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 | 567,749,497 | 6,807 | alexey-milovidov | We don't depend on system LLVM anymore: https://github.com/ClickHouse/ClickHouse/pull/8258 | 2019-12-20T01:32:48 | 2019-12-20T01:32:48 | {} | MEMBER |
ClickHouse | ClickHouse | 567,923,261 | 8,315 | alexey-milovidov | > I am very suspicious of such a conclusion, as a result, has any one made any benchmark on similar solution as Ceph? Which means that to use a distributed storage(either object storage or block) to directly serve as the data layer of ClickHouse.
We also provide ClickHouse over Network Block Device (as an option) in... | 2019-12-20T13:24:59 | 2019-12-20T13:26:15 | {} | MEMBER |
ClickHouse | ClickHouse | 566,706,770 | 8,210 | KochetovNicolai | Conflicts with 19.16 and 19.17 and 19.18 | 2019-12-17T19:13:43 | 2019-12-20T13:52:53 | {} | MEMBER |
ClickHouse | ClickHouse | 567,961,366 | 5,773 | alexey-milovidov | Unfinished, should be continued. | 2019-12-20T15:22:06 | 2019-12-20T15:22:06 | {} | MEMBER |
ClickHouse | ClickHouse | 567,965,444 | 8,320 | alexey-milovidov | Did you configured the `name` attribute as `injective`? | 2019-12-20T15:33:41 | 2019-12-20T15:33:41 | {} | MEMBER |
ClickHouse | ClickHouse | 567,990,939 | 8,322 | den-crane | https://clickhouse.yandex/docs/en/operations/table_engines/replacingmergetree/
_Data deduplication occurs only during a merge. Merging occurs in the background at an unknown time, so you can't plan for it. **Some of the data may remain unprocessed.** Although you can run an unscheduled merge using the OPTIMIZE query... | 2019-12-20T16:47:34 | 2019-12-20T16:53:51 | {
"+1": 1,
"heart": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,161,651 | 8,320 | amosbird | Neither works on current master. And current master requires database name specified in dictGet* functions like `'default.dictId'`. Could you test against current master to see if it's still a confusion? | 2019-12-21T07:58:49 | 2019-12-21T07:58:49 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 568,364,000 | 8,159 | gubinjie | @weiqxu
Hello, I have a few questions about your answer.
1.Can you say hdfs-client.xml can be understood as hdfs-site.xml?
2. Which specific directory of clickhouse do you mean by working directory? Is it / etc / clickhouse-server?
3. Do you still need to configure the path in the config.xml file of clickhouse to ... | 2019-12-23T05:46:55 | 2019-12-23T05:47:35 | {} | NONE |
ClickHouse | ClickHouse | 568,407,533 | 7,695 | CurtizJ | Fixed by code refactoring in #8130. | 2019-12-23T08:54:31 | 2019-12-23T08:54:31 | {
"+1": 1,
"hooray": 1
} | MEMBER |
ClickHouse | ClickHouse | 567,755,373 | 8,021 | k1my0 | > > If I use a real table (has columns a lot), only method 2 will produce the right result.
> Could you make a sample to show it? Your sample returns the same result for each method.
It's works.
Before i got Error like this
> `Code: 179, e.displayText() = DB::Exception: Different expressions with the same a... | 2019-12-20T02:02:13 | 2019-12-20T02:20:20 | {} | NONE |
ClickHouse | ClickHouse | 567,909,262 | 8,311 | alexey-milovidov | Perf tests have failed for reason. Need to investigate. | 2019-12-20T12:30:53 | 2019-12-20T12:30:53 | {} | MEMBER |
ClickHouse | ClickHouse | 567,940,668 | 8,315 | filimonov | > My collegues from Altinity have tested ClickHouse with FUSE over S3 and mabe they can share the results. This setup works but it's not near to be practical.
ClickHouse works quite well in readonly mode with fuse + goofys above s3. That means data should be moved there manually, after that ClickHouse can read it. ... | 2019-12-20T14:20:51 | 2019-12-20T14:20:51 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,001,055 | 8,322 | masosky | That explains a lot.
It happened because probably the inserts were made 1 by 1.
Thank you very much | 2019-12-20T17:00:17 | 2019-12-20T17:00:17 | {} | NONE |
ClickHouse | ClickHouse | 568,160,915 | 8,328 | yd1261137568 | > 不支持19.5.3.1。它可能已经修复。
> 尝试在19.15+上重现
I see thank you very much | 2019-12-21T07:45:05 | 2019-12-21T07:45:05 | {} | NONE |
ClickHouse | ClickHouse | 568,188,262 | 8,082 | alexey-milovidov | Thank you! | 2019-12-21T15:14:35 | 2019-12-21T15:14:35 | {} | MEMBER |
ClickHouse | ClickHouse | 568,241,739 | 8,302 | yurriy | The server generates keys if it cannot read them, thus certificates are not required.
This warning is supposed to help in case if user has provided a client with a server key, and server has generated other keys. In this case the server won't be able to decrypt a password.
I'll remove this warning on startup, since w... | 2019-12-22T08:18:38 | 2019-12-22T08:18:38 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,246,262 | 8,332 | alexey-milovidov | > Unit tests — fail: 2
Glitch while interacting with ZooKeeper in CI environment.
> Functional stateless tests (release, processors) — fail:1
Wasn't introduced in this PR.
> Description check — Category section not found
Already Ok. | 2019-12-22T09:37:26 | 2019-12-22T09:37:26 | {} | MEMBER |
ClickHouse | ClickHouse | 568,389,365 | 7,799 | Slach | doesn't reproduce on `19.19.1.1956` testing release installed from official Yandex RPM repo | 2019-12-23T07:43:30 | 2019-12-23T07:43:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,415,345 | 8,298 | filimonov | #7955 ? | 2019-12-23T09:22:47 | 2019-12-23T09:22:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,264,704 | 8,258 | alexey-milovidov | --The list of libraries was stripped too much in 1d0854c, need to refine this list.-- | 2019-12-18T23:50:51 | 2019-12-19T23:42:25 | {} | MEMBER |
ClickHouse | ClickHouse | 567,756,968 | 8,309 | alexey-milovidov | Don't merge because it will break "unbundled" build. | 2019-12-20T02:10:56 | 2019-12-20T02:10:56 | {} | MEMBER |
ClickHouse | ClickHouse | 567,949,020 | 8,293 | filimonov | Duplicate #8215
The roadmap should be translated soon. Stay tuned. | 2019-12-20T14:45:48 | 2019-12-20T14:45:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,960,355 | 5,773 | alexey-milovidov | It get enabled only for `c` sources. | 2019-12-20T15:19:13 | 2019-12-20T15:19:13 | {} | MEMBER |
ClickHouse | ClickHouse | 567,963,710 | 8,301 | alexey-milovidov | You can either check and fix this single case or rewrite all path handling from std::string to std::filesystem::path. | 2019-12-20T15:28:51 | 2019-12-20T15:28:51 | {} | MEMBER |
ClickHouse | ClickHouse | 568,011,033 | 8,322 | masosky | Once I understood this scenario, I would like to know what is the best solution.
1. Guarantee in each query (add more rules) in order to select the not duplicated results. That means each query will have an extra cost.
2. Execute a cron process that "optimize final" each table.
If you have other solutions please... | 2019-12-20T17:31:41 | 2019-12-20T17:33:16 | {} | NONE |
ClickHouse | ClickHouse | 567,931,033 | 8,108 | stavrolia | Thank you for your report. Fixed. | 2019-12-20T13:50:36 | 2019-12-20T13:50:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,074,630 | 8,313 | svladykin | @den-crane
Yes, data itself contains timestamp but it is not monotonically increasing, a batch may contain any timestamp within the last 24 hours. So, it does not help much.
How `etl_time` is different from `insert_time`? In case of clickhouse replica failure or data writer failure we may end up with batch dupli... | 2019-12-20T20:19:36 | 2019-12-20T20:19:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,744,341 | 5,765 | alexey-milovidov | What will happen if you:
- remove CMakeCache.txt;
- reconfigure with added cmake option `-D USE_INTERNAL_MYSQL_LIBRARY=1`?
If it will help, we will enable this option by default.
PS. Sorry, we don't have any Mac machine in our team. | 2019-12-20T01:06:51 | 2019-12-20T01:06:51 | {} | MEMBER |
ClickHouse | ClickHouse | 567,744,402 | 5,765 | alexey-milovidov | @qoega | 2019-12-20T01:07:13 | 2019-12-20T01:07:13 | {} | MEMBER |
ClickHouse | ClickHouse | 568,001,584 | 8,313 | den-crane | Usually data itself has some natural datetime attributes. I have 5! different timestamps in my data so I can check data consistency among 4 different DWH systems (2 clusters of CH + 2 clusters of Vertica).
You can fill `insert_time DateTime` on ETL stage before inserting into CH -- `etl_time` | 2019-12-20T17:01:58 | 2019-12-20T17:08:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,278,407 | 8,343 | Felixoid | Here are proper tests still missing, I'll implement them later | 2019-12-22T16:29:46 | 2019-12-22T16:29:46 | {} | MEMBER |
ClickHouse | ClickHouse | 568,407,115 | 7,617 | CurtizJ | Implemented in master #8130. | 2019-12-23T08:52:47 | 2019-12-23T08:52:47 | {} | MEMBER |
ClickHouse | ClickHouse | 567,614,837 | 8,278 | th0th | > > 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
> Su... | 2019-12-19T18:49:25 | 2019-12-19T18:49:25 | {} | NONE |
ClickHouse | ClickHouse | 567,745,638 | 5,773 | alexey-milovidov | Now it can be continued: https://github.com/ClickHouse/ClickHouse/pull/8258 | 2019-12-20T01:13:36 | 2019-12-20T01:13:36 | {} | MEMBER |
ClickHouse | ClickHouse | 567,786,762 | 8,285 | vkingnew | @alexey-milovidov That's a Good news for enterprise users !
v19.14.8.9-lts , Whether this version is the first LTS Version from clickhouse born so far ? | 2019-12-20T04:52:12 | 2019-12-20T04:52:12 | {} | NONE |
ClickHouse | ClickHouse | 568,166,757 | 8,082 | moon03432 | @alexey-milovidov hi, test has been added in 00829_bitmap_function.sql and reference | 2019-12-21T09:28:28 | 2019-12-21T09:28:28 | {} | NONE |
ClickHouse | ClickHouse | 568,275,351 | 8,340 | cjdelisle | I work around this issue by using the fully qualified name in the query, as:
```
echo "CREATE TEMPORARY TABLE xt1 AS SELECT * FROM test1.x" | curl --user default:`cat ~/sec/clickhouse | tr -d '\n'` -sS --data-binary @- 'http://192.168.1.27:8123/?database=test1&session_id=12345'
``` | 2019-12-22T16:00:51 | 2019-12-22T16:00:51 | {} | NONE |
ClickHouse | ClickHouse | 568,288,267 | 8,313 | filimonov | Maybe Kafka timestamp (accessible via _timestamp virtual column) will help you?
Also some minimal support for transactions is planned. | 2019-12-22T17:56:06 | 2019-12-22T17:56:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,342,531 | 8,159 | weiqxu | HDFS HA mode is already support. you need put the hdfs-client.xml in the working directory. and then use the namespace uri in the HDFS URI as you used in Hadoop componments. | 2019-12-23T03:35:11 | 2019-12-23T03:35:11 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,615,767 | 8,248 | KochetovNicolai | Build is broken :( | 2019-12-19T18:52:01 | 2019-12-19T18:52:01 | {} | MEMBER |
ClickHouse | ClickHouse | 566,704,220 | 8,194 | KochetovNicolai | Conflicts with 19.16 and 19.17 and 19.18 | 2019-12-17T19:07:16 | 2019-12-20T13:48:54 | {} | MEMBER |
ClickHouse | ClickHouse | 548,806,118 | 7,576 | alexey-milovidov | This can be implemented by sending the same `TSTP` signal. But in signal handler we can look at the source of this signal. | 2019-11-01T14:22:38 | 2019-12-22T20:23:57 | {} | MEMBER |
ClickHouse | ClickHouse | 568,336,067 | 7,195 | listenerri | I found a workaround, referring to the [test sql script](https://github.com/ClickHouse/ClickHouse/commit/7f28e9bcfb44fa08c56edfc6b872afde26f5b4ab#diff-626c5d0c090623bc2c5fd634049e9ea8) in this PR: #6324
The content of `test sql script` (Works well for recursive MV):
```
DROP TABLE IF EXISTS test.src;
DROP TABLE IF ... | 2019-12-23T02:53:06 | 2019-12-23T02:53:06 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 568,351,817 | 8,336 | kaijianding | `FROM dwh.source_table_local ... TO dwh.target_table_local` is the right way | 2019-12-23T04:38:20 | 2019-12-23T04:38:20 | {} | NONE |
ClickHouse | ClickHouse | 568,409,832 | 8,336 | godintsov | @kaijianding `FROM dwh.source_table ... TO dwh.target_table` works as well.
I want the data to be resharded after mat. view, so I want to insert into Distributed table inside the mat. view, but if selecting from local table I got duplicates. | 2019-12-23T09:03:30 | 2019-12-23T09:03:30 | {} | NONE |
ClickHouse | ClickHouse | 568,005,079 | 7,666 | SaltTan | I have just realized that the exception only happens when I sort the data:
```
SELECT
version(),
uptime()
FROM cluster(segmented, system, one)
┌─version()──┬─uptime()─┐
│ 19.16.6.17 │ 1383195 │
│ 19.13.7.57 │ 3813312 │
│ 19.13.7.57 │ 1744510 │
│ 19.13.7.57 │ 3813483 │
│ 19.13.7.57 │ 1742709 │
... | 2019-12-20T17:12:41 | 2019-12-20T17:12:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,037,837 | 8,154 | laingawbl | https://github.com/ClickHouse/ClickHouse/pull/8252 | 2019-12-20T18:33:41 | 2019-12-20T18:33:41 | {} | NONE |
ClickHouse | ClickHouse | 568,041,750 | 8,311 | alexey-milovidov | ORDER BY became much slower. | 2019-12-20T18:47:37 | 2019-12-20T18:47:37 | {} | MEMBER |
ClickHouse | ClickHouse | 568,055,986 | 8,322 | den-crane | I use 5 different solutions in one project. Different solutions fit for different cases.
1. re-create table by inserting to another table without duplicates.
2. select from T final
3. select argMax(*,)
4. cron / optimize
5. order by limit by
| 2019-12-20T19:20:52 | 2019-12-20T21:29:33 | {
"+1": 4,
"heart": 3
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,203,265 | 8,313 | svladykin | @den-crane
We write data from Kafka to Clickhouse and collect batches of fixed size per partition separately. This way we guarantee at least once processing of a batch. Probably we could use `kafka_time` but we have more than 1 partition and partitions may go at different speeds. | 2019-12-21T18:53:09 | 2019-12-21T18:53:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,245,455 | 8,311 | alexey-milovidov | Performance test is Ok and sorting performance was fought back almost two times. | 2019-12-22T09:24:39 | 2019-12-22T09:24:39 | {} | MEMBER |
ClickHouse | ClickHouse | 568,369,208 | 8,159 | weiqxu | @gubinjie you can simplify rename the copy the hdfs-site.xml to hdfs-client.xml. the file should be put in working directory or you can use environment variable $LIBHDFS3_CONF to point the configuration file.
Actually, it use the libhdfs3 which from Apache HAWQ to read the HDFS. the configuration should be same. | 2019-12-23T06:12:18 | 2019-12-23T06:12:18 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,749,255 | 5,169 | alexey-milovidov | We don't depend on system LLVM anymore: https://github.com/ClickHouse/ClickHouse/pull/8258 | 2019-12-20T01:31:39 | 2019-12-20T01:31:39 | {} | MEMBER |
ClickHouse | ClickHouse | 567,899,714 | 8,298 | 4ertus2 | #7556 #8254 #7915 #8100 | 2019-12-20T11:52:05 | 2019-12-20T11:52:05 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,908,674 | 8,308 | alexey-milovidov | Perf test - results are untrue. | 2019-12-20T12:28:26 | 2019-12-20T12:28:26 | {} | MEMBER |
ClickHouse | ClickHouse | 567,960,087 | 8,295 | matwey | Yes, they are. | 2019-12-20T15:18:23 | 2019-12-20T15:18:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,965,064 | 7,969 | chukynax | 19.17.5.18
so after lots of testing i finally made cluster of kafka tables run without cppkafka error .... but in all of the versions i recieve only part of all the messages sent, if i sent 10 000 i get about 9100 from php script, and 9758 from python script, though in kafka consumer i can see all sent messages 10 00... | 2019-12-20T15:32:35 | 2019-12-20T15:32:35 | {} | NONE |
ClickHouse | ClickHouse | 568,154,035 | 8,228 | ashkank66 | That's right, thank you. | 2019-12-21T05:25:32 | 2019-12-21T05:25:32 | {} | NONE |
ClickHouse | ClickHouse | 568,227,153 | 8,329 | amosbird | This leads to breaking changes.
`select 'name' as n, dictGet(n, 'i8', k)` is parsed with dict name `n` instead of `name`.
Perhaps an option is needed.
| 2019-12-22T02:21:10 | 2019-12-22T02:21:10 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 568,343,247 | 8,242 | ghost | @alexey-milovidov
I fixed an additional bug that referencing an invalid offset in bloom_filter with array.
Thank you. | 2019-12-23T03:39:39 | 2019-12-23T03:54:51 | {} | NONE |
ClickHouse | ClickHouse | 567,616,701 | 8,016 | KochetovNicolai | Conflicts with 19.17
@filimonov can you please create a special pr for kafka fixes to 19.17? | 2019-12-19T18:54:23 | 2019-12-19T18:54:23 | {} | MEMBER |
ClickHouse | ClickHouse | 566,202,087 | 5,385 | kolen | `-Wno-ctad-maybe-unsupported` is still a problem in latest master (0bc1c6d91d966778f2443640282d79b7919242ad).
```
[119/4432] Building CXX object dbms/CMakeFiles/clickhouse_malloc.dir/src/Common/malloc.cpp.o
FAILED: dbms/CMakeFiles/clickhouse_malloc.dir/src/Common/malloc.cpp.o
/Library/Developer/CommandLineTools/u... | 2019-12-16T19:17:06 | 2019-12-20T01:19:26 | {} | NONE |
ClickHouse | ClickHouse | 567,943,528 | 8,301 | filimonov | Rather it's an issue.
Motivation: when people look into logs and see SOME error message related to the filesystem operations, and a message itself has 'malformed' path they start to think that malformed path is the reason of the issue (while it's just a cosmetic). | 2019-12-20T14:29:28 | 2019-12-20T14:29:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,963,270 | 8,301 | alexey-milovidov | Let's fix that. | 2019-12-20T15:27:38 | 2019-12-20T15:27:38 | {} | MEMBER |
ClickHouse | ClickHouse | 568,245,714 | 8,311 | alexey-milovidov | And there was unexpected speedup in some queries. For example, formatting in Pretty formants was improved about 40%. Overall performance improvement is about 2%. | 2019-12-22T09:28:10 | 2019-12-22T09:32:58 | {
"laugh": 1
} | MEMBER |
ClickHouse | ClickHouse | 567,646,150 | 8,278 | den-crane | # SummingMergeTree
```
CREATE MATERIALIZED VIEW results_histogram_by_hour
ENGINE = SummingMergeTree()
PARTITION BY toYYYYMMDD(date_time)
ORDER BY (date_time, item_id, value)
POPULATE
AS SELECT toStartOfHour(date_time) AS date_time,
item_id,
intDiv(value,100) * 100 AS value,
sum... | 2019-12-19T20:15:35 | 2019-12-19T20:18:19 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,739,861 | 7,546 | alexey-milovidov | `unknown warning option '-Wno-ctad-maybe-unsupported'`
Fixed in https://github.com/ClickHouse/ClickHouse/commit/cea2903e72d4378b382f9b8ae616c01f512483dc#diff-eabb2b894359aa83d33168e3117959dcR55 | 2019-12-20T00:44:47 | 2019-12-20T00:44:47 | {} | MEMBER |
ClickHouse | ClickHouse | 567,744,652 | 8,309 | alexey-milovidov | Requires https://github.com/ClickHouse/ClickHouse/pull/8311 | 2019-12-20T01:08:37 | 2019-12-20T01:08:37 | {} | MEMBER |
ClickHouse | ClickHouse | 567,947,411 | 8,295 | filimonov | So you mean that util are linked with more threads? | 2019-12-20T14:41:10 | 2019-12-20T14:41:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,000,962 | 8,240 | alexey-milovidov | Stress test - issue with DateTime64 type that was added by @Enmk | 2019-12-20T16:59:59 | 2019-12-20T16:59:59 | {} | MEMBER |
ClickHouse | ClickHouse | 568,153,072 | 8,328 | den-crane | 19.5.3.1 is not supported. It could be fixed already.
Try to reproduce it with 19.15+ | 2019-12-21T05:05:44 | 2019-12-21T05:05:44 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,203,601 | 8,313 | svladykin | @alexey-milovidov
Maybe it would be a good idea if deduplication will consider columns with `DEFAULT` expression but ignore `MATERIALIZED` columns. This way it will be possible to have 'insert_time MATERILAIZED now()' column and still have working deduplication. | 2019-12-21T18:58:20 | 2019-12-21T18:58:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,274,893 | 7,997 | ghuname | I hope that when optimization is in the question that clickhouse knows that it should be execute ones on each shard because of replication. | 2019-12-22T15:56:05 | 2019-12-22T15:56:05 | {} | NONE |
ClickHouse | ClickHouse | 568,287,890 | 8,299 | filimonov | > It is fixed in #7975
> Btw, which program did you use to display protocol packets?
Looks like wireshark | 2019-12-22T17:52:45 | 2019-12-22T17:52:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,321,811 | 7,195 | listenerri | I confirmed that the problem started with tag: `19.11.12.69`. Guess was introduced by some commit to fix `comp-matview`. The last working tag was: `19.11.11.57`.
|CH version|MV2 based on MV1|Use multiple `Join` in MV|
|----|----|----|
|`19.11.11.57`|√|×|
|`19.11.12.69`|×|√| | 2019-12-23T01:14:27 | 2019-12-23T02:49:41 | {} | NONE |
ClickHouse | ClickHouse | 568,414,438 | 7,136 | zhang2014 | @alexey-milovidov
I currently trying to fix all known `PredicateOptimizer` bug in github issues list, but it may take one to two weeks. | 2019-12-23T09:19:32 | 2019-12-23T09:19:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,250,127 | 8,333 | CurtizJ | I found, that it was broken in #7392.
Query result using [build](https://clickhouse-builds.s3.yandex.net/0/18a72fa91a714e1a3383a48de17f954f6a140664/report.html) from merge commit of this PR.
```
:) select version()
┌─version()────┐
│ 19.17.1.1552 │
└──────────────┘
1 rows in set. Elapsed: 0.005 sec.
... | 2019-12-22T10:32:45 | 2019-12-22T10:32:45 | {} | MEMBER |
ClickHouse | ClickHouse | 567,608,058 | 8,303 | nikitamikhaylov | The changes affected only `has` function. Need to test it and then rewrite all other functions in CacheDictionary. | 2019-12-19T18:30:41 | 2019-12-19T18:30:41 | {} | MEMBER |
ClickHouse | ClickHouse | 566,691,035 | 8,079 | KochetovNicolai | Conflicts with 19.16 and 19.17 | 2019-12-17T18:32:48 | 2019-12-19T19:07:54 | {} | MEMBER |
ClickHouse | ClickHouse | 567,745,644 | 8,300 | kolen | This issue is not about `-Wno-ctad-maybe-unsupported`, that error just interfered to reproduce described problem cleanly. I updated output in this issue with build log from current master (cea2903e72d4378b382f9b8ae616c01f512483dc).
This issue is about `error: unknown directive`, `error: invalid alignment value`, ...... | 2019-12-20T01:13:37 | 2019-12-20T01:13:37 | {} | NONE |
ClickHouse | ClickHouse | 567,804,087 | 8,085 | tisonkun | Still the problem reproduced. Any news at least what is the cause? | 2019-12-20T06:16:10 | 2019-12-20T06:16:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,914,907 | 5,653 | filimonov | ```
create table json_insert (a UInt32 DEFAULT 1, b UInt32 DEFAULT 1) Engine=Log;
insert into json_insert FORMAT JSONEachRow {"a":2}
```
```
Exception on client:
Code: 9. DB::Exception: Mismatch column sizes while adding defaults
Stack trace:
0. 0x55b34754d147 StackTrace::StackTrace() /build/obj-x86_64-linux-... | 2019-12-20T12:53:08 | 2019-12-20T12:59:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 567,924,749 | 8,281 | alexey-milovidov | Ok. This warning is way too much for us. | 2019-12-20T13:30:30 | 2019-12-20T13:30:30 | {} | MEMBER |
ClickHouse | ClickHouse | 567,958,456 | 5,773 | alexey-milovidov | `01019_Buffer_and_max_memory_usage` - flap. | 2019-12-20T15:13:31 | 2019-12-20T15:13:31 | {} | MEMBER |
ClickHouse | ClickHouse | 568,223,638 | 8,314 | alexey-milovidov | For OPTIMIZE, it is 100% ok.
For DROP/DETACH/REPLACE PARTITION or CLEAR COLUMN IN PARTITION it is questionable, because a race condition exists: it's possible that some replicas will become active and after this query there are active replicas that still didn't execute the command. And it will be unclear what semant... | 2019-12-22T00:58:14 | 2019-12-22T00:59:55 | {} | MEMBER |
ClickHouse | ClickHouse | 568,239,618 | 8,299 | yurriy | It is fixed in #7975
Btw, which program did you use to display protocol packets? | 2019-12-22T07:37:26 | 2019-12-22T07:37:26 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,246,590 | 8,326 | alexey-milovidov | `SELECT runningAccumulate()` leads to crash. | 2019-12-22T09:41:09 | 2019-12-22T09:41:09 | {} | MEMBER |
ClickHouse | ClickHouse | 568,246,758 | 8,130 | alexey-milovidov | We can also add a perf test. | 2019-12-22T09:43:58 | 2019-12-22T09:43:58 | {} | MEMBER |
ClickHouse | ClickHouse | 567,749,337 | 5,169 | alexey-milovidov | Also we started to support LLVM 9 and 8. | 2019-12-20T01:32:03 | 2019-12-20T01:32:03 | {} | MEMBER |
ClickHouse | ClickHouse | 567,909,770 | 5,773 | alexey-milovidov | Perf tests are irrelevant to this commit but need to check this additionaly:
```
SELECT count() FROM system.numbers WHERE NOT ignore(exp2(toFloat64(number)))
``` | 2019-12-20T12:32:49 | 2019-12-20T12:32:49 | {} | MEMBER |
ClickHouse | ClickHouse | 568,148,936 | 7,195 | listenerri | Is there any progress ? | 2019-12-21T03:41:16 | 2019-12-21T03:41:16 | {} | NONE |
ClickHouse | ClickHouse | 553,420,826 | 7,603 | kwai-code | sorry for paste screenshot
----------------------------
```
2019.11.13 22:03:41.428343 [ 157 ] {15D56D1912F3E289} <Error> virtual DB::UnionBlockInputStream::~UnionBlockInputStream(): Code: 394, e.displayText() = DB::Exception: Query was cancelled:
Code: 394, e.displayText() = DB::Exception: Query was cancelled (ver... | 2019-11-13T14:11:55 | 2019-12-23T09:27:31 | {} | NONE |
ClickHouse | ClickHouse | 567,739,571 | 8,148 | alexey-milovidov | Unknown flag: fixed in https://github.com/ClickHouse/ClickHouse/commit/cea2903e72d4378b382f9b8ae616c01f512483dc#diff-eabb2b894359aa83d33168e3117959dcR55
Build instruction: obsolete. | 2019-12-20T00:43:22 | 2019-12-20T00:43:22 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 567,739,980 | 7,995 | alexey-milovidov | `unknown warning option '-Wno-ctad-maybe-unsupported'`
Fixed in https://github.com/ClickHouse/ClickHouse/commit/cea2903e72d4378b382f9b8ae616c01f512483dc#diff-eabb2b894359aa83d33168e3117959dcR55 | 2019-12-20T00:45:18 | 2019-12-20T00:45:18 | {} | MEMBER |
ClickHouse | ClickHouse | 567,785,897 | 8,293 | vkingnew | @den-crane It's not easy to read for non-Russian users，It's better to translate it into English and update it to official docments.As so far I just to read it by translate software,maybe it not exact. | 2019-12-20T04:47:12 | 2019-12-20T04:47:50 | {} | NONE |
ClickHouse | ClickHouse | 567,984,347 | 8,277 | Anwert | Got it, thank you for answer | 2019-12-20T16:27:42 | 2019-12-20T16:27:42 | {} | NONE |
ClickHouse | ClickHouse | 568,255,963 | 7,136 | alexey-milovidov | @zhang2014 This is still relevant. | 2019-12-22T12:02:35 | 2019-12-22T12:02:35 | {} | MEMBER |
ClickHouse | ClickHouse | 568,086,574 | 8,313 | den-crane | If you re-generate a batch at ETL the second time how you guarantee the same row-order and the same row-set for deduplication? | 2019-12-20T21:02:29 | 2019-12-20T21:02:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 568,404,749 | 8,267 | alesapin | Ohh, tried to fix known issue https://github.com/ClickHouse/ClickHouse/issues/8233. Our server doesn't correctly stop after full functional stateless tests run, so we don't have `.profraw` coverage file. @Akazz investigating the issue.
This pull request can be merged. | 2019-12-23T08:43:32 | 2019-12-23T08:43:32 | {} | MEMBER |
ClickHouse | ClickHouse | 569,112,494 | 8,329 | alexey-milovidov | I don't think it's good to add a setting. Because this setting is going to be very rarely used but we would have to always keep it in mind. | 2019-12-26T18:41:39 | 2019-12-26T18:41:39 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.