[
https://issues.apache.org/jira/browse/HDFS-15770?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17511612#comment-17511612
]
Srinivasu Majeti commented on HDFS-15770:
-----------------------------------------
Hi [~arpiagariu] , I remember we discussed on this some time back but wanted to
double confirm that this is an expected behavior to list Live file path when
there are missing/corrupted blocks specific to snapshot specific file . fsck to
report all snapshot files that missing block relates to needs an enhancement
and design change in the snapshot feature framework. Please confirm/comment and
any chance that it can be addressed in recent future .
> fsck to support printing snapshot name for missing blocks from snapshot only
> files
> ----------------------------------------------------------------------------------
>
> Key: HDFS-15770
> URL: https://issues.apache.org/jira/browse/HDFS-15770
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: fs
> Affects Versions: 3.2.0, 3.3.0
> Reporter: Srinivasu Majeti
> Priority: Minor
>
> Today when there are blockids belonging to an older snapshot are different
> that that of its corresponding live file and block from snapshot file is
> missing , FSCK reports a blockid missing against the live file but no clue if
> that block id belongs to the live file or specific snapshot file path alone
> .[ in the case where file in snapshot and live file system are different with
> some overwrite option ]. It would be nice to show a flag in the fsck output
> that the blockid is missing on live file or snapshot file when using
> -includesnapshots option.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
