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 | 569,587,615 | 8,221 | kreuzerkrieg | > We only rolled back `-Wmissing-declarations` and related changes because it requires to collect registration functions in headers, and when we edit these headers adding a new function, it leads to long rebuild time.
Legit step in case you dont PCH your build. BTW, AFAIR there was some opened issue regarding Unity ... | 2019-12-30T05:26:21 | 2019-12-30T05:26:21 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 569,703,824 | 8,290 | CurtizJ | Huge part of the failed tests checks some implementation defined things like checksums, computed granularity, ProfileEvents, etc. So, it's expected that they will fail with compact parts. Tests, that uses `checkDataPart` function also fails as this function is not implemented for compact parts yet.
Other failing sta... | 2019-12-30T15:06:10 | 2019-12-30T15:06:10 | {
"+1": 1
} | MEMBER |
ClickHouse | ClickHouse | 570,191,570 | 8,490 | vincenzon | @den-crane I made the fix you suggested, where c1 is not null, and it is working. Thank you for looking into this. | 2020-01-02T12:16:09 | 2020-01-02T12:16:09 | {} | NONE |
ClickHouse | ClickHouse | 569,562,931 | 8,455 | amosbird | ```
localhost:9001, queries 78, QPS: 3.155, RPS: 1369816.586, MiB/s: 1211.481, result RPS: 0.000, result MiB/s: 0.000.
0.000% 0.277 sec. 0.270 sec.
10.000% 0.284 sec. 0.282 sec.
20.000% 0.291 sec. 0.291 sec.
30.000% 0.298 sec. 0.297 sec.
40.000% 0.307... | 2019-12-30T02:00:16 | 2019-12-30T02:00:16 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 569,566,117 | 7,696 | alexey-milovidov | Fixed. | 2019-12-30T02:33:57 | 2019-12-30T02:33:57 | {} | MEMBER |
ClickHouse | ClickHouse | 569,844,805 | 8,014 | volfco | @achimbab I don't have a solid idea of the conditions that existed for this to happen. So let's assume #8242 fixes it and I'll re-open this or link it to a new issue if something like this happens again | 2019-12-31T01:34:05 | 2019-12-31T01:34:05 | {} | NONE |
ClickHouse | ClickHouse | 569,871,831 | 7,769 | alexey-milovidov | Do I understand correctly, that:
- we avoid filtering data twice if the filter in PREWHERE does not filter much?
> As a result, any viable predicates can be moved into prewhere without fancy heuristic
But we already doing it (almost: we exclude conditions on the first primary key column).
And the choice of cond... | 2019-12-31T06:15:09 | 2019-12-31T06:15:09 | {} | MEMBER |
ClickHouse | ClickHouse | 569,973,997 | 8,490 | vincenzon | @den-crane the version is: 19.17.5 revision 54428
I can't share the real tables, but edited versions are:
```
CREATE TABLE t1 (`cA` String, `c1` String)
ENGINE = MergeTree
PARTITION BY substring(cA, 1, 7) ORDER BY (cA, c1)
SETTINGS index_granularity = 8192
```
```
CREATE TABLE t2 (`cB` String, `c2` Float6... | 2019-12-31T18:53:24 | 2019-12-31T18:53:24 | {} | NONE |
ClickHouse | ClickHouse | 570,091,938 | 7,792 | vzakaznikov | @KochetovNicolai, I've merged your changes from https://github.com/ClickHouse/ClickHouse/pull/8217 and made a few changes to simplify extractDependentTable function. Could you see if we could merge this PR so that subqueries are supported in LV tables? Thanks! | 2020-01-01T23:19:42 | 2020-01-01T23:19:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,117,812 | 8,498 | den-crane | Make a View.
create view ptx_db.vor_xc_order_item_years_view as select
pay_nextyear, sku, account, country, asin, company_sk, sumMerge....
group by ....
and use this vor_xc_order_item_years_view in your dictionary. | 2020-01-02T04:50:27 | 2020-01-02T04:50:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,376,887 | 8,506 | den-crane | >B::Exception: Memory limit (total) exceeded: would use 74.51 GiB (attempt to allocate chunk of >4217732 bytes), maximum: 74.51 GiB (version 19.17.4.11)
**(total) exceeded**
It could be a problem from other sessions which set max_memory_usage_for_all_queries=74GB because CH has an issue with `max_memory_usage_for_... | 2020-01-02T22:22:48 | 2020-01-02T22:22:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,436,720 | 8,488 | den-crane | 10.230.226.20, 10.230.226.21 are replicas in the same shard so it's a replication delay because CH has eventual consistency. You can try stabilize results using load_balancing -- in_order or using quorum insert.
Usually replication delay is about several seconds check ch logs for errors.
And check delay:
http://... | 2020-01-03T02:17:30 | 2020-01-03T02:24:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,605,647 | 8,514 | alexey-milovidov | > each test uses new random database, which is not true
That was added a few month ago. | 2020-01-03T15:31:49 | 2020-01-03T15:31:49 | {} | MEMBER |
ClickHouse | ClickHouse | 569,739,102 | 8,474 | alexey-milovidov | Because of removal of PREWHERE columns, that was added by @KochetovNicolai
We should not remove columns that have dependents. | 2019-12-30T17:19:57 | 2019-12-30T17:19:57 | {} | MEMBER |
ClickHouse | ClickHouse | 569,741,395 | 8,460 | alexey-milovidov | Missing integration test. | 2019-12-30T17:28:52 | 2019-12-30T17:28:52 | {} | MEMBER |
ClickHouse | ClickHouse | 569,745,057 | 8,221 | alexey-milovidov | > opened issue regarding Unity build/cotire - any progress?
No progress at all. | 2019-12-30T17:42:38 | 2019-12-30T17:42:38 | {} | MEMBER |
ClickHouse | ClickHouse | 569,967,281 | 8,486 | amosbird | oops, missing
```
DROP TABLE test_prewhere_default_column;
DROP TABLE test_prewhere_column_type;
```
at the end... | 2019-12-31T18:02:19 | 2019-12-31T18:02:19 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 570,111,904 | 8,492 | zhicwu | Thanks @ramazanpolat, it works for sure when the table is small(e.g. less than 1m rows?). However, it's not in our case since push-down is not supported at this point. Also please excuse my bad example, adhoc query is usually more complex(e.g. with table joins or aggregations etc.). | 2020-01-02T03:40:42 | 2020-01-02T03:40:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,356,429 | 8,358 | tdingus | In order for openssl to work, there has to be a key <secure>1</secure> as indicated in the document:
https://clickhouse.yandex/docs/en/operations/table_engines/distributed/
| 2020-01-02T21:26:20 | 2020-01-02T21:26:20 | {} | NONE |
ClickHouse | ClickHouse | 570,371,333 | 8,506 | genzgd | Sorry, I was thinking of it as max_memory_usage "per_query" since that's how it's referenced in the documentation and in error logs:
```
SELECT
name,
value
FROM system.settings
WHERE name LIKE 'max%mem%'
┌─name─────────────────────────────┬─value────────┐
│ max_memory_usage │ 10737... | 2020-01-02T22:08:09 | 2020-01-02T22:08:09 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,433,220 | 8,488 | wsxGit | after run several times
`select count(), hostName()
from tables where toYYYYMMDD(time) = 20191212
group by hostName()`
We found that the results of **10.230.226.20** is **inconsistent** with **10.230.226.21**,
but **several minutes later**, result is **consistent**
| 2020-01-03T01:53:44 | 2020-01-03T02:07:35 | {} | NONE |
ClickHouse | ClickHouse | 569,563,180 | 6,528 | alexey-milovidov | It's implemented by @kononencheg in #8326 | 2019-12-30T02:02:48 | 2019-12-30T02:02:48 | {} | MEMBER |
ClickHouse | ClickHouse | 569,620,558 | 7,057 | orloffv | ping? | 2019-12-30T09:07:23 | 2019-12-30T09:07:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 569,741,115 | 8,458 | alexey-milovidov | Sanitizer report and integration tests failure look relevant. | 2019-12-30T17:27:53 | 2019-12-30T17:27:53 | {} | MEMBER |
ClickHouse | ClickHouse | 570,098,665 | 8,488 | wsxGit | @den-crane
### config.xml
`<?xml version="1.0"?>
<!-- Ansible managed -->
<yandex>
<remote_servers>
<cluster_uom_alert>
<shard>
<internal_replication>true</internal_replication>
<replica>
<host>10.230.226.16</host>
<port>9000</port>
... | 2020-01-02T01:03:23 | 2020-01-02T01:56:51 | {} | NONE |
ClickHouse | ClickHouse | 570,112,599 | 8,466 | hczhcz | Running untrusted SQL code is generally not safe.
Two possible ideas:
* Give each user an isolated ClickHouse instance
* Reconstruct a restricted query from the user's input | 2020-01-02T03:48:30 | 2020-01-02T03:48:30 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,354,794 | 8,501 | alexey-milovidov | assert in `Functional stateless tests (debug)`, unrelated, will be fixed separately. | 2020-01-02T21:22:04 | 2020-01-02T21:22:04 | {} | MEMBER |
ClickHouse | ClickHouse | 570,624,840 | 4,717 | SaltTan | I have checked 19.17.6.36.
With `dictionaries_lazy_load=false` CH starts up and then loads dictionaries asynchronously:
```
2020.01.03 16:13:58.949511 [ 1 ] {} <Information> Application: Listening http://[::]:8123
2020.01.03 16:13:58.949537 [ 1 ] {} <Information> Application: Listening for connections with native p... | 2020-01-03T16:34:47 | 2020-01-03T16:34:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 569,563,364 | 6,528 | alexey-milovidov | @zombiemonkey
> Have realized this isn't going to work because of processing per block. Is there any way to force consecutive blocks to synchronize?
Yes, all "stateful" functions: `runningDifference`, `runningAccumulate`, `neighbor` are working semi-legally from architecture standpoint. It should be fixed in mor... | 2019-12-30T02:04:58 | 2019-12-30T02:04:58 | {} | MEMBER |
ClickHouse | ClickHouse | 569,563,925 | 2,628 | alexey-milovidov | @nikitamikhaylov
Is it fixed? | 2019-12-30T02:11:43 | 2019-12-30T02:11:43 | {} | MEMBER |
ClickHouse | ClickHouse | 569,760,186 | 8,329 | amosbird | I guess we already have those exclusions..
```
SELECT 'CSV' AS CSV
FROM file('/dev/null', CSV, 'i int')
Ok.
```
I'm inclined to error out when there is an ambiguity with aliases and identifier names. | 2019-12-30T18:41:07 | 2019-12-30T18:41:07 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 569,872,415 | 7,769 | alexey-milovidov | I've merged this PR, but there are some drawbacks:
1. There is no performance tests added.
2. The threshold 0.6 is not controlled by setting and there is no comments about how it was chosen.
3. Some style issues like `num_rows = total_rows_per_granule = new_filter->size();`
Could you please address these is fol... | 2019-12-31T06:20:42 | 2019-12-31T06:20:42 | {} | MEMBER |
ClickHouse | ClickHouse | 569,872,492 | 7,769 | alexey-milovidov | If the issue #8474 is fixed, could you please also add a test? | 2019-12-31T06:21:33 | 2019-12-31T06:21:33 | {} | MEMBER |
ClickHouse | ClickHouse | 569,986,079 | 7,262 | kiransunkari | Hi, I want to take this issue. | 2019-12-31T20:26:40 | 2019-12-31T20:26:40 | {} | NONE |
ClickHouse | ClickHouse | 570,263,647 | 8,501 | alexey-milovidov | Failed perftest of `exp2` - unrelated. | 2020-01-02T16:39:36 | 2020-01-02T16:39:36 | {} | MEMBER |
ClickHouse | ClickHouse | 570,376,780 | 8,506 | genzgd | To be clear, I assumed that settings was not being referenced, since error messages for that setting look like `Memory limit (for query) exceeded`, not `Memory limit (total) exceeded`. Examples of the "total" memory limit being exceeded are very rare based on my Google search. | 2020-01-02T22:22:31 | 2020-01-02T22:22:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 569,787,418 | 8,478 | alexey-milovidov | Ok. Added a star just in case. | 2019-12-30T20:28:02 | 2019-12-30T20:28:02 | {} | MEMBER |
ClickHouse | ClickHouse | 569,944,195 | 8,488 | den-crane | I believe your cluster is misconfigured.
Please provide XML with remote_servers and DDL `show create table` for your Distributed table.
And `select count() from shard_table` from each node (replica/shard) | 2019-12-31T15:17:56 | 2019-12-31T15:18:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 569,996,895 | 8,491 | den-crane | Not a CH issue. This is not CSV.
CH supports CSV with any delimiter and with string quoting by " or ' with escaping by the quote with same quote type ( `aaaa|"sdfsdfsdfd""fsdsf"|"sdfsd"` ).
Try TSV. CH TSV does not need quoting. | 2019-12-31T22:01:10 | 2019-12-31T22:02:53 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,440,500 | 8,488 | wsxGit | execute` curl http://10.230.226.20:8123/replicas_status`
always return **OK**
execute` curl http://10.230.226.20:8123/replicas_status?verbose=1`
Some tables appear occasionally **Absolute delay: 47. Relative delay: 0.** | 2020-01-03T02:40:08 | 2020-01-03T02:42:02 | {} | NONE |
ClickHouse | ClickHouse | 570,520,677 | 6,638 | alexey-milovidov | @akuzm Did you fail? | 2020-01-03T09:46:16 | 2020-01-03T09:46:16 | {} | MEMBER |
ClickHouse | ClickHouse | 570,535,304 | 8,514 | azat | P.S. FWIW I though that each test uses new random database, which is not true, but this could significantly relax the naming scheme in tests, thoughts? | 2020-01-03T10:38:43 | 2020-01-03T10:38:43 | {} | MEMBER |
ClickHouse | ClickHouse | 570,592,975 | 8,509 | den-crane | >Not really sure what you wanted to achieve here.
I meant this:
```
CREATE TABLE IF NOT EXISTS dataDate201912_local (
statdate Date,stattime DateTime,ip UInt32,domain String,type UInt8,flow UInt64,request UInt64)
ENGINE = SummingMergeTree PARTITION BY statdate ORDER BY (statdate,stattime,ip,domain,type);
i... | 2020-01-03T14:49:45 | 2020-01-03T14:54:21 | {
"hooray": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 569,967,447 | 8,418 | alexey-milovidov | After https://github.com/ClickHouse/ClickHouse/pull/8483 this task can be postponed. | 2019-12-31T18:03:38 | 2019-12-31T18:03:38 | {} | MEMBER |
ClickHouse | ClickHouse | 570,381,683 | 8,506 | den-crane | I don't remember the issue number.
STR:
```
for i in `seq 1 5000`; do echo -n " Result: "; clickhouse-client --max_memory_usage_for_all_queries=100000 -q "select sleep(1)"; done
clickhouse-client --max_memory_usage_for_all_queries=0 --max_memory_usage=0 --max_memory_usage_for_user=0 -q "create table X Engine=... | 2020-01-02T22:38:28 | 2020-01-02T22:43:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,406,255 | 8,488 | den-crane | try to run several times
```
select count(), hostName()
from tables where toYYYYMMDD(time) = 20191212
group by hostName()
```
to get the difference, what nodes how many rows return | 2020-01-02T23:57:35 | 2020-01-02T23:57:35 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,461,506 | 8,509 | den-crane | Your SummingMergeTree sums flow column, but AggregatingMergeTree calcs max.
Not really sure what you wanted to achieve here. You can calculate maxState in SummingMergeTree as well.
| 2020-01-03T04:48:52 | 2020-01-03T04:48:52 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,533,724 | 8,467 | yuzhichang | @alexey-milovidov Any comments for this PR? By the way, I don't think the failed cases relate to my change. | 2020-01-03T10:33:28 | 2020-01-03T10:33:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 569,579,302 | 8,429 | zhang2014 | > It's about all the clients that use Native format ?
I checked it again. It will affect XML, TabSeparated, Native, JSON output formats and ODBC Driver.
> Maybe we can do it in a different way:
>
> - do not use custom names inside DataType objects; use canonical names as before;
> - save user-specified names ... | 2019-12-30T04:26:23 | 2019-12-30T04:28:39 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 569,971,779 | 8,490 | den-crane | Please provide
1. ClickHouse version
2. show create table t1 / t2 / t3 | 2019-12-31T18:37:24 | 2019-12-31T18:37:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,056,612 | 8,494 | den-crane | Duplicate https://github.com/ClickHouse/ClickHouse/issues/4470
comment the ssl_conf = ssl_sect line in your /etc/ssl/openssl.cnf under [default_conf] | 2020-01-01T14:14:41 | 2020-01-01T14:14:41 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,092,182 | 8,492 | ramazanpolat | > 1. adhoc query
> ```sql
> select * from jdbc('datasource://some-db', '',
> 'select * from some_table order by created desc limit 10')
> ```
Why not using the following way?
```sql
select * from jdbc('datasource://some-db', '', 'some_table') order by created desc limit 10
``` | 2020-01-01T23:23:40 | 2020-01-01T23:23:40 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,563,619 | 7,792 | KochetovNicolai | Cool! | 2020-01-03T12:49:40 | 2020-01-03T12:49:40 | {} | MEMBER |
ClickHouse | ClickHouse | 570,570,195 | 7,666 | vapongit | some more examples (19.17.6.36).
1. Correct
```sql
SELECT
version() AS v,
hostName() AS h,
uptime() AS u
FROM cluster(cluster, system, one)
```
```
┌─v──────────┬─h────────────┬──────u─┐
│ 19.17.6.36 │ clickhouse05 │ 609636 │
└────────────┴──────────────┴────────┘
┌─v──────────┬─h───────────... | 2020-01-03T13:21:50 | 2020-01-03T13:21:50 | {} | NONE |
ClickHouse | ClickHouse | 569,559,337 | 6,034 | alexey-milovidov | Need info on query and dataset.
(Dataset can be prepared with `clickhouse-obfuscator`)
What is the query result size?
How the execution time changes if you enable `distributed_aggregation_memory_efficient`? | 2019-12-30T01:14:55 | 2019-12-30T01:14:55 | {} | MEMBER |
ClickHouse | ClickHouse | 569,772,747 | 8,474 | alexey-milovidov | Sorry, I have missed that PR. Will keep up soon. | 2019-12-30T19:30:36 | 2019-12-30T19:30:36 | {} | MEMBER |
ClickHouse | ClickHouse | 569,844,378 | 8,014 | ghost | @volfco
Some bugs are fixed at https://github.com/ClickHouse/ClickHouse/pull/8242.
If you are yet interested in the bloom_filter, please retest the perfect storm of pending merges.
And if the result of the test is fine then please close this issue,
Thank you. | 2019-12-31T01:30:04 | 2019-12-31T01:32:22 | {} | NONE |
ClickHouse | ClickHouse | 570,100,887 | 8,488 | wsxGit | ### other config
`Node 226.16
<yandex>
<macros>
<replica>01</replica>
<shard>01</shard>
<replica_alert>01</replica_alert>
<shard_alert>01</shard_alert>
</macros>
</yandex>
Node:226.17
<yandex>
<macros>
<replica>02</replica>
<shard>01</shard>
<replica_alert>02</replica_al... | 2020-01-02T01:35:51 | 2020-01-02T01:37:15 | {} | NONE |
ClickHouse | ClickHouse | 570,383,690 | 8,506 | genzgd | Will do,thanks so much! | 2020-01-02T22:44:10 | 2020-01-02T22:44:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,574,007 | 8,320 | crakjie | @alexey-milovidov No I did not. In this case I could have understood.
@amosbird mmmh big change for existing queries. That was actually tested with 19.15.3.6 | 2020-01-03T13:37:29 | 2020-01-03T13:37:29 | {} | NONE |
ClickHouse | ClickHouse | 570,593,077 | 1,118 | victor-perov | Hello @blinkov @alexey-milovidov!
Could you guys please share any news about plans on that feature?
We are trying to predict each query cost (based on the possible impact on ClickHouse cluster: such as read_rows, read_bytes, parts touched etc) on-fly in order to determine whether it's possible to run the query or n... | 2020-01-03T14:50:05 | 2020-01-03T14:50:05 | {
"+1": 16,
"eyes": 1
} | NONE |
ClickHouse | ClickHouse | 569,655,937 | 8,473 | Inasayang | sorry，Wrong configure | 2019-12-30T11:45:49 | 2019-12-30T11:45:49 | {} | NONE |
ClickHouse | ClickHouse | 569,872,213 | 7,769 | alexey-milovidov | Now we can go further and implement:
1. Do "prewhere" style filtering in MergeTree tables without modifying query AST (without actually involving PREWHERE).
2. Implement multi-step "prewhere" filtering when we read columns and evaluate conditions one by one, deciding whether to do filtering on each step.
Multist... | 2019-12-31T06:18:44 | 2019-12-31T06:18:44 | {} | MEMBER |
ClickHouse | ClickHouse | 569,876,602 | 6,943 | alexey-milovidov | @CurtizJ This issue should be addressed in your PR with implementation of compact data parts. | 2019-12-31T06:55:42 | 2019-12-31T06:55:42 | {} | MEMBER |
ClickHouse | ClickHouse | 570,379,331 | 8,506 | genzgd | That actually looks like the problem, we do have a different profile which just happens to have a 74.51 GB value for that `max_memory_usage_for_all_queries` setting. Do you have a link to that issue you mentioned? | 2020-01-02T22:30:51 | 2020-01-02T22:32:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,518,494 | 4,586 | alexey-milovidov | Implementation proposal:
- create a property of IDatabase that can say that all tables in database are "ephemeral" and should not be removed (by `removeTable`) on database drop.
| 2020-01-03T09:37:40 | 2020-01-03T09:37:40 | {} | MEMBER |
ClickHouse | ClickHouse | 569,602,760 | 8,461 | filimonov | Documentation should be fixed. Maybe test should be added (if absent) | 2019-12-30T07:22:54 | 2019-12-30T07:22:54 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 569,758,776 | 8,329 | 4ertus2 | > What happens after we found out it's an alias? Do we need to error out or use the aliased value instead?
I have no opinion which of them is better. But if we do not care about alias at all we'll have an exclusion in our aliases logic.
| 2019-12-30T18:35:35 | 2019-12-30T18:35:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,070,230 | 8,490 | den-crane | minimal example:
```
CREATE TABLE t1 (`cA` String, `c1` String) ENGINE = MergeTree
ORDER BY (cA, c1);
insert into t1 select 'AAAAAAAAAAA', 'BBBBBB';
for i in `seq 1 50`; do echo -n " Result: "; clickhouse-client --empty_result_for_aggregation_by_empty_set=0 -q "select count() from t1 where c1 in (select 'BBB... | 2020-01-01T17:38:39 | 2020-01-01T19:52:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,519,240 | 4,717 | alexey-milovidov | @filimonov Please validate the current behaviour (it was changed multiple times). | 2020-01-03T09:40:27 | 2020-01-03T09:40:27 | {} | MEMBER |
ClickHouse | ClickHouse | 570,609,324 | 8,514 | azat | > That was added a few month ago.
I guess you are talking about "each test run uses new random database", not "each test" (496376a6bcf0a42b8c) | 2020-01-03T15:44:12 | 2020-01-03T15:44:12 | {} | MEMBER |
ClickHouse | ClickHouse | 569,562,580 | 7,867 | alexey-milovidov | @nikitamikhaylov We should exclude all columns that are arguments of partition key expression from summing.
If the user explicitly specified that columns in the list of columns to sum - throw an exception. | 2019-12-30T01:56:08 | 2019-12-30T01:56:08 | {} | MEMBER |
ClickHouse | ClickHouse | 569,994,874 | 8,491 | BigBig-joe | in sqoop cmd, I specify escaped-by = \\\ , but the first \ is the escape for the second \ . so in the file, data shows like
> "9AAAAAADW5__A6BA_BARGAINS \\"N\\" DEALZ"
the test data looks like below
> "Regular","No","Domestic Inc, seller","BARGAINS \\"N\\" DEALZ"
and add the sqoop CMD
`/usr/local/sqoop/b... | 2019-12-31T21:39:42 | 2019-12-31T21:44:27 | {} | NONE |
ClickHouse | ClickHouse | 570,075,244 | 8,490 | den-crane | @alesapin
Seems the issue related to adaptive index granularity (only for parts with .mrk2) 19.10.+
No issue
```
CREATE TABLE t1 (`cA` String, `c1` String) ENGINE = MergeTree
ORDER BY (cA, c1)
settings index_granularity_bytes=0;
```
Issue
```
CREATE TABLE t1 (`cA` String, `c1` String) ENGINE = MergeTr... | 2020-01-01T18:56:34 | 2020-01-01T19:31:02 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,123,065 | 8,498 | leoones | > > Make a View.
> > create view ptx_db.vor_xc_order_item_years_view as select
> > pay_nextyear, sku, account, country, asin, company_sk, sumMerge....
> > group by ....
> > and use this vor_xc_order_item_years_view in your dictionary.
>
> I created the view and applied it in the dictionary, but the value obtaine... | 2020-01-02T05:45:42 | 2020-01-02T05:45:42 | {} | NONE |
ClickHouse | ClickHouse | 570,167,854 | 7,477 | alexey-milovidov | @Vinatorul Hi and happy new year! Hope you have success building the code? | 2020-01-02T10:23:48 | 2020-01-02T10:23:48 | {} | MEMBER |
ClickHouse | ClickHouse | 570,517,388 | 1,752 | alexey-milovidov | This issue will be addressed in #6787 | 2020-01-03T09:33:45 | 2020-01-03T09:33:45 | {} | MEMBER |
ClickHouse | ClickHouse | 570,601,386 | 8,512 | den-crane | You missed `storage_configuration` tag.
```
<storage_configuration>
<disks>
<data1>
<path>/data1/clickhouse/data</path>
</data1>
<data>
<path>/data/ck9099/data</path>
</data>
</disks>
</storage_configuration>
<path>/data1/clickhous... | 2020-01-03T15:18:12 | 2020-01-03T15:26:45 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,665,204 | 8,491 | den-crane | `--format_csv_delimiter`
```
create table Y(A String, B String) Engine=Memory;
=> cat text.csv
aaaa|bbb
"ffff""x"|'dfdf'ff'''fff
=> cat text.csv |clickhouse-client --format_csv_delimiter='|' --format_csv_allow_single_quotes=0 -q "insert into Y format CSV"
select * from Y;
┌─A──────┬─B──────────────┐
│ ... | 2020-01-03T18:58:59 | 2020-01-03T18:58:59 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,433,604 | 8,217 | vzakaznikov | Changes in this PR have been merged to https://github.com/ClickHouse/ClickHouse/pull/7792. This PR can be closed. | 2020-01-03T01:56:24 | 2020-01-03T01:56:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,522,712 | 7,851 | alexey-milovidov | Fixed in 19.17. | 2020-01-03T09:53:32 | 2020-01-03T09:53:32 | {} | MEMBER |
ClickHouse | ClickHouse | 569,615,018 | 2,628 | nikitamikhaylov | It is not fixed. No Exception is being catched now. https://github.com/ClickHouse/ClickHouse/blob/458f596e187c75c4e77669bfc0831cfe20b3b986/dbms/src/Databases/DatabaseOnDisk.cpp#L145-L155
Will send PR. | 2019-12-30T08:36:10 | 2019-12-30T08:36:10 | {} | MEMBER |
ClickHouse | ClickHouse | 569,716,246 | 8,329 | amosbird | > Then we could check if Identifier is an alias.
What happens after we found out it's an alias? Do we need to error out or use the aliased value instead? If latter is favored, that means one might accidentally have an alias supplied in a dictGet function while he interprets that as a table name. If `dictGet(a, ..)` ... | 2019-12-30T15:53:17 | 2019-12-30T15:53:17 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 569,762,999 | 8,474 | amosbird | It has been already covered in this pr https://github.com/ClickHouse/ClickHouse/pull/7769 . Perhaps it's time to review :) | 2019-12-30T18:53:03 | 2019-12-30T18:53:03 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 569,989,773 | 8,491 | BigBig-joe | > COLUMN_A : "ABC, DE" , COLUMN_B: "EF "G" "
this is what it like in table.
yes, I tried this, but not work. here is a data sample.
> "9AAAAAADW5__A6BA_BARGAINS \\"N\\" DEALZ"
> --escaped-by \\\ --enclosed-by '"'
the thing is , I can't change the CSV file format . that's just what I can use and provided... | 2019-12-31T20:55:52 | 2019-12-31T20:57:31 | {} | NONE |
ClickHouse | ClickHouse | 570,066,619 | 8,495 | ramazanpolat | Closed because of #8496 | 2020-01-01T16:44:34 | 2020-01-01T16:44:34 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,151,821 | 5,075 | siddharth-goel-guavus | The issue is still there | 2020-01-02T09:14:25 | 2020-01-02T09:14:25 | {
"+1": 2
} | NONE |
ClickHouse | ClickHouse | 570,238,479 | 8,498 | leoones | > Make a View.
>
> create view ptx_db.vor_xc_order_item_years_view as select
> pay_nextyear, sku, account, country, asin, company_sk, sumMerge....
> group by ....
>
> and use this vor_xc_order_item_years_view in your dictionary.
Thank you very much for your reply | 2020-01-02T15:19:36 | 2020-01-02T15:19:36 | {} | NONE |
ClickHouse | ClickHouse | 570,463,177 | 8,509 | jencoldeng | > Your SummingMergeTree sums flow column, but AggregatingMergeTree calcs max.
>
> Not really sure what you wanted to achieve here. You can calculate maxState in SummingMergeTree as well.
In [Date] table dataDate201912 , I want to summary the result of dimensions.
For example, when ip="192.168.0.2" and domain="w... | 2020-01-03T05:01:40 | 2020-01-03T05:01:40 | {} | NONE |
ClickHouse | ClickHouse | 569,747,518 | 8,455 | alexey-milovidov | Yes, I have also checked `Vertical` format on my server and there is no regression. Something is strange in CI. | 2019-12-30T17:51:41 | 2019-12-30T17:51:41 | {} | MEMBER |
ClickHouse | ClickHouse | 569,765,766 | 8,455 | alexey-milovidov | I have updated performance test baseline to the current state. | 2019-12-30T19:03:49 | 2019-12-30T19:03:49 | {} | MEMBER |
ClickHouse | ClickHouse | 569,988,719 | 8,491 | den-crane | ```
"ABC, DE" , "EF "G" "
```
This is not CSV.
```
COLUMN_A : "ABC, DE" , COLUMN_B: "EF "G" "
```
This is also not CSV.
You should either change delimiter to not comma or quoting not to double quotes.
--format_csv_delimiter
--format_csv_allow_single_quotes
--format_csv_allow_double_quotes
Try `--escap... | 2019-12-31T20:46:39 | 2019-12-31T20:48:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,361,199 | 8,506 | den-crane | There is no such parameter `max_memory_usage_per_query`
Check your settings
```
SELECT
name,
value
FROM system.settings
WHERE name LIKE 'max%mem%'
┌─name─────────────────────────────┬─value───────┐
│ max_memory_usage │ 50659012608 │
│ max_memory_usage_for_user │ 0 ... | 2020-01-02T21:41:33 | 2020-01-02T21:41:33 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,458,128 | 8,510 | den-crane | check this issue https://github.com/ClickHouse/ClickHouse/issues/8488#issuecomment-570436720 | 2020-01-03T04:27:20 | 2020-01-03T04:27:20 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 569,737,609 | 8,303 | nikitamikhaylov | I wrote a test for bunch updating, but queries with **send_logs_level = debug** don't send logs from *ExternalDictionaries* and this test is Ok only if update thread sleeps for about a second.
Test:
https://paste.yandex-team.ru/902774
grep server logs:
https://paste.yandex-team.ru/902777
dictionary:
https:/... | 2019-12-30T17:13:59 | 2019-12-30T17:16:10 | {} | MEMBER |
ClickHouse | ClickHouse | 569,994,209 | 8,491 | den-crane | I am confused. Can you show whole row sample for
` --escaped-by \\ --enclosed-by '"'`
and
for `--escaped-by \" --enclosed-by '"'` | 2019-12-31T21:34:22 | 2019-12-31T21:34:22 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,122,923 | 8,498 | leoones | > Make a View.
>
> create view ptx_db.vor_xc_order_item_years_view as select
> pay_nextyear, sku, account, country, asin, company_sk, sumMerge....
> group by ....
>
> and use this vor_xc_order_item_years_view in your dictionary.
I created the view and applied it in the dictionary, but the value obtained is... | 2020-01-02T05:44:08 | 2020-01-02T05:44:08 | {} | NONE |
ClickHouse | ClickHouse | 570,457,845 | 8,488 | den-crane | >Absolute delay: 47
pretty big delay, check CH logs at both replicas and network between replicas. | 2020-01-03T04:25:38 | 2020-01-03T04:25:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 570,658,978 | 8,491 | BigBig-joe | Sorry to reply so late as on a leave yesterday.
I think I can use the customize delimiter instead. but, how to use the custom format settings? I tried like this, but not work. does I get something wrong?
`clickhouse-client --format_custom_field_delimiter='||' --query="INSERT INTO default.File_Q_salesorderdetai... | 2020-01-03T18:35:59 | 2020-01-03T18:35:59 | {} | NONE |
ClickHouse | ClickHouse | 592,549,715 | 9,289 | nikitamikhaylov | Conflicts in 20.1
| 2020-02-28T14:57:55 | 2020-02-28T14:57:55 | {} | MEMBER |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.