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 | 774,192,518 | 19,252 | filimonov | > No difference
Great, thanks! 👍 | 2021-02-05T18:01:37 | 2021-02-05T18:01:37 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,499,846 | 16,049 | bharatnc | Resolved via #20107. | 2021-02-06T16:06:57 | 2021-02-06T16:06:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,536,419 | 20,152 | alexey-milovidov | > AST fuzzer (UBSan) — ../src/Functions/array/mapPopulateSeries.cpp:193:49: runtime error: signed integer overflow: 9223372036854775807 + 1 cannot be represented i
https://github.com/ClickHouse/ClickHouse/issues/20161 | 2021-02-06T20:11:25 | 2021-02-06T20:11:25 | {} | MEMBER |
ClickHouse | ClickHouse | 774,538,024 | 20,094 | alexey-milovidov | > Stress test (thread) — Sanitizer assert (in stderr.log)
https://github.com/ClickHouse/ClickHouse/issues/20032 | 2021-02-06T20:21:16 | 2021-02-06T20:21:16 | {} | MEMBER |
ClickHouse | ClickHouse | 774,550,500 | 20,160 | alexey-milovidov | > AST fuzzer (MSan) — SUMMARY: MemorySanitizer: use-of-uninitialized-value obj-x86_64-linux-gnu/../contrib/libcxx/include/string:2379:11 in std::__1::basic_string
https://github.com/ClickHouse/ClickHouse/issues/20165 | 2021-02-06T22:04:10 | 2021-02-06T22:04:10 | {} | MEMBER |
ClickHouse | ClickHouse | 774,552,839 | 20,037 | alexey-milovidov | > At least I'd like to add ignore(available()) where possible.
Yes, it will be much better. | 2021-02-06T22:27:32 | 2021-02-06T22:27:32 | {} | MEMBER |
ClickHouse | ClickHouse | 774,677,958 | 20,131 | alexey-milovidov | > and all you will see is garbage
It should never display garbage - this is a bug. It may overflow and display arbitrary time but not garbage.
We don't have this issue for DateTime type and I remember that I fixed it long time ago. | 2021-02-07T13:52:38 | 2021-02-07T13:52:38 | {} | MEMBER |
ClickHouse | ClickHouse | 774,249,700 | 20,124 | alesapin | Tests not important. | 2021-02-05T19:46:43 | 2021-02-05T19:46:43 | {} | MEMBER |
ClickHouse | ClickHouse | 774,477,683 | 19,649 | vdimir |
Testing parralel download with 800MB file:
```
select count(*) from s3('https://test-bucket-vdimir.s3.mds.yandex.net/test_vdimir_1/tsv/dat.tsv', 'LineAsString', 'x String')
```
<details>
<summary>Single thread</summary>
```
set max_read_buffer_size = 1048576;
set max_download_threads = 1;
```
```
1 r... | 2021-02-06T13:29:10 | 2021-02-06T13:29:10 | {} | MEMBER |
ClickHouse | ClickHouse | 774,546,988 | 20,159 | mzealey | This is with both 20.11 and 21.1.
Your example is working as described for me.
However you are looking from the tail of the table. If you look from the beginning for example:
```
SELECT *
FROM logs
WHERE toDateTime(time) > toDateTime('2021-01-26 17:00:00')
ORDER BY time ASC
LIMIT 10
10 rows in set. Ela... | 2021-02-06T21:31:37 | 2021-02-06T21:31:37 | {} | NONE |
ClickHouse | ClickHouse | 774,561,696 | 20,162 | alexey-milovidov | Yes, after `optimize_on_insert` is enabled there is no reason to write GROUP BY in mat.view definition. | 2021-02-06T23:50:55 | 2021-02-06T23:50:55 | {} | MEMBER |
ClickHouse | ClickHouse | 774,583,469 | 20,091 | hexiaoting | I want to try it. | 2021-02-07T03:15:16 | 2021-02-07T03:15:16 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,863,840 | 20,018 | vitlibar | > hi alexey-milovidov vitlibar , I have fixed the bug, today I will commit the pr.
Sorry, I didn't see your message. I've already created the PR. | 2021-02-05T07:57:41 | 2021-02-05T07:57:41 | {} | MEMBER |
ClickHouse | ClickHouse | 773,883,233 | 19,252 | alexey-milovidov | Sorry, wrong query... I will test another one. | 2021-02-05T08:37:23 | 2021-02-05T08:37:23 | {} | MEMBER |
ClickHouse | ClickHouse | 773,992,024 | 20,018 | JackyWoo | > > hi alexey-milovidov vitlibar , I have fixed the bug, today I will commit the pr.
>
> Sorry, I didn't see your message. I've already created the PR.
It does not mater. Thanks for your pr. | 2021-02-05T12:03:20 | 2021-02-05T12:03:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,417,188 | 20,120 | alesapin | Unrelated test failure. | 2021-02-06T07:23:34 | 2021-02-06T07:23:34 | {} | MEMBER |
ClickHouse | ClickHouse | 774,549,630 | 20,092 | alexey-milovidov | @ucasFL I corrected the description.
Do we have optimization for `avg(x + const)`? | 2021-02-06T21:56:07 | 2021-02-06T21:56:07 | {} | MEMBER |
ClickHouse | ClickHouse | 774,583,466 | 20,016 | JackyWoo | hi tavplubix, I got what you mean. Because of both DDL worker thread and clean thread will operate the zookeeper queue, I think handle queue competition will make sense. | 2021-02-07T03:15:14 | 2021-02-07T03:15:14 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,634,413 | 20,102 | nshfany | Hello, the same problem has happend on CH version: ( 20.8.6.6) | 2021-02-07T08:30:21 | 2021-02-07T08:30:21 | {} | NONE |
ClickHouse | ClickHouse | 773,868,751 | 20,087 | ervikrant06 | okay thanks for sharing the information. I will change to any instead of avg.
I am sending the metrics of different retention, my table is replicated.
~~~
clickhouse1-0.clickhouse-hs.ns-vaggarwal.svc.cluster.local :) show create table graphite
SHOW CREATE TABLE graphite
Query id: d0734bca-40fb-4c9b-accf-... | 2021-02-05T08:07:57 | 2021-02-05T08:07:57 | {} | NONE |
ClickHouse | ClickHouse | 774,233,165 | 20,011 | azat | >@alexey-milovidov @vdimir DateTime64 will be fixed separately, since this is not only about UBsan reports, but there is just no specializations for DataTypeDateTime64 from numeric, and it simply uses raw cast
#20131 addresses overflows in dates | 2021-02-05T19:14:25 | 2021-02-05T19:14:25 | {} | MEMBER |
ClickHouse | ClickHouse | 774,439,946 | 20,139 | sonirico | @nvartolomei You are right indeed. Surely I messed up while I was testing. I'm editing the explanation. Anyway, I still consider that configuring access management prior to spawning up the environment is useful to leave prepared the installation at once (I think of helm charts, ansible playbooks...), in spite of counti... | 2021-02-06T10:04:04 | 2021-02-06T10:04:04 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,549,726 | 20,092 | alexey-milovidov | ```
milovidov-desktop :) EXPLAIN SYNTAX SELECT sum(number * 2) FROM system.numbers
EXPLAIN SYNTAX
SELECT sum(number * 2)
FROM system.numbers
Query id: f2a66469-d3a4-4548-813b-042798d334dd
┌─explain────────────────┐
│ SELECT sum(number) * 2 │
│ FROM system.numbers │
└────────────────────────┘
2 rows... | 2021-02-06T21:57:17 | 2021-02-06T21:57:17 | {} | MEMBER |
ClickHouse | ClickHouse | 773,852,296 | 19,562 | qoega | same place
``` sql
SELECT intDiv(number, 1.)
FROM numbers(2 * 9223372036854775807, 1)
```
```
../base/common/../common/extended_types.h:110:28: runtime error: 1.84467e+19 is outside the range of representable values of type 'long'
SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior ../base/common/../common... | 2021-02-05T07:31:15 | 2021-02-05T07:31:15 | {} | MEMBER |
ClickHouse | ClickHouse | 773,890,358 | 18,247 | ghuname | > Can you reproduce the issue?
I tried to reproduce it by creation of the table with ReplicatedCollapsingMergeTree engine, insert some rows and drop it, all the time using on cluster commands.
After repeating the process for dozen times, I had no success to produce the error.
@SaltTan What was your idea in orde... | 2021-02-05T08:51:22 | 2021-02-05T08:51:22 | {} | NONE |
ClickHouse | ClickHouse | 773,929,463 | 20,051 | tavplubix | Unlike `StorageSystemPartsBase`, `StorageSystemTables` uses only virtual methods of `IStorage` and does not cast storage to `MergeTreeData`, so it should work for MaterializeMySQL without replacing storage pointer. The issue is that `StorageMaterializeMySQL` [does not set](https://github.com/ClickHouse/ClickHouse/blob/... | 2021-02-05T10:01:10 | 2021-02-05T10:01:10 | {} | MEMBER |
ClickHouse | ClickHouse | 774,039,083 | 20,100 | KochetovNicolai | #20073 | 2021-02-05T13:40:19 | 2021-02-05T13:40:19 | {} | MEMBER |
ClickHouse | ClickHouse | 774,574,755 | 19,407 | templarzq | @CurtizJ test finished. | 2021-02-07T01:50:07 | 2021-02-07T01:50:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,662,489 | 21 | alexey-milovidov | Please also check these performance results: https://github.com/ClickHouse/ClickHouse/pull/20001#issuecomment-774661963 | 2021-02-07T11:59:23 | 2021-02-07T11:59:23 | {} | MEMBER |
ClickHouse | ClickHouse | 773,963,126 | 3,123 | crakjie | Documentation update is missing | 2021-02-05T11:02:31 | 2021-02-05T11:02:31 | {} | NONE |
ClickHouse | ClickHouse | 774,028,163 | 20,043 | den-crane | Interesting,
CH before the reading data (before execution starts) knows that it need to process 6211 marks.
6211 * 8192 = 50,880,512
I'll try to reproduce it. | 2021-02-05T13:18:59 | 2021-02-05T13:18:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,034,339 | 20,051 | stigsb | I ended up just copying the in-memory metadata from the nested table. Can't think of any cases where that could cause other problems, what do you think @tavplubix ? | 2021-02-05T13:31:03 | 2021-02-05T13:31:03 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,534,184 | 20,150 | alexey-milovidov | > AST fuzzer (UBSan) — ../src/Functions/array/arrayCumSum.cpp:87:63: runtime error: signed integer overflow: -9223372036854775808 + -9223372036854775808 cannot be
https://github.com/ClickHouse/ClickHouse/pull/20160 | 2021-02-06T20:02:02 | 2021-02-06T20:02:02 | {} | MEMBER |
ClickHouse | ClickHouse | 774,536,564 | 20,161 | alexey-milovidov | More illustrative example:
```
SELECT mapPopulateSeries([0xFFFFFFFFFFFFFFFF], [0], 0xFFFFFFFFFFFFFFFF)
``` | 2021-02-06T20:12:31 | 2021-02-06T20:12:31 | {} | MEMBER |
ClickHouse | ClickHouse | 774,586,905 | 20,129 | keen-wolf | I'd like to take this one~ | 2021-02-07T03:28:32 | 2021-02-07T03:28:32 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,592,891 | 20,129 | alexey-milovidov | Thank you, let's do it! | 2021-02-07T04:27:29 | 2021-02-07T04:27:29 | {
"rocket": 1
} | MEMBER |
ClickHouse | ClickHouse | 774,593,654 | 20,091 | alexey-milovidov | Great, it will be wonderful! | 2021-02-07T04:28:10 | 2021-02-07T04:28:10 | {} | MEMBER |
ClickHouse | ClickHouse | 774,666,254 | 20,179 | vmihailenco | Thanks for the answer. I guess I will continue to use it with caution since otherwise it is working well enough and S3 pricing is considerably lower (with zero maintenance).
If there are any workarounds (like ability to split parts) - please let me know. | 2021-02-07T12:26:44 | 2021-02-07T12:27:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,884,187 | 19,252 | alexey-milovidov | No difference:
`clickhouse-benchmark --query "SELECT count() FROM test.hits WHERE NOT ignore(toIPv6(IPv6NumToString(ClientIP6)))"`
21.1:
```
localhost:9000, queries 253, QPS: 10.733, RPS: 95244262.006, MiB/s: 1453.325, result RPS: 10.733, result MiB/s: 0.000.
0.000% 0.080 sec.
10.000% 0.083... | 2021-02-05T08:39:18 | 2021-02-05T08:39:18 | {} | MEMBER |
ClickHouse | ClickHouse | 773,893,339 | 20,104 | amosbird | > If there is any resource or document which go deeper into the subject of indexes, I would be more than interested !
I'd recommend browsing through https://github.com/ClickHouse/clickhouse-presentations . It takes time but the journey will be fruitful. | 2021-02-05T08:57:23 | 2021-02-05T08:57:23 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 774,038,308 | 20,076 | pan3793 | > I wonder , why do you think that murmurHash3_32 seed=1361930890 ?
I see 0 ?
Oops! You are right, my fault, the search result actually in arrow submodule ... | 2021-02-05T13:38:51 | 2021-02-05T13:38:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,101,152 | 20,036 | abyss7 | Consumed by #20037 | 2021-02-05T15:26:39 | 2021-02-05T15:26:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,258,479 | 20,114 | nvartolomei | @qoega i don't think it is possible. A better idea might be to introduce a lock for synchronizing drop with shutdown/restarting thread. | 2021-02-05T20:03:39 | 2021-02-05T20:03:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,417,157 | 20,120 | alesapin | `Cannot open file /etc/clickhouse-server/config.d/no_file.txt, errno: 2, strerror: No such file or directory (version 21.3.1.5923). Stack trace:`
`¯\_(ツ)_/¯` | 2021-02-06T07:23:16 | 2021-02-06T07:23:16 | {} | MEMBER |
ClickHouse | ClickHouse | 774,561,831 | 20,162 | alexey-milovidov | Maybe it is still good for illustrative purposes.
Also maybe there are cases when GROUP BY in mat.view differs to the ORDER BY key of AggregatingMergeTree but I cannot imagine if they are useful. | 2021-02-06T23:52:13 | 2021-02-06T23:52:13 | {} | MEMBER |
ClickHouse | ClickHouse | 773,861,422 | 20,104 | amosbird | > Is this normal ? Is there anything I can do to have more homogeneous result ?
Yes
> Does the order of the columns in the ORDER BY clause of my table creation code has an impact on it ?
Yes
> Could the reason be that name in the index is non-monotonic ?
No
Most likely you `word_id = 545` is clusterd ... | 2021-02-05T07:52:02 | 2021-02-05T07:52:02 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 773,876,747 | 12,312 | bharatnc | Looks like this is fixed by the above referenced PR i.e https://github.com/ClickHouse/ClickHouse/pull/14515. Uses `TO` assignment now. | 2021-02-05T08:24:31 | 2021-02-05T08:24:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,937,211 | 5,329 | stfaun | @alexey-milovidov Is the optimization metioned here complete? | 2021-02-05T10:14:05 | 2021-02-05T10:14:40 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 774,035,337 | 20,076 | den-crane | I wonder , why do you think that `murmurHash3_32 seed=1361930890` ?
I see 0 ?
void MurmurHash3_x86_32 ( const void * key, int len, uint32_t seed, void * out
MurmurHash3_x86_32(data, size, 0, bytes);
https://github.com/ClickHouse/ClickHouse/blob/bd81f43ecb33dfd160f54b97cebf533156d9565e/src/Functions/FunctionsHa... | 2021-02-05T13:33:02 | 2021-02-05T13:33:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,261,783 | 20,130 | nvartolomei | #12300 | 2021-02-05T20:10:33 | 2021-02-05T20:10:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,538,138 | 20,159 | den-crane | What CH version do you use? **Pruning was improved drastically in 20.9+**
Before 20.9 you need to use a condition without **toDateTime** `where (time) > toDateTime64('2021-04-26 17:00:00',3)`
I could not reproduce it:
```sql
21.2.1.5834
CREATE TABLE logs (
time DateTime64(3) CODEC(DoubleDelta, Z... | 2021-02-06T20:22:23 | 2021-02-06T20:23:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,540,810 | 16,049 | den-crane | date_sub -- still not
select date_sub(YEAR, 3, toDate('2018-01-01'));
| 2021-02-06T20:44:16 | 2021-02-06T20:44:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 766,589,362 | 19,409 | sunny19930321 | > xplains nothing. It does not help.
same question，The HTTP connection is fine, but the TCP connection is down | 2021-01-25T06:51:08 | 2021-02-07T06:42:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,881,382 | 20,104 | nflorentin | @amosbird thanks a lot for your quick answer and for your work.
If there is any resource or document which go deeper into the subject of indexes, I would be more than interested !
Have a nice day. | 2021-02-05T08:33:56 | 2021-02-05T08:33:56 | {} | NONE |
ClickHouse | ClickHouse | 773,882,419 | 19,252 | alexey-milovidov | We usually don't tolerate 10% regressions in perf test. But I think this function is not of high importance.
Nevertheless, I will check the code... | 2021-02-05T08:35:56 | 2021-02-05T08:35:56 | {} | MEMBER |
ClickHouse | ClickHouse | 774,278,686 | 20,138 | den-crane | 1. MV is an insert trigger. It gets data from insert. IT NEVER READS THE SOURCE TABLE.
ENGINE OF THE SOURCE TABLE DOES NOT MATTER.
2. AggregatingMergeTree MUST BE USED with AggregateFunctions OR SimpleAggregateFunction
Now it simply corrupts `max(timestamp)` , sum(amount) DURING merges.
Check this:
https://de... | 2021-02-05T20:44:30 | 2021-02-05T20:44:30 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,353,533 | 20,139 | nvartolomei | > Prior to this PR, changing this setting by altering the /etc/clickhouse-server/users.d/default-user.xml would force to trigger a server restart to apply changes.
```sql
CREATE USER demo
Received exception from server (version 20.11.6):
Code: 497. DB::Exception: Received from localhost:9000. DB::Exception: de... | 2021-02-06T00:06:34 | 2021-02-06T00:08:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,537,854 | 20,161 | alexey-milovidov | 1209c02869 | 2021-02-06T20:19:39 | 2021-02-06T20:19:39 | {} | MEMBER |
ClickHouse | ClickHouse | 774,542,983 | 16,049 | bharatnc | > date_sub -- still not
>
> select date_sub(YEAR, 3, toDate('2018-01-01'));
Thanks - Missed this for a similar func `date_diff`. Added via #20163. | 2021-02-06T21:02:52 | 2021-02-06T21:12:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,575,713 | 20,168 | alexey-milovidov | ```
milovidov-desktop :) SELECT sum(x) FROM test_memory
SELECT sum(x)
FROM test_memory
Query id: 14833c4a-436f-40d6-8cce-f1a09db81158
┌─────sum(x)─┐
│ 1000000000 │
└────────────┘
1 rows in set. Elapsed: 0.140 sec. Processed 1.00 billion rows, 8.00 GB (7.13 billion rows/s., 57.04 GB/s.)
milovidov-des... | 2021-02-07T01:59:28 | 2021-02-07T01:59:28 | {} | MEMBER |
ClickHouse | ClickHouse | 774,677,027 | 20,174 | alexey-milovidov | BTW, why extractAllGroups or roundBankers are case insensitive? It is simply a mistake.
Only functions compatible with SQL standard or MySQL/Postgres/SQLite, etc should be case insensitive.
ClickHouse native functions should not be case insensitive. | 2021-02-07T13:46:27 | 2021-02-07T13:46:27 | {} | MEMBER |
ClickHouse | ClickHouse | 774,038,167 | 20,051 | tavplubix | Yes, it should work, let's wait for test results | 2021-02-05T13:38:35 | 2021-02-05T13:38:35 | {} | MEMBER |
ClickHouse | ClickHouse | 774,470,133 | 20,092 | ucasfl | In fact, for `sum`, `sum(x * const) = sum(x) * const` and `sum(x / const) = sum(x) / const`, but `sum(x +/- const)` isn't equal to `sum(x) +/- const`. @alexey-milovidov | 2021-02-06T12:28:12 | 2021-02-06T12:29:46 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 774,494,184 | 20,014 | ahutsunshine | @den-crane Thank you for your great answer. If I have created multiple ReplicatedMergeTree tables in the clickhouse cluster, do you have a way to make one of the tables as readonly mode rather than all tables? | 2021-02-06T15:29:09 | 2021-02-06T15:29:09 | {} | NONE |
ClickHouse | ClickHouse | 774,549,750 | 20,092 | alexey-milovidov | Only first query is rewritten by default. | 2021-02-06T21:57:34 | 2021-02-06T21:57:34 | {} | MEMBER |
ClickHouse | ClickHouse | 774,554,142 | 20,150 | alexey-milovidov | > AST fuzzer (ASan) — Received signal 11
https://github.com/ClickHouse/ClickHouse/issues/20046 | 2021-02-06T22:38:50 | 2021-02-06T22:38:50 | {} | MEMBER |
ClickHouse | ClickHouse | 774,561,231 | 20,158 | alexey-milovidov | `SELECT fo,22222?LUTAY(SELECT(NOT CAUTAY(SELECT(NOT CAST(NOTT(NOT CAST(NOT NOT LEfT(NOT coARRAYlumnsFLuTAY(SELECT(NO0?LUTAY(SELECT(NOT CAUTAY(SELECT(NOT CAST(NOTT(NOT CAST(NOT NOT LEfT(NOT coARRAYlumnsFLuTAY(SELECT(NOTAYTAY(SELECT(NOTAYEFAULT(fo,22222?LUTAY(%SELECT(NOT CAST(NOT NOTAYTAY(SELECT(NOTAYEFAULT(fo,22222?LUTA... | 2021-02-06T23:45:41 | 2021-02-06T23:45:53 | {} | MEMBER |
ClickHouse | ClickHouse | 774,595,129 | 20,173 | alexey-milovidov | A test showing the issue:
```
clickhouse-benchmark --query_profiler_real_time_period_ns 10000 --min_bytes_to_use_mmap_io 1 --checksum_on_read 0 --query "SELECT sum(ignore(WatchID, UserID, ClientIP, EventTime, EventDate, CounterID)) FROM test.hits"
```
```
SELECT arrayStringConcat(arrayMap(x -> demangle(address... | 2021-02-07T04:31:01 | 2021-02-07T04:31:21 | {} | MEMBER |
ClickHouse | ClickHouse | 773,888,139 | 19,562 | alexey-milovidov | Reopen when a new report will come. | 2021-02-05T08:47:07 | 2021-02-05T08:47:07 | {} | MEMBER |
ClickHouse | ClickHouse | 773,956,767 | 20,043 | ghost | 1. Set logs to debug:
```
SET send_logs_level = 'debug'
Query id: a418c88f-4b73-4e1d-baa3-e539d04db02b
[labs-den-personal] 2021.02.05 11:48:55.117742 [ 12523 ] {a418c88f-4b73-4e1d-baa3-e539d04db02b} <Debug> executeQuery: (from 127.0.0.1:55946, using production parser) set send_logs_level = 'debug';
[labs-den-p... | 2021-02-05T10:50:26 | 2021-02-05T10:50:26 | {} | NONE |
ClickHouse | ClickHouse | 774,416,139 | 19,792 | olgarev | @Avogar Hi, will you review this, pls? | 2021-02-06T07:14:24 | 2021-02-06T07:14:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,417,365 | 20,120 | alesapin | Will try to investigate later. | 2021-02-06T07:24:57 | 2021-02-06T07:24:57 | {} | MEMBER |
ClickHouse | ClickHouse | 774,676,740 | 20,174 | alexey-milovidov | I think it will be more visually pleasant if they will be not lowercased, but normalized to canonical name.
Can we extract canonical name from function object (getName()) or from factory? | 2021-02-07T13:44:20 | 2021-02-07T13:44:20 | {} | MEMBER |
ClickHouse | ClickHouse | 773,901,311 | 20,063 | hexiaoting | @alexey-milovidov I have added more tests. | 2021-02-05T09:11:57 | 2021-02-05T09:11:57 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 773,902,279 | 20,104 | nflorentin | @amosbird thank you for taking time to answer my questions
| 2021-02-05T09:13:49 | 2021-02-05T09:13:49 | {} | NONE |
ClickHouse | ClickHouse | 774,229,870 | 20,114 | qoega | Can we add a test for this case? | 2021-02-05T19:07:19 | 2021-02-05T19:07:19 | {} | MEMBER |
ClickHouse | ClickHouse | 774,386,718 | 10,162 | cw9 | For my own record, this has been fixed | 2021-02-06T03:01:28 | 2021-02-06T03:01:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,475,788 | 20,092 | ucasfl | There is nothing need to do, because this has been done in #11899. @alexey-milovidov | 2021-02-06T13:13:40 | 2021-02-06T13:15:46 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 774,538,053 | 20,094 | alexey-milovidov | https://github.com/ClickHouse/ClickHouse/issues/20158 | 2021-02-06T20:21:37 | 2021-02-06T20:21:37 | {} | MEMBER |
ClickHouse | ClickHouse | 774,644,433 | 20,092 | ucasfl | > Only first query is rewritten by default.
Yes, I checked that we have optimization for `sum`, `min` and `max` except `avg`. | 2021-02-07T09:48:58 | 2021-02-07T09:48:58 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 774,326,897 | 20,114 | nvartolomei | I think I spotted another bug. Table can be greated with auxiliary zookeper but on drop we try to drop it from the default zookeeper... cc @fastio #17070
Trying to test this case in #20141 | 2021-02-05T22:40:15 | 2021-02-05T22:47:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,544,025 | 20,163 | bharatnc | cc: @den-crane @vdimir | 2021-02-06T21:09:15 | 2021-02-06T21:09:15 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,664,024 | 20,179 | alexey-milovidov | S3 volume is experimental feature in development stage, it has obvious drawbacks that should be eliminated before production readiness. They use O(N) memory for files - it is just unbearable, and before it is fixed, you should not consider the feature for any serious usage. | 2021-02-07T12:10:20 | 2021-02-07T12:10:20 | {} | MEMBER |
ClickHouse | ClickHouse | 774,046,416 | 3,123 | den-crane | @crakjie ??? https://clickhouse.tech/docs/en/sql-reference/dictionaries/external-dictionaries/external-dicts-dict-layout/#range-hashed
..This storage method works the same way as hashed and allows using date/time **(arbitrary numeric type)** ranges
..(Date / DateTime / UInt64 / Int32 / others).
```
<range_min>... | 2021-02-05T13:54:22 | 2021-02-05T13:54:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,463,329 | 20,123 | tavplubix | Stress test (thread) - segfault in external aggregation, looks similar to #18875
```
2021.02.05 17:48:47.651660 [ 18595 ] {3acd53e4-614a-4cbb-9305-e2eec6660de3} <Debug> executeQuery: (from [::1]:34904, using production parser) (comment: /usr/share/clickhouse-test/queries/0_stateless/00107_totals_after_having.sql) SEL... | 2021-02-06T12:00:53 | 2021-02-06T12:00:53 | {} | MEMBER |
ClickHouse | ClickHouse | 774,536,233 | 13,166 | alexey-milovidov | #20161 | 2021-02-06T20:10:19 | 2021-02-06T20:10:19 | {} | MEMBER |
ClickHouse | ClickHouse | 774,553,609 | 20,078 | alexey-milovidov | Wow, the performance results are fantastic! | 2021-02-06T22:34:12 | 2021-02-06T22:34:12 | {} | MEMBER |
ClickHouse | ClickHouse | 773,935,503 | 20,066 | tavplubix | AST fuzzer (TSan) - looks similar to #19233 | 2021-02-05T10:11:47 | 2021-02-05T10:11:47 | {} | MEMBER |
ClickHouse | ClickHouse | 774,576,445 | 20,169 | alexey-milovidov | Our existing perf test should show improvements. | 2021-02-07T02:04:58 | 2021-02-07T02:04:58 | {} | MEMBER |
ClickHouse | ClickHouse | 774,661,963 | 20,001 | alexey-milovidov | Here are performance results:
https://clickhouse-test-reports.s3.yandex.net/20001/967af339e1ca75143e32f3f4a5b255dd00f40427/performance_comparison/report.html
42 queries are faster, there are no slower queries.
For faster queries improvements about 1.5 times are typical.
The average difference across all queries... | 2021-02-07T11:55:16 | 2021-02-07T11:55:16 | {} | MEMBER |
ClickHouse | ClickHouse | 774,676,850 | 20,174 | alexey-milovidov | Example: `extractallgroups` does not look beautiful. | 2021-02-07T13:45:10 | 2021-02-07T13:45:10 | {} | MEMBER |
ClickHouse | ClickHouse | 773,886,085 | 20,018 | alexey-milovidov | We can take your two PRs and cross-compare them :) | 2021-02-05T08:43:01 | 2021-02-05T08:43:01 | {} | MEMBER |
ClickHouse | ClickHouse | 773,887,826 | 19,562 | alexey-milovidov | @qoega Your report is invalid because there is no line 110 in https://github.com/ClickHouse/ClickHouse/blob/master/base/common/extended_types.h | 2021-02-05T08:46:29 | 2021-02-05T08:46:29 | {} | MEMBER |
ClickHouse | ClickHouse | 773,943,511 | 20,051 | stigsb | Thanks for the feedback @tavplubix, I'll do that instead. | 2021-02-05T10:25:48 | 2021-02-05T10:25:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 774,123,548 | 11,461 | ortho | I've tried to build ClickHouse using native apple clang compiler on two different source versions one is `v21.1.3.32-stable` and the other main branch (commit: `0db2aa3c2e8aa39eb814b8c8a22e904cb8528025`). The result is build success with some warnings. My environment is
**Operating system**
MacOS BigSur 11.2
```... | 2021-02-05T16:02:11 | 2021-02-05T16:02:11 | {} | NONE |
ClickHouse | ClickHouse | 774,554,118 | 20,046 | alexey-milovidov | https://clickhouse-test-reports.s3.yandex.net/20150/4092916db68de1264bb2afb3f1a27f681c43e28c/fuzzer_asan/server.log | 2021-02-06T22:38:33 | 2021-02-06T22:38:33 | {} | MEMBER |
ClickHouse | ClickHouse | 774,667,294 | 20,162 | mzealey | Seems to work perfectly in my testing - no need for Null engines or GROUP BY's any more. Great work! | 2021-02-07T12:35:00 | 2021-02-07T12:35:00 | {} | NONE |
ClickHouse | ClickHouse | 770,357,928 | 19,233 | alexey-milovidov | https://clickhouse-test-reports.s3.yandex.net/19862/e23dc60e31f5856b4b8f910fdbaba6b885b715d1/fuzzer_asan/server.log | 2021-01-31T10:10:03 | 2021-01-31T10:10:03 | {} | MEMBER |
ClickHouse | ClickHouse | 770,364,677 | 19,861 | alexey-milovidov | > AST fuzzer (UBSan) — ../src/Functions/FunctionDateOrDateTimeAddInterval.h:141:18: runtime error: signed integer overflow: 47764 + 9223372036854775807 cannot be r
#19859 | 2021-01-31T11:08:39 | 2021-01-31T11:08:39 | {} | MEMBER |
ClickHouse | ClickHouse | 770,374,425 | 17,847 | fibersel | Right now I can't understand why, but LZSSE demonstrates very poor performance.
Authors also claim, that codec is targeted at x64 arhitecture:
https://github.com/ConorStokes/LZSSE/blob/master/lzsse2/lzsse2.h#L67 | 2021-01-31T12:25:55 | 2021-01-31T12:25:55 | {} | CONTRIBUTOR |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.