bharatviswa504 edited a comment on pull request #1677:
URL: https://github.com/apache/ozone/pull/1677#issuecomment-746694383


   > > > > The default will be -2, all older buckets will have -2, new buckets 
created after this feature will have -1. So, if some one sets >quota on older 
buckets in this way, we can figure it out, just an idea saying here.
   > > > 
   > > > 
   > > > If we want to disallow the quota support for older buckets, one idea 
to do at code level.
   > > 
   > > 
   > > User using old Ozone version may still has the desire to use the new 
Quota feature. It's better not shut the door completely.
   > > @adoroszlai , I'm not sure if it's feasible that we provide a option of 
calculating the bytesInUsed for each bucket during 1.0.0 -> 1.1. upgrade to 
user? So that user has the choice either tolerate the inaccurate bytesInUsed or 
a relative longer upgrade time to get accurate bytesInUsed of buckets.
   > 
   > Hi @ChenSammi, I think it's dangerous recalculating usedBytes in upgrade . 
If the amount of data is too large, it will greatly increase the time, and if 
the processing is not good, it will cause the upgrade to fail.
   > If we want use Quota in old buckets, we can add a command to recalculate 
bucket usedBytes to minimize the impact. User can decide for himself which 
buckets to recalculate.
   
   >Yes during upgrade might not be a feasible idea, it will take a long time 
for an upgrade if there are many buckets/keys in the cluster.
   
   
   > If we want to use Quota in old buckets, we can add a command to 
recalculate bucket usedBytes to minimize the impact. User can decide for 
himself which buckets to recalculate.
   
   Even this will be a costly operation, during that operation we should 
acquire bucket read lock and do the calculation, so all writes will be stalled 
in the system. (As during this operation, we should not allow new writes to get 
accurate bytesused).
   
   
   


----------------------------------------------------------------
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.

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