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 | 795,068,011 | 21,136 | filimonov | > By the way, if performance of the parsing is important inside a browser
Not so important. In browsers typically only a small amount of data can be processed anyway.
So anything reasonably good (better supported by browser natively) is ok.
1) SSE: https://caniuse.com/eventsource
2) base64: https://caniuse... | 2021-03-10T08:18:36 | 2021-03-10T08:22:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,726,173 | 20,213 | nikitamikhaylov | @MaxJPowers The only issue I know with timeout is https://github.com/ClickHouse/ClickHouse/pull/14105 and it is already in 20.8. Please, give us a reproduction.. | 2021-03-10T16:51:22 | 2021-03-10T16:51:22 | {} | MEMBER |
ClickHouse | ClickHouse | 796,347,409 | 16,669 | sliontc | I'v updated to the latest version. | 2021-03-11T01:16:58 | 2021-03-11T01:16:58 | {} | NONE |
ClickHouse | ClickHouse | 794,763,482 | 21,136 | alexey-milovidov | clickhouse-cpp will bring C++ standard library with vector, string, etc. and the JS version will be huge (maybe in order of megabytes), also it will require conversion from bytes to JavaScript strings, that will spend some time... Moreover I cannot say that clickhouse-cpp is performant. | 2021-03-10T02:33:26 | 2021-03-10T02:33:41 | {} | MEMBER |
ClickHouse | ClickHouse | 795,042,730 | 21,468 | alesapin | > @alesapin
> I don't use secure zookeeper connection.
> The value is `Len error 1913604`.
> number of logs bytes
> 8314 1913604
> 63 1930672
Ok, it's not a bug or race. Just for some reason, clickhouse tries to send more than 1MB to ZooKeeper. Maybe you can provide your table schema and parts count in each ... | 2021-03-10T07:53:51 | 2021-03-10T07:54:30 | {} | MEMBER |
ClickHouse | ClickHouse | 795,051,299 | 21,556 | OmarBazaraa | > Can you show results of running query on new version? (GB scanned)
```
100 rows in set. Elapsed: 0.075 sec. Processed 11.63 million rows, 369.61 MB (154.78 million rows/s., 4.92 GB/s.)
``` | 2021-03-10T08:02:42 | 2021-03-10T08:02:42 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,175,943 | 19,766 | ghost | An exceptional case is:
data:
| timestamp | id | page |
| -- | -- | ------ |
| 1 | 1 | home |
| 2 | 1 | home |
| 3 | 1 | giftbox |
| 4 | 1 | wish |
| 1 | 2 | home |
| 2 | 2 | giftbox |
| 3 | 2 | wish |
1. Get first pages.
```
select id, sequenceNextNode(0)(timestamp, page) from table group by id
... | 2021-03-10T09:55:44 | 2021-03-10T10:21:11 | {} | NONE |
ClickHouse | ClickHouse | 795,387,259 | 21,567 | yakout | Thanks Issue resolved.
when I run
`sudo service clickhouse-server status`
I found the correct path of the config file (I was editing the wrong one)
```
└─274033 /usr/bin/clickhouse-server --config=/etc/clickhouse-server/config.xml --pid-file=/run/clickhouse-server/clickhouse-server.pid
``` | 2021-03-10T13:07:25 | 2021-03-10T13:07:25 | {} | NONE |
ClickHouse | ClickHouse | 795,814,242 | 21,456 | nikitamikhaylov | Yandex synchronization check failure - inconsistent behavior of a tool to sync ClickHouse into Arcadia | 2021-03-10T17:49:52 | 2021-03-10T17:49:52 | {} | MEMBER |
ClickHouse | ClickHouse | 796,031,401 | 21,607 | siradjev | #19320 has similar request so they probably can be implemented together. | 2021-03-10T20:18:27 | 2021-03-10T20:18:27 | {} | NONE |
ClickHouse | ClickHouse | 794,750,478 | 5,582 | alexey-milovidov | You mean - avg cannot be calculated correctly within the limitations of GraphiteMergeTree?
Should we close with won't fix? | 2021-03-10T02:19:04 | 2021-03-10T02:19:04 | {} | MEMBER |
ClickHouse | ClickHouse | 794,823,498 | 21,567 | ngsiolei | Please have a look on preprocessed config `/var/lib/clickhouse/preprocessed_configs/config.xml`, which was generated by all config files, see if `max_concurrent_queries` was overridden.
Ref: [How to check whether Clickhouse server-settings is really applied?](https://stackoverflow.com/questions/65950979/how-to-check... | 2021-03-10T03:46:02 | 2021-03-10T03:46:02 | {
"+1": 3
} | NONE |
ClickHouse | ClickHouse | 795,300,107 | 20,213 | MaxJPowers | @nikitamikhaylov I've installed 20.8.12.2 as that's the latest version available on centos 7.6, is that fine? | 2021-03-10T11:45:49 | 2021-03-10T11:45:49 | {} | NONE |
ClickHouse | ClickHouse | 795,556,654 | 21,591 | filimonov | Or use `<listen_try>1</listen_try>` to ignore the exception. | 2021-03-10T15:00:17 | 2021-03-10T15:00:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,420,847 | 16,798 | sunbc996 | https://blog.csdn.net/qq_27639777/article/details/111005838
// 创建分布式表
CREATE TABLE stone.bitmap_test ON CLUSTER cluster_name (
ds Int64,
user String,
roaring_bitmap AggregateFunction(groupBitmap, UInt32)
MATERIALIZED base64Decode(user)
)
ENGINE = ReplicatedMergeTree('xxx', '{replica}')
P... | 2021-03-11T03:44:29 | 2021-03-11T03:44:29 | {} | NONE |
ClickHouse | ClickHouse | 796,496,595 | 21,586 | cyber-moon | Alright I get the point. Thanks for clarifying! | 2021-03-11T06:31:06 | 2021-03-11T06:31:06 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,512,563 | 12,609 | ababutski | Just want to share my pain about this here. I managed to properly set up clickhouse kafka engine with event hub kafka endpoint.
1. First of all add clickhouse instance ip to event hub namespace firewall. Better test connectivity with kafkacat before moving on.
2. Second, setting up kafka engine config. My config
```... | 2021-03-09T21:58:52 | 2021-03-09T21:59:24 | {
"+1": 1
} | NONE |
ClickHouse | ClickHouse | 794,678,237 | 5,582 | den-crane | 21.3.1.6159
```
SET optimize_on_insert = 0;
TRUNCATE TABLE graphite_one_insert
INSERT INTO graphite_one_insert VALUES ('12', 1, 1615330649, toDate(1615330648), 0), ('13', 3, 1615330650, toDate(1615330648), 0)
INSERT INTO graphite_one_insert VALUES ('12', 3, 1615330650, toDate(1615330648), 0), ('13', 1, 16153... | 2021-03-10T00:56:41 | 2021-03-10T00:56:56 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,064,179 | 17,236 | egorbl4 | @den-crane But does the way ( to grant priveleges to user created by users.xml ) exist ? | 2021-03-10T08:15:11 | 2021-03-10T08:15:37 | {
"eyes": 7
} | NONE |
ClickHouse | ClickHouse | 795,158,195 | 20,347 | gyuton | @nikitamikhaylov Hi! Any additions yet? | 2021-03-10T09:39:36 | 2021-03-10T09:39:36 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,616,888 | 21,594 | Maximus43 | > Do you use some memory limitation on LXC/cgroup level?
Yes, container has limit 131072 Mbytes for RAM
> Do you check `free` in container? Or outside?
It was inside the container.
Here is "free" on host machine (outside):
``` total used free shared buff/cache available
... | 2021-03-10T15:36:45 | 2021-03-10T15:37:49 | {} | NONE |
ClickHouse | ClickHouse | 794,942,804 | 21,577 | ggchangan | From the above picture, you can see that the performance in 20.12.7.3 is worse than 20.4.4.18. Let make troubleshoot as an example, the <2000.ms ratio of 20.12.7.3 is worse than 20.4.4.18:
- 20.4.4.18
catalog | round | <=100.0ms | <=100.0ms ratio | >100.0ms <=500.0ms | <500.0ms ratio | >500.0ms <=1000.0ms | <1000.0... | 2021-03-10T06:10:40 | 2021-03-10T07:09:27 | {} | NONE |
ClickHouse | ClickHouse | 796,270,690 | 21,611 | chenziliang | CH 20.11.1.
We are not sure if there are mem leak actually. Since we are only doing ingestion, it seems not quite making sense CH runs out of memory and crash (more reasonable way may be rejecting the ingest request instead?).
I will try to collect the usage graph.
Thanks a lot ! | 2021-03-10T23:08:32 | 2021-03-10T23:08:32 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,937,418 | 21,536 | filimonov | https://github.com/ClickHouse/ClickHouse/issues/14735 | 2021-03-10T06:04:46 | 2021-03-10T06:04:46 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,014,578 | 21,549 | alesapin | No idea what's wrong with replace partition test:
```
node5/logs/clickhouse-server.log
224:2021.03.09 17:58:50.637165 [ 11 ] {a96a1319-67e5-4b75-bad2-227e7803bb90} <Debug> default.test_table (4f0c7ad5-9a5e-4f77-bb62-1ce4a0a0dc5c): Cloning part /var/lib/clickhouse/store/1e0/1e02db04-68ef-4c80-b2a6-552070f088a9/201706... | 2021-03-10T07:25:02 | 2021-03-10T07:25:02 | {} | MEMBER |
ClickHouse | ClickHouse | 796,216,389 | 21,611 | den-crane | Clickhouse version?
Why do you think that you are experiencing a mem. leak?
Do you have some graph to see memory usage in time?
If CH uses 56GB or 560GB it does not mean that it's a memory leak.
| 2021-03-10T22:16:29 | 2021-03-10T22:16:29 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,259,639 | 21,612 | nikitamikhaylov | Locally this catches a data race (test_grpc_protocol with TSAN)
```
m ==================
18 ^[[36mnode_1 |^[[0m WARNING: ThreadSanitizer: data race (pid=7)
19 ^[[36mnode_1 |^[[0m Read of size 8 at 0x7b9000065bb0 by thread T112:
20 ^[[36mnode_1 |^[[0m #0 google::protobuf::internal::ArenaStringPtr::Get()... | 2021-03-10T22:50:15 | 2021-03-10T22:51:44 | {} | MEMBER |
ClickHouse | ClickHouse | 794,772,268 | 10,519 | alexey-milovidov | @pitwegner Mac OS build of clickhouse-client does not support format Parquet.
You can run Linux build in Docker or on Linux VM. | 2021-03-10T02:43:46 | 2021-03-10T02:43:46 | {} | MEMBER |
ClickHouse | ClickHouse | 794,779,109 | 21,538 | ngsiolei | For Kafka engine authentication, server-level settings in ClickHouse config file is the way to go now. | 2021-03-10T02:51:52 | 2021-03-10T02:51:52 | {} | NONE |
ClickHouse | ClickHouse | 794,937,202 | 21,536 | filimonov | https://github.com/ClickHouse/ClickHouse/issues/18091 | 2021-03-10T06:04:31 | 2021-03-10T06:04:31 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,981,342 | 21,575 | joakapp | Thanks a lot | 2021-03-10T06:51:30 | 2021-03-10T06:51:30 | {} | NONE |
ClickHouse | ClickHouse | 795,289,795 | 20,213 | MaxJPowers | @nikitamikhaylov So any advice? I'm still dealing with the same issue. | 2021-03-10T11:36:13 | 2021-03-10T11:36:13 | {} | NONE |
ClickHouse | ClickHouse | 795,807,195 | 17,413 | UnamedRus | Simpler example:
```
CREATE TABLE pushdown_test (key UInt32) ENGINE=MergeTree() ORDER BY key;
INSERT INTO pushdown_test SELECT * FROM numbers(100000);
CREATE VIEW v_pushdown_test AS SELECT * FROM pushdown_test;
CREATE TABLE m_v_pushdown_test AS pushdown_test ENGINE=Merge(default,'^v_pushdown_test');
SELECT * ... | 2021-03-10T17:44:54 | 2021-03-10T18:27:04 | {
"eyes": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,336,201 | 21,611 | den-crane | You provided zero information. No logs. No graphs. No queries. No table DDLs.
Clickhouse usually does not crash. And has memory tracker to prevent OOM. CH sets memory limit for itself as 0.9*RAM and fails queries with an error " Memory limit (total) exceeded: would use ... GiB".
You can change .9. /etc/clickhouse... | 2021-03-11T00:56:12 | 2021-03-11T00:56:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,495,854 | 21,617 | alexey-milovidov | The message is correct. | 2021-03-11T06:29:33 | 2021-03-11T06:29:33 | {} | MEMBER |
ClickHouse | ClickHouse | 796,512,839 | 21,136 | traceon | > Can you run wasm in curl? In Python http client? In perl http client? In Ruby http client? Etc.
Well, you can't decode binary or BASE64 in curl too, etc. I guess, we are solving different problems here...
Regarding heavy calculations on client side: one example would be rendering data as plots. | 2021-03-11T07:02:45 | 2021-03-11T07:02:45 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,477,044 | 19,575 | dankondr | > @dankondr Are you going to continue working on this PR? Do you need some help maybe?
Hi, yes, it is my coursework, I was just lacking the time to make it, sorry, I will do it in the near future. | 2021-03-09T21:27:20 | 2021-03-09T21:27:20 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,781,754 | 21,538 | Lipeng522 | OK，I know，thank you | 2021-03-10T02:54:59 | 2021-03-10T02:54:59 | {} | NONE |
ClickHouse | ClickHouse | 794,931,052 | 21,526 | azat | >Testflows check — failed: 754, passed: 2692, other: 25
Looks like it is broken/flaky? | 2021-03-10T05:57:42 | 2021-03-10T05:57:42 | {} | MEMBER |
ClickHouse | ClickHouse | 794,933,154 | 21,561 | filimonov | Open commit in the ClickHouse master branch which does some changes in code with a green mark near it.
Click on the green mark, open 'Special build check' it has a binary and a build log.
For example:
https://clickhouse-builds.s3.yandex.net/0/5a6cc876e4947f4b883faab9c15a1645ef413242/clickhouse_special_build_check/... | 2021-03-10T06:00:10 | 2021-03-10T06:00:10 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,946,332 | 21,514 | filimonov | What is happineng in the server logs at for the same query? How much time does the query take? Does retry help? | 2021-03-10T06:14:35 | 2021-03-10T06:15:16 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,038,053 | 235 | develar | Is it possible to build macOS for release branches? https://github.com/ClickHouse/ClickHouse/tree/21.2
> ClickHouse special build check — Not ran for backport or release branches
Build from master is not good to use clickhouse to test/develop app on macOS since better to use the same version as in production. A... | 2021-03-10T07:48:56 | 2021-03-10T07:50:10 | {} | NONE |
ClickHouse | ClickHouse | 795,081,874 | 235 | develar | If someone cannot build stable 21.2 clickhouse on macOS — see https://github.com/ClickHouse/ClickHouse/issues/21128#issuecomment-784971469 | 2021-03-10T08:31:21 | 2021-03-10T08:31:29 | {} | NONE |
ClickHouse | ClickHouse | 795,282,196 | 19,370 | hatrena | Hi,
I wanted to know if you could have any updates on this? | 2021-03-10T11:29:24 | 2021-03-10T11:29:24 | {} | NONE |
ClickHouse | ClickHouse | 796,356,651 | 21,533 | alexey-milovidov | TODO fix arithmetic in `swap` method. | 2021-03-11T01:34:48 | 2021-03-11T01:34:48 | {} | MEMBER |
ClickHouse | ClickHouse | 796,362,473 | 21,611 | chenziliang | > You provided zero information. No logs. No graphs. No queries. No table DDLs.
> Clickhouse usually does not crash. And has memory tracker to prevent OOM. CH sets memory limit for itself as 0.9*RAM and fails queries with an error " Memory limit (total) exceeded: would use ... GiB".
>
> You can change .9. /etc/clic... | 2021-03-11T01:46:28 | 2021-03-11T01:46:28 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,363,466 | 21,608 | alexey-milovidov | Only need to find some picture... | 2021-03-11T01:48:24 | 2021-03-11T01:48:24 | {} | MEMBER |
ClickHouse | ClickHouse | 796,424,741 | 21,552 | den-crane | ```
select [1,2,3],[1.1,2.2,3.3],['a','b','c'] format CSV
"[1,2,3]","[1.1,2.2,3.3]","['a','b','c']"
```
You sent: `"[1]","[2]","[a,b,c]"` instead of `"[1,2,3]","[1.1,2.2,3.3]","['a','b','c']"` | 2021-03-11T03:53:07 | 2021-03-11T03:53:07 | {
"+1": 2
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,466,808 | 21,136 | filimonov | > it is supported out of the box in all major browsers
Can you run wasm in curl? In Python http client? In perl http client? In Ruby http client? Etc.
> lots of data can be processed
There is no such need normally. Browser is a 'presentation layer'. Heavy calculations happen on a database layer and few rows of... | 2021-03-11T05:29:54 | 2021-03-11T05:45:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,476,164 | 21,583 | filimonov | Also relates to https://github.com/ClickHouse/ClickHouse/issues/11749 | 2021-03-11T05:50:27 | 2021-03-11T05:50:27 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,182,769 | 21,475 | CurtizJ | Continued in #21562. | 2021-03-09T17:15:15 | 2021-03-09T17:15:15 | {} | MEMBER |
ClickHouse | ClickHouse | 794,758,513 | 21,560 | alexey-milovidov | Actually it's useful only for dictionaries created from XML and only if you did not create the Dictionary database:
```
CREATE DATABASE dictionaries ENGINE = Dictionary
```
But may be still worth doing. | 2021-03-10T02:28:00 | 2021-03-10T02:28:00 | {} | MEMBER |
ClickHouse | ClickHouse | 794,950,670 | 14,582 | cyberhuman | The bug is in `StorageMergeTree::movePartitionToTable`.
It does this:
```
dest_table_storage->renameTempPartAndReplace(part, &increment, &transaction, lock);
```
but AFAIU it must use `&dest_table_storage->increment` instead. | 2021-03-10T06:19:22 | 2021-03-10T06:19:22 | {} | NONE |
ClickHouse | ClickHouse | 794,957,609 | 21,136 | traceon | Actually, only the stream parsing is enough, and it is not hard to implement in a standalone C++ code. I also mean native wasm bytecode, not JS as a target. | 2021-03-10T06:26:55 | 2021-03-10T06:26:55 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,566,775 | 21,586 | filimonov | For me it's EXPECTED behaviour :) You asked docker to start `bash` INSEAD of `clickhouse-server`, and so he did.
So why do you expect it will be working? Do some other databases docker images work like that? | 2021-03-10T15:05:49 | 2021-03-10T15:06:00 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,143,921 | 21,609 | UnamedRus | > However, in 21.1.2, the replica path template has a hardcoded table name after creation:
This was done on purpose, because when people tried to RENAME those tables it was breaking replication because {database} and {table} macroses have been updated with new table name.
https://github.com/ClickHouse/ClickHouse/is... | 2021-03-10T21:27:29 | 2021-03-10T21:30:01 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,207,452 | 21,609 | den-crane | @alvarotuso it's impossible to rename key in ZK. ZK does not have such feature.
| 2021-03-10T22:10:06 | 2021-03-10T22:10:06 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,395,351 | 21,591 | dubin555 | Solved. if anyone had the same issue, please try to modify the listen.xml(full path:/etc/clickhouse-server/config.d/listen.xml), do not modify the listen_host in the config.xml. | 2021-03-11T02:51:01 | 2021-03-11T02:51:01 | {
"+1": 4
} | NONE |
ClickHouse | ClickHouse | 796,420,127 | 21,614 | den-crane | it's taken. It does not matter that they are not showing in the " CREATE TABLE "
Check ` select zookeeper_path, table from system.replicas where table = 'test'`; | 2021-03-11T03:42:50 | 2021-03-11T03:42:50 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,479,156 | 21,616 | amosbird | Converting from DateTime to Date is not time-zone neutral. | 2021-03-11T05:56:54 | 2021-03-11T05:56:54 | {} | COLLABORATOR |
ClickHouse | ClickHouse | 794,737,927 | 15,090 | shiyi23 | > Yes.
thanks your reply | 2021-03-10T02:05:00 | 2021-03-10T02:05:00 | {} | NONE |
ClickHouse | ClickHouse | 794,857,153 | 21,575 | den-crane | --format_csv_delimiter=$'\x03' | 2021-03-10T04:28:58 | 2021-03-10T04:28:58 | {
"+1": 1
} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,923,719 | 21,577 | filimonov | Please share samples of problematic queries & schema. | 2021-03-10T05:49:08 | 2021-03-10T05:49:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,002,390 | 21,577 | ggchangan | @filimonov the performance test is based on case set and there're so many query and schema. Is there any setting changes from 20.4.4.18 to 20.12.7.3 which lead throughput of ClickHouse 20.12.7.3 is lower than ClickHouse 20.4.4.18? | 2021-03-10T07:12:55 | 2021-03-10T07:12:55 | {} | NONE |
ClickHouse | ClickHouse | 795,015,295 | 21,553 | alesapin | ```test_dictionaries_update_and_reload/test.py::test_reload_after_loading```
not related to changes. | 2021-03-10T07:25:41 | 2021-03-10T07:25:41 | {} | MEMBER |
ClickHouse | ClickHouse | 795,130,325 | 19,766 | ghost | @vdimir
I found an exceptional case.
So I will fix the core part of this aggregation function.
Sorry, I will make change it as soon as possible. | 2021-03-10T09:14:51 | 2021-03-10T09:17:01 | {} | NONE |
ClickHouse | ClickHouse | 795,163,931 | 19,766 | vdimir | @achimbab
> I found an exceptional case.
Do you mean that there are cases when algorithm provides wrong result? Could you give an example?
> So I will fix the core part of this aggregation function.
> Sorry, I will make change it as soon as possible.
Sure, it's up to you.
| 2021-03-10T09:44:42 | 2021-03-10T09:44:55 | {} | MEMBER |
ClickHouse | ClickHouse | 795,559,550 | 21,594 | filimonov | Hm. We have one more report with LXC recently. See https://github.com/ClickHouse/ClickHouse/issues/20608.
Do you use some memory limitation on LXC/cgroup level?
Do you check `free` in container? Or outside? | 2021-03-10T15:02:16 | 2021-03-10T15:03:24 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,855,858 | 21,570 | sevirov | @vdimir Please, review this PR. | 2021-03-10T18:18:07 | 2021-03-10T18:18:07 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,000,786 | 21,561 | janantos | Thanks for hints @filimonov, after little bit digging I can confirm I can build ClickHouse in FreeBSD pot jail. (clang11, freebsd 12.2)
```
cmake --debug-trycompile --verbose=1 -DCMAKE_VERBOSE_MAKEFILE=1 -LA -DCMAKE_BUILD_TYPE=Release -DENABLE_CHECK_HEAVY_BUILDS=0 -DCMAKE_TOOLCHAIN_FILE=../cmake/freebsd/toolchain-x... | 2021-03-10T19:58:08 | 2021-03-10T20:18:33 | {} | NONE |
ClickHouse | ClickHouse | 796,532,410 | 21,577 | filimonov | > Is there any setting changes from 20.4.4.18 to 20.12.7.3 which lead throughput of ClickHouse 20.12.7.3 is lower than ClickHouse 20.4.4.18?
ClickHouse development pace is very high. There are thousands of changes in between.
Here is the (not comprehensive) list of new settings / settings which change their def... | 2021-03-11T07:37:16 | 2021-03-11T07:38:08 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,963,247 | 21,136 | filimonov | > Actually, only the stream parsing is enough, and it is not hard to implement in a standalone C++ code. I also mean native wasm bytecode, not JS as a target.
It's interesting as a research project but sounds too exotic / out of the scope for that particular issue.
We need a clean & standard way to talk clickhous... | 2021-03-10T06:32:59 | 2021-03-10T06:34:04 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,015,783 | 21,548 | alesapin | `test_dictionaries_dependency_xml/test.py::test_get_data` not related to changes. | 2021-03-10T07:26:10 | 2021-03-10T07:26:21 | {} | MEMBER |
ClickHouse | ClickHouse | 794,960,839 | 21,136 | filimonov | > But server sent events don't support binary data: https://stackoverflow.com/questions/33529685/using-sse-to-send-image-data
> So, we cannot use it for arbitrary formats. It can find limited use for TSV/CSV/JSONEachRow (in fact, even these formats are binary in ClickHouse). But it is also unclear how does it deal wit... | 2021-03-10T06:30:23 | 2021-03-10T07:56:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,555,895 | 21,591 | filimonov | If you have ipv6 disabled you can not use IPv6 addresses `::` and `::1` in `<listen_host>` | 2021-03-10T14:59:51 | 2021-03-10T14:59:51 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,206,866 | 21,609 | UnamedRus | > Does that mean that when renaming, the replicas are kept in the old path? I wonder if renaming the path would have worked as well
Yes. (in new versions of clickhouse) | 2021-03-10T22:09:43 | 2021-03-10T22:09:43 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,333,010 | 16,669 | den-crane | @sliontc no solution. It's a bug. | 2021-03-11T00:50:38 | 2021-03-11T00:50:38 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,496,006 | 20,182 | alexey-milovidov | Changelog is being prepared here: #21618 | 2021-03-11T06:29:52 | 2021-03-11T06:29:52 | {} | MEMBER |
ClickHouse | ClickHouse | 794,960,127 | 21,514 | cyberhuman | > What is happineng in the server logs at for the same query?
For example:
```
2021.03.10 05:47:21.140195 [ 10627 ] {f3e8f241-d02c-49cf-a6b7-17ed322b013e} <Error> executeQuery: Code: 241, e.displayText() = DB::Exception: Memory limit (total) exceeded: would use 504.21 GiB (attempt to allocate chunk of 2147486608 b... | 2021-03-10T06:29:37 | 2021-03-10T06:29:37 | {} | NONE |
ClickHouse | ClickHouse | 795,102,629 | 5,582 | Felixoid | The `optimize_on_insert=0` is not set, so CH applies merges on the fly. And the behavior now (since #16954) very close to the whisper: [last write wins](https://grafana.com/blog/2016/03/03/25-graphite-grafana-and-statsd-gotchas/#graphite.quantization)
Only the case when it's not: values for the same point come in on... | 2021-03-10T08:50:20 | 2021-03-10T09:16:52 | {} | MEMBER |
ClickHouse | ClickHouse | 795,192,259 | 21,136 | traceon | > It's interesting as a research project but sounds too exotic / out of the scope for that particular issue.
We need a clean & standard way to talk clickhouse via HTTP (from a browser and other standard HTTP clients / without installing some strange stuff).
I understand your concerns, however, the thing with WASM i... | 2021-03-10T10:10:23 | 2021-03-10T10:10:23 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,183,890 | 21,609 | alvarotuso | Hey @UnamedRus thanks for the reply
Does that mean that when renaming, the replicas are kept in the old path? I wonder if renaming the path would have worked as well
Anyways, thanks for the explanation! we'll look for a workaround | 2021-03-10T21:52:41 | 2021-03-10T21:52:41 | {} | NONE |
ClickHouse | ClickHouse | 796,404,787 | 2,568 | ManchesterUnited16 | how to drop all replicas | 2021-03-11T03:11:40 | 2021-03-11T03:11:40 | {} | NONE |
ClickHouse | ClickHouse | 794,827,614 | 21,468 | ghost | @alesapin
I don't use secure zookeeper connection.
The value is `Len error 1913604`.
| number of logs | bytes |
| ----- | ---- |
| 8314 | 1913604 |
| 63 | 1930672 |
| 2021-03-10T03:51:20 | 2021-03-10T04:07:26 | {} | NONE |
ClickHouse | ClickHouse | 795,633,476 | 21,594 | Maximus43 | I'm slightly confused with startup message:
```
2021.03.10 15:17:19.060659 [ 1272 ] {} <Information> Application: Available RAM: 503.85 GiB; physical cores: 12; logical cores: 24.
2021.03.10 15:17:19.060687 [ 1272 ] {} <Information> Application: Ready for connections.
```
It shows amount of RAM on the host machi... | 2021-03-10T15:47:46 | 2021-03-10T15:49:20 | {} | NONE |
ClickHouse | ClickHouse | 796,037,032 | 21,599 | filimonov | 1) Motivation is not 100% clear. The only clear improvement i see is ability to source the entrypoint w/o executing it. So you can for example create some adjusted entrypoint which will source the main one. It's nice and common practice.
2) Backward incompatible? before you could start for example clickhouse-client f... | 2021-03-10T20:21:11 | 2021-03-10T20:23:12 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,581,429 | 5,582 | Felixoid | Ok, I found a minimal reproducible case, AFAIS
```
<graphite_rollup_example>
<default>
<function>avg</function>
<retention>
<age>0</age>
<precision>4</precision>
</retention>
</default>
</graphite_rollup_example>
CREATE TABLE default.graphite_one_... | 2021-03-09T23:07:53 | 2021-03-09T23:20:00 | {} | MEMBER |
ClickHouse | ClickHouse | 794,937,813 | 21,578 | filimonov | https://github.com/ClickHouse/ClickHouse/issues/14735 | 2021-03-10T06:05:13 | 2021-03-10T06:05:13 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 795,156,898 | 21,329 | vdimir | We already have some functions that overlaps with some implemented in this PR.
<details>
<summary>See IPAddressDictionary.cpp and FunctionsCoding.h
</summary>
https://github.com/ClickHouse/ClickHouse/blob/e27715e55e471dc4d7f825eb4097b468a9a4a451/src/Dictionaries/IPAddressDictionary.cpp#L83-L243
https://git... | 2021-03-10T09:38:30 | 2021-03-10T09:38:48 | {} | MEMBER |
ClickHouse | ClickHouse | 795,775,531 | 21,552 | aol-59 | > What will be if you disable `input_format_parallel_parsing` setting?
same error :( | 2021-03-10T17:24:08 | 2021-03-10T17:24:08 | {} | NONE |
ClickHouse | ClickHouse | 796,123,690 | 21,611 | chenziliang | One example of the fatal log is
```
2021.03.10 21:39:19.745189 [ 200054 ] {} <Error> auto DB::IBackgroundJobExecutor::jobExecutingTask()::(anonymous class)::operator()() const: Code: 241, e.displayText() = DB::Exception: Memory limit (total) exceeded: would use 56.59 GiB (attempt to allocate chunk of 4412234 bytes... | 2021-03-10T21:15:48 | 2021-03-10T21:15:48 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 796,326,058 | 16,669 | sliontc | Hey guys, I got the same issue, any solution? | 2021-03-11T00:38:43 | 2021-03-11T00:38:43 | {} | NONE |
ClickHouse | ClickHouse | 794,674,974 | 5,582 | den-crane | ```
SET optimize_on_insert = 0;
INSERT INTO default.graphite_one_insert VALUES ('12', 1, 1615330649, toDate(1615330648), 0), ('12', 3, 1615330650, toDate(1615330648), 0) ;
INSERT INTO graphite_one_insert VALUES ('13', 3, 1615330650, toDate(1615330648), 0);
INSERT INTO graphite_one_insert VALUES ('13', 1, 161... | 2021-03-10T00:52:49 | 2021-03-10T00:52:49 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,926,862 | 21,574 | filimonov | I guess we need to introduce smth like 'maintenance' (maybe also 'failover') mode for a node. | 2021-03-10T05:52:47 | 2021-03-10T05:52:47 | {} | CONTRIBUTOR |
ClickHouse | ClickHouse | 794,929,554 | 21,529 | azat | Rebased to fix conflicts in `tests/queries/skip_list.json` (previous `HEAD` was c0a02c9ce9833362485638c24007318923eca09c) | 2021-03-10T05:55:58 | 2021-03-10T05:56:41 | {} | MEMBER |
ClickHouse | ClickHouse | 796,407,744 | 21,608 | alexey-milovidov | Maybe add some examples? | 2021-03-11T03:17:46 | 2021-03-11T03:17:46 | {} | MEMBER |
ClickHouse | ClickHouse | 800,016,192 | 20,387 | alesapin | `test_query_deduplication` -- I don't see how it's related to changes. Looks like a flaky test because failed only with address sanitizer. | 2021-03-16T07:14:58 | 2021-03-16T07:14:58 | {} | MEMBER |
ClickHouse | ClickHouse | 800,254,873 | 20,179 | alexey-milovidov | DiskS3 is experimental and it won't be considered production ready until it will use O(1) memory.
I have no idea why someone is advertising this feature. | 2021-03-16T13:24:34 | 2021-03-16T13:24:34 | {} | MEMBER |
ClickHouse | ClickHouse | 800,265,731 | 21,782 | flyingmutant | How can you determine if the date is correct or not? Isn't it the job of `toDate`? | 2021-03-16T13:38:41 | 2021-03-16T13:38:41 | {} | NONE |
ClickHouse | ClickHouse | 800,288,193 | 21,782 | proggga | Thanks for the help, but you said:
> My opinion: CH developers decided that it's a responsibility of a client for the sake of performance.
this is your opinion, not theirs, not either is community opinion
if you sure, that CHdev decided that, please send a link to the official documentation | 2021-03-16T14:08:34 | 2021-03-16T14:08:34 | {} | NONE |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.