 566,604,753 | 8,245 | jackpgao | It will be much more powerful if it does. | 2019-12-17T15:58:30 | 2019-12-17T15:58:30 | {} | NONE |
ClickHouse | ClickHouse | 566,872,152 | 8,258 | alexey-milovidov | Now we can use LLVM with sanitizers and with libc++.
Drawbacks:
- the build time was slowed down, we must ensure that CI will continue to work;
- if the user wants to build with system LLVM, only versions 8 and 9 are supported.
It's in question, should we enable it (`USE_EMBEDDED_COMPILER`) by default (or disab... | 2019-12-18T05:20:13 | 2019-12-18T05:21:06 | {} | MEMBER |
ClickHouse | ClickHouse | 566,981,126 | 8,228 | KochetovNicolai | > /var/lib/clickhouse/metadata/db_name.sql:ENGINE = MySQL('host:port', 'database', 'user', 'password')
That means that you have `MySQL` database with name `db_name`, which has incorrect configuration (instead of `'host:port', 'database', 'user', 'password'` must be real values). And this database can't connect to My... | 2019-12-18T10:51:56 | 2019-12-18T10:51:56 | {} | MEMBER |
ClickHouse | ClickHouse | 567,030,377 | 8,269 | alesapin | suspended | 2019-12-18T13:26:30 | 2019-12-18T13:26:30 | {} | MEMBER |
ClickHouse | ClickHouse | 567,118,273 | 7,477 | alexey-milovidov | Let's use `TaskStatsInfoGetter::getCurrentTID()`. To avoid the cost of system call, we can merge https://github.com/ClickHouse/ClickHouse/pull/8276 | 2019-12-18T16:53:43 | 2019-12-18T16:53:43 | {} | MEMBER |
ClickHouse | ClickHouse | 567,251,328 | 8,258 | alexey-milovidov | Compiled files are just 90 MB:
```
$ find contrib/llvm/ -name *.o | xargs wc -c | grep total
90898800 total
$ find contrib/llvm/ -name *.a | xargs wc -c | grep total
91132122 total
``` | 2019-12-18T23:02:29 | 2019-12-18T23:02:29 | {} | MEMBER |
ClickHouse | ClickHouse | 567,294,256 | 2,345 | agaurav | @alexey-milovidov it has been labelled `st-accepted`.
has it been already taken ? | 2019-12-19T01:51:44 | 2019-12-19T01:51:44 | {} | NONE |
ClickHouse | ClickHouse | 567,359,275 | 7,707 | amosbird | @stavrolia there is one test in this pr. what tests do you expect? | 2019-12-19T06:47:45 | 2019-12-19T06:47:45 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 567,529,831 | 8,296 | alexey-milovidov | We can add a setting `rows_before_limit_exact` that will force using `always_read_till_end` in `LimitBlockInputStream`. | 2019-12-19T15:12:53 | 2019-12-19T15:12:53 | {} | MEMBER |
ClickHouse | ClickHouse | 567,548,818 | 8,299 | Slach | dear @yurriy could you look to this issue? | 2019-12-19T15:56:52 | 2019-12-19T15:56:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,569,539 | 8,223 | den-crane | @krnaveen14 my opinion is not relevant. I am just a member of a CH community not a developer. I just wondered. | 2019-12-17T14:40:28 | 2019-12-17T14:40:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,063,323 | 8,267 | alesapin | it doesn't help :( | 2019-12-18T14:49:28 | 2019-12-18T14:49:28 | {} | MEMBER |
ClickHouse | ClickHouse | 567,109,234 | 8,275 | alesapin | We don't check consistency between types in table and dictionary even for ClickHouse source. So we can create a dictionary with `Date` types and table with `DateTime` types and get very confusing behavior. | 2019-12-18T16:32:22 | 2019-12-18T16:32:22 | {} | MEMBER |
ClickHouse | ClickHouse | 567,264,979 | 8,258 | alexey-milovidov | The test:
```
dbms/programs/clickhouse local --compile_expressions 1 --min_count_to_compile_expression 0 --query "SELECT number + number + number FROM system.numbers LIMIT 10"
``` | 2019-12-18T23:51:59 | 2019-12-18T23:51:59 | {} | MEMBER |
ClickHouse | ClickHouse | 567,427,240 | 8,283 | alesapin | You have to specifiy `db 'prod'`, not `database 'prod'`. | 2019-12-19T10:17:05 | 2019-12-19T10:17:05 | {} | MEMBER |
ClickHouse | ClickHouse | 566,480,270 | 6,168 | YiuRULE | I rebased from master. (and resolved some merge conflict)
We have normally errors but not due to the patch | 2019-12-17T10:26:10 | 2019-12-17T10:26:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,850,862 | 8,242 | alexey-milovidov | > It's not an issue for this PR, but how do you think about #8022?
I made that PR two weeks ago, but nobody left any comment.
I wonder how to do that. :-)
It's on queue, we will continue (either me or someone from my team). | 2019-12-18T03:27:22 | 2019-12-18T03:27:22 | {} | MEMBER |
ClickHouse | ClickHouse | 567,039,034 | 8,194 | excitoon | @KochetovNicolai there is no need to backport it to 19.16 because 19.16 has no pool for background moves. | 2019-12-18T13:49:42 | 2019-12-18T13:49:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,459,372 | 8,167 | CurtizJ | I suggest to make this functionality optional and implement it via mutations. | 2019-12-19T11:54:02 | 2019-12-19T11:54:02 | {} | MEMBER |
ClickHouse | ClickHouse | 567,586,807 | 8,148 | zhang2014 | I have also investigated this problem, and I'm not sure if it helps you(hopefully):
#7621
https://reviews.llvm.org/D56731
https://github.com/llvm/llvm-project/commit/73b51ae160af6b94d4468331ff3fb6855cff3b18
this commit exists at `master` `llvmorg-10-init` `llvmorg-9.0.1-rc3` `llvmorg-9.0.1-rc2` `llvmorg-9.0.1... | 2019-12-19T17:33:01 | 2019-12-19T17:33:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,682,110 | 8,151 | alexey-milovidov | > But unwinding without exception works just fine.
Does it successfully unwind stack through (above) this function? | 2019-12-17T18:10:36 | 2019-12-17T18:10:36 | {} | MEMBER |
ClickHouse | ClickHouse | 566,809,858 | 8,242 | ghost | @alexey-milovidov
I surrounded the test with `SYSTEM STOP MERGES`, `SYSTEM START MERGES` and set `allow_experimental_data_skipping_indices` for this test.
After you make it the default, I will remove it if needed.
I learned a lot about ClickHouse.
Thank you. | 2019-12-18T00:23:34 | 2019-12-18T00:26:44 | {} | NONE |
ClickHouse | ClickHouse | 566,912,067 | 8,027 | jiugem | Thanks, I have recompiled successfully according to the contents of commits.
At 2019-12-18 01:42:59, "alexey-milovidov" <notifications@github.com> wrote:
How to get pre-built binaries:
Open the list of commits: https://github.com/ClickHouse/ClickHouse/commits/master
Click on green check mark.
Click on the details ... | 2019-12-18T07:27:35 | 2019-12-18T07:27:35 | {} | NONE |
ClickHouse | ClickHouse | 567,006,640 | 8,093 | huang--yijian | > @huang--yijian
> Yes, it is safe. We have removed all the dependencies that may prevent installation.
> The `clickhouse` binary itself doesn't have any dependencies except old enough version of `libc` that already exists in the system.
Thank you. We will install ClickHouse RPMs manually by 'rpm -Fvh'. | 2019-12-18T12:11:25 | 2019-12-18T12:11:25 | {} | NONE |
ClickHouse | ClickHouse | 567,072,276 | 5,771 | andim2 | Some post-mortem writeup (HTH):
I also had this compile error with llvm::sort etc., when trying to build IWYU (branch clang_8.0 , fe5275aa7b6a1372), on Ubuntu 18.04.3 LTS x64, with distro packages.
In my case it turned out that I had an inconsistent llvm config (IIRC, the resulting CMakeCache.txt listed LLVM_DIR w... | 2019-12-18T15:09:36 | 2019-12-18T15:09:36 | {} | NONE |
ClickHouse | ClickHouse | 567,076,113 | 8,267 | alesapin | Don't merge. | 2019-12-18T15:18:41 | 2019-12-18T15:18:41 | {} | MEMBER |
ClickHouse | ClickHouse | 567,205,480 | 8,277 | den-crane | It's how the insert deduplication works.
The last insert did not happen because it's the same as 1st.
```
set insert_deduplicate=0
.....
SELECT *
FROM table_name
┌─CounterID─┬─UserID─┬─sign─┐
│ 1 │ 1 │ 1 │
└───────────┴────────┴──────┘
```
https://clickhouse.yandex/docs/en/operation... | 2019-12-18T20:45:16 | 2019-12-18T20:50:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,348,667 | 7,969 | filimonov | How many partitions in the topic? | 2019-12-19T06:05:43 | 2019-12-19T06:05:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,845,821 | 936 | zhangguiyu | Correct me if I am wrong, based on the above comments I summarize the current status as of 2019.12.18
1) [x] 3-node zookeeper + 3-node cluster
config = 2 replicas, 0 shards
config = <internal_replication>true</internal_replication>
2) [x] CREATE/DROP ReplicatedMergeTree ON CLUSTER mycluster
working... | 2019-12-18T03:02:51 | 2019-12-18T03:04:04 | {} | NONE |
ClickHouse | ClickHouse | 567,118,625 | 7,477 | alexey-milovidov | And maybe better to move the code around: move it inside `getThreadNumber()` and use it in `TaskStatsInfoGetter`. | 2019-12-18T16:54:36 | 2019-12-18T16:54:36 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 567,452,435 | 8,285 | vkingnew | wow,the question is answered so quickly. Thanks!
Is LTS version open to common users？ And whether provided the packages like DEB and RPM for download ? | 2019-12-19T11:30:40 | 2019-12-19T11:30:40 | {} | NONE |
ClickHouse | ClickHouse | 567,571,375 | 8,301 | alexey-milovidov | Not an issue. | 2019-12-19T16:52:30 | 2019-12-19T16:52:30 | {} | MEMBER |
ClickHouse | ClickHouse | 566,532,740 | 7,812 | alesapin | We can do this simplier without changes in part names by default https://github.com/ClickHouse/ClickHouse/pull/8250/files. | 2019-12-17T13:04:10 | 2019-12-17T13:04:10 | {} | MEMBER |
ClickHouse | ClickHouse | 566,632,542 | 8,245 | alex-zaitsev | Yes, we know. Stay tuned. | 2019-12-17T16:17:05 | 2019-12-17T16:17:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,641,970 | 8,242 | alexey-milovidov | Performance test is failed due to `allow_experimental_data_skipping_indices`.
I will make it the default and merge this PR. | 2019-12-17T16:32:58 | 2019-12-17T16:32:58 | {} | MEMBER |
ClickHouse | ClickHouse | 566,651,499 | 8,151 | alexey-milovidov | Maybe our libunwind is somewhat flawed.
We see a similar issue here: https://github.com/ClickHouse/ClickHouse/pull/8218#issuecomment-565864097
It looks like it cannot correctly interpret "CFI" in certain cases. | 2019-12-17T16:55:26 | 2019-12-17T16:55:26 | {} | MEMBER |
ClickHouse | ClickHouse | 566,667,955 | 8,151 | amosbird | But unwinding without exception works just fine. I'd suspect a clang bug.... | 2019-12-17T17:34:37 | 2019-12-17T17:34:37 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 567,292,820 | 7,707 | stavrolia | @amosbird Could please add test for this? | 2019-12-19T01:45:24 | 2019-12-19T01:45:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,487,152 | 5,131 | slimtom95 | Any update on this? | 2019-12-17T10:43:53 | 2019-12-17T10:43:53 | {} | NONE |
ClickHouse | ClickHouse | 566,680,174 | 8,151 | alexey-milovidov | > compile executeQuery with -fno-omit-frame-point works
Let's do it (we can specify this flag for a single source file in CMake) as a first solution.
... and wait for more bugs :) | 2019-12-17T18:05:40 | 2019-12-17T18:05:40 | {} | MEMBER |
ClickHouse | ClickHouse | 566,699,402 | 8,164 | KochetovNicolai | Conflicts with 19.16 | 2019-12-17T18:54:49 | 2019-12-17T18:54:49 | {} | MEMBER |
ClickHouse | ClickHouse | 566,890,512 | 7,799 | Slach | Same problem on 19.17.4.11
following query works if remove `INNER JOIN`
of if remove `i.session2_stat_crm_tracker IN (...)`
I didn't understand error message and how corresponding JOIN and WHERE in different columns (`i.pays` in JOIN and `i.session2_stat_crm_tracker` in WHERE)
```sql
SELECT toDate(i.date) ... | 2019-12-18T06:41:21 | 2019-12-18T09:43:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,982,963 | 7,898 | KochetovNicolai | Everything else is ok. | 2019-12-18T10:57:00 | 2019-12-18T10:57:00 | {} | MEMBER |
ClickHouse | ClickHouse | 567,363,650 | 7,969 | chukynax | 3
in clickhouse queue with kafka_consuners 3 also
| 2019-12-19T07:03:44 | 2019-12-19T08:18:00 | {} | NONE |
ClickHouse | ClickHouse | 567,548,677 | 8,293 | den-crane | Q1. https://github.com/ClickHouse/ClickHouse/blob/master/docs/ru/extended_roadmap.md Russian only.
Q2. `JOIN not limited by available memory` Almost ready. AFAIK expected in Q1 2020 in release.
Q3. CH license does not restrict it. | 2019-12-19T15:56:29 | 2019-12-19T15:56:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,523,538 | 8,116 | blinkov | @godfreyd why do we need this as separate section? it's just topK aggregate function with -Array combinator:
https://clickhouse.yandex/docs/en/query_language/agg_functions/reference/#topk-n-column
https://clickhouse.yandex/docs/en/query_language/agg_functions/combinators/#agg-functions-combinator-array | 2019-12-17T12:35:52 | 2019-12-17T12:35:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,742,508 | 7,386 | akuzm | Test failures are flaps or broken in master. The run time improved after rewriting the assertion, now it's about 2h:20m, the same as in master. | 2019-12-17T20:48:06 | 2019-12-17T20:48:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,789,349 | 7,676 | alexey-milovidov | @stavrolia will add support for wildcards. | 2019-12-17T23:05:43 | 2019-12-17T23:05:43 | {} | MEMBER |
ClickHouse | ClickHouse | 566,961,578 | 8,228 | ashkank66 | This is all the responses:
root@ashkanPC:/home/ashkan# grep -r -i mysql /etc/clickhouse-server/
/etc/clickhouse-server/users.xml: Restrictions of SHA256: impossibility to connect to ClickHouse using MySQL JS client (as of July 2019).
root@ashkanPC:/home/ashkan# grep -i mysql /etc/metrika.xml
grep:... | 2019-12-18T09:58:43 | 2019-12-18T09:58:43 | {} | NONE |
ClickHouse | ClickHouse | 566,975,635 | 8,264 | alesapin | Already done https://github.com/ClickHouse/ClickHouse/pull/8032. Just try fresh 19.17. | 2019-12-18T10:36:23 | 2019-12-18T10:37:41 | {} | MEMBER |
ClickHouse | ClickHouse | 567,213,259 | 7,091 | silviucpp | @filimonov any idea if `max_execution_time` it suppose to work also with queries like `CREATE TEMPORARY TABLE` ?
For sure right now its not working but I don't know if it's a bug or by design | 2019-12-18T21:07:18 | 2019-12-18T21:07:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,596,192 | 8,293 | alexey-milovidov | We welcome the use of ClickHouse in AliCloud as in other cloud providers, services, appliances and commercial products. | 2019-12-19T17:57:22 | 2019-12-19T17:57:22 | {} | MEMBER |
ClickHouse | ClickHouse | 566,369,285 | 6,549 | den-crane | I've added a new replica at my stage (and killed a source node). All downloaded parts' files in that replica have file privileges `-r--r-----`.
```
alter table XX update colX = [0] where colX = [-1] and date = '2019-12-12';
Code: 76, e.displayText() = DB::ErrnoException: Cannot open file
/var/lib/clickhouse/d... | 2019-12-17T03:56:44 | 2019-12-17T14:57:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,674,236 | 8,254 | 4ertus2 | #7556 | 2019-12-17T17:50:46 | 2019-12-17T17:50:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 566,861,273 | 8,258 | alexey-milovidov | Now it works. | 2019-12-18T04:23:21 | 2019-12-18T04:23:21 | {} | MEMBER |
ClickHouse | ClickHouse | 566,982,685 | 7,898 | KochetovNicolai | There is no `01035_avg_weighted.sql ` now. Did you forget to add it? | 2019-12-18T10:56:11 | 2019-12-18T10:56:11 | {} | MEMBER |
ClickHouse | ClickHouse | 567,054,463 | 8,267 | alexey-milovidov | Stress test: @Enmk issue with function `now64`. | 2019-12-18T14:28:44 | 2019-12-18T14:28:44 | {} | MEMBER |
ClickHouse | ClickHouse | 567,335,545 | 8,282 | maskshell | @alesapin
@KochetovNicolai | 2019-12-19T05:09:37 | 2019-12-19T05:09:37 | {} | NONE |
ClickHouse | ClickHouse | 567,560,771 | 8,148 | alexey-milovidov | This PR breaks clang builds: https://clickhouse-builds.s3.yandex.net/8148/54ddf4b0a03a12c6d2e6a5e5e78d88087ca18b1d/report.html | 2019-12-19T16:25:49 | 2019-12-19T16:25:49 | {} | MEMBER |
ClickHouse | ClickHouse | 566,919,628 | 8,061 | ghuname | I hope so. | 2019-12-18T07:54:38 | 2019-12-18T07:54:38 | {} | NONE |
ClickHouse | ClickHouse | 567,221,289 | 7,707 | alexey-milovidov | > It also makes sense for dictGet functions to accept tables
Now we have a green light for that! | 2019-12-18T21:29:35 | 2019-12-18T21:29:35 | {} | MEMBER |
ClickHouse | ClickHouse | 562,480,060 | 7,995 | nicelulu | https://github.com/ClickHouse/ClickHouse/blob/9123cce70307c87cf74d1f4ee03db607efb607ac/contrib/libcxx-cmake/CMakeLists.txt#L47
Maybe because of this? | 2019-12-06T08:26:37 | 2019-12-06T08:26:37 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.