[
https://issues.apache.org/jira/browse/HDDS-8685?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
ASF GitHub Bot updated HDDS-8685:
---------------------------------
Labels: pull-request-available (was: )
> CreateCheckpoint needs to be idemp[otent
> ----------------------------------------
>
> Key: HDDS-8685
> URL: https://issues.apache.org/jira/browse/HDDS-8685
> Project: Apache Ozone
> Issue Type: Sub-task
> Reporter: Prashant Pogde
> Assignee: Aswin Shakil
> Priority: Major
> Labels: pull-request-available
>
> Create snapshot can be called multiple times with log replay. We need to
> make sure that it's idempotent. Consider scenarios
>
> * Create rocksDB checkpoint itself can cause issues.
> - RocksDB will create hardlinks in a temp directory
> - after links are created, temp dir is renamed to snapshot-id
> - we make entry in snapshot info table
>
> Operation could fail at any time due to OM crash/restart and log replay will
> start from the beginning here. We need to make sure
> * We need to make sure clean up the temp directory. un-removed hardlinks
> will pose a problem because we sst files with these hardlink will not get
> cleaned up when the snapshots are deleted due to extra link.
> * CreateCheckpoint doesn't try to recreated the checkpoint again. (rocksDB
> operation would fail)
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
