[
https://issues.apache.org/jira/browse/IMPALA-12303?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
daicheng updated IMPALA-12303:
------------------------------
Description:
Environment
AlibabaCloud centos7
machine: 32core, 64G
A single sql mem_limit 3G
Queue max_resources: 61440 mb
I submit 10 sql to impala with impala-shell at once, I expect those 10 SQL
could run concurrently, but some of them just Queued and waitting for running
lilke this:
!image-2023-07-21-11-49-20-998.png|width=971,height=501!
some sql waitting for running too long time, almost 25s.
!image-2023-07-21-11-52-25-880.png!
and this meche has enough resources to run those sql at one time:
!image-2023-07-21-11-55-31-828.png|width=961,height=186!
The time difference between queued and admission completes is too high, why?
was:
Environment
AlibabaCloud centos7
machine: 32core, 64G
A single sql mem_limit 3G
Queue max_resources: 61440 mb
I submit 10 sql to impala with impala-shell at once, I expect those 10 SQL
could run concurrently, but some of them just Queued and waitting for running
lilke this:
!image-2023-07-21-11-49-20-998.png!
some sql waitting for running too long time, almost 25s.
!image-2023-07-21-11-52-25-880.png!
and this meche has enough resources to run those sql at one time:
!image-2023-07-21-11-55-31-828.png|width=961,height=186!
The time difference between queued and admission completes is too high, why?
> the Time between the query being queued and the admission completes is high
> ---------------------------------------------------------------------------
>
> Key: IMPALA-12303
> URL: https://issues.apache.org/jira/browse/IMPALA-12303
> Project: IMPALA
> Issue Type: Bug
> Affects Versions: Impala 4.1.1
> Reporter: daicheng
> Priority: Major
> Attachments: image-2023-07-21-11-49-20-998.png,
> image-2023-07-21-11-52-25-880.png, image-2023-07-21-11-55-31-828.png
>
>
> Environment
> AlibabaCloud centos7
> machine: 32core, 64G
> A single sql mem_limit 3G
> Queue max_resources: 61440 mb
> I submit 10 sql to impala with impala-shell at once, I expect those 10 SQL
> could run concurrently, but some of them just Queued and waitting for
> running lilke this:
> !image-2023-07-21-11-49-20-998.png|width=971,height=501!
> some sql waitting for running too long time, almost 25s.
>
> !image-2023-07-21-11-52-25-880.png!
> and this meche has enough resources to run those sql at one time:
> !image-2023-07-21-11-55-31-828.png|width=961,height=186!
> The time difference between queued and admission completes is too high, why?
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
