[
https://issues.apache.org/jira/browse/HDDS-5126?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jackson Yao updated HDDS-5126:
------------------------------
Description:
in my test environment, 400000 containers exist. when bootstrap a new recon,
every container will be checked and added to recondb.
but , for now , recon check all the containers in a container report one by one
, and each check will take a rpc call to scm. this is too slow and in my test
environment , it leads to recon oom, because too many containers to be consumed
are waitting in the message queue .
this Jira will use a batch method to handle this problem.
> Recon should check new containers of a container repot with batch
> -----------------------------------------------------------------
>
> Key: HDDS-5126
> URL: https://issues.apache.org/jira/browse/HDDS-5126
> Project: Apache Ozone
> Issue Type: Bug
> Reporter: Jackson Yao
> Assignee: Jackson Yao
> Priority: Major
>
> in my test environment, 400000 containers exist. when bootstrap a new recon,
> every container will be checked and added to recondb.
> but , for now , recon check all the containers in a container report one by
> one , and each check will take a rpc call to scm. this is too slow and in my
> test environment , it leads to recon oom, because too many containers to be
> consumed are waitting in the message queue .
> this Jira will use a batch method to handle this problem.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
