GeorgeJahad commented on PR #4680:
URL: https://github.com/apache/ozone/pull/4680#issuecomment-1553512855

   
   > it could still append new compaction log entries (and hardlink new SST 
files) in CompactionBegin/CompletedListener.
   
   I don't think the new sst files are a serious problem.  The question is are 
the new compaction log entries a problem.  That is what I need help 
determining.  If they are not a problem, then we don't need to pause 
compactions.
   
   The issue is after the bootstrap process takes the checkpoint of the active 
fs, any compactions before the tarball is finished will create entries to the 
comaction logs.  Those entries don't correspond the active fs in the tarball.  
Once the follower loads the tarball, it will start doing it's own compactions 
which may differ from the compaction log entries.  It could cause conflicting 
entries in the compaction log, which might break the dag.  That is the problem 
we may need to avoid by invoking pauseBackgroundWork().
   
   For example, the leader could compact files a and b into file c while the 
follower could compact them into file d.  What is the compaction dag supposed 
to do with conflicting entries like that?  That is why I'm leaning towards 
stopping compactions.
   
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to