flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "YUJIANBO (Jira)" <j...@apache.org>
Subject [jira] [Updated] (FLINK-22806) The checkpoint state of FLINKSQL is getting bigger and bigger
Date Mon, 31 May 2021 03:26:00 GMT

     [ https://issues.apache.org/jira/browse/FLINK-22806?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

YUJIANBO updated FLINK-22806:
-----------------------------
    Description: 
I have added the parameter :
   1、state.idle.duration: 3600,  
   2、state.checkpoints.num-retained:3
   3、sql:select count(1), LISTAGG(concat(m,n)) from tabeA group a,b

I set the checkpoint state for one hour, but the size of the folder directory /checkpoint/shared
is still growing.  I observed it for two days and guessed that there was expired data in the
 /checkpoint/shared folder that had not been cleared?

What else can limit the growth of state?

  was:
I have added the parameter :
   1、state.idle.duration: 3600,  
   2、state.checkpoints.num-retained:3
   3、sql:select count(*), LISTAGG(concat(m,n)) from tabeA group a,b

I set the checkpoint state for one hour, but the size of the folder directory /checkpoint/shared
is still growing.  I observed it for two days and guessed that there was expired data in the
 /checkpoint/shared folder that had not been cleared?

What else can limit the growth of state?


> The checkpoint state of FLINKSQL is getting bigger and bigger
> -------------------------------------------------------------
>
>                 Key: FLINK-22806
>                 URL: https://issues.apache.org/jira/browse/FLINK-22806
>             Project: Flink
>          Issue Type: Bug
>          Components: Runtime / Checkpointing, Runtime / State Backends
>    Affects Versions: 1.12.0
>            Reporter: YUJIANBO
>            Priority: Major
>
> I have added the parameter :
>    1、state.idle.duration: 3600,  
>    2、state.checkpoints.num-retained:3
>    3、sql:select count(1), LISTAGG(concat(m,n)) from tabeA group a,b
> I set the checkpoint state for one hour, but the size of the folder directory /checkpoint/shared
is still growing.  I observed it for two days and guessed that there was expired data in the
 /checkpoint/shared folder that had not been cleared?
> What else can limit the growth of state?



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Mime
View raw message