sumitagrawl commented on PR #4817:
URL: https://github.com/apache/ozone/pull/4817#issuecomment-1590795787

   > > > Can we think of triggering quota repair at bucket level in Recon 
instead of CLI, will be more intuitive and user friendly as well
   > > 
   > > 
   > > How would OM learn about the repaired values? Right now information 
flows from OM to Recon, not the other way around. The proposed CLI would be run 
on each OM host, and would be provided the OM DB to operate on so it can run 
offline while OM is shut down.
   > 
   > @errose28 thanks for clarifying, however in future we want to make Recon a 
tool which can trigger actions other way around as well, so if we can provide 
om db file path to recon while quota repair, would it make sense and push back 
the file to OM db file same path after repair ? Am I missing something?
   
   @devmadhuu There are below reason where this kind of operation normally is 
avoided, even from recon,
   1. If customer have 1 billion keys, it will take 30 min as per current perf 
measurement for each bucket. It will lock the bucket for the duration and no 
write operation will be allowed.
   2. This will be debug operation if count goes wrong where it make take this 
much time
   3. If Recon integration is required, may need think another solution then 
CLI trigger from Recon as part of this.
   
   This can be discussed as part of 
[HDDS-8824](https://issues.apache.org/jira/browse/HDDS-8824) for CLI cases and 
further support.


-- 
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