keith-turner commented on issue #5047:
URL: https://github.com/apache/accumulo/issues/5047#issuecomment-2520606151

   > Is this check originating in the TabletServer to check it's own lock? If 
so, each process knows which resource group it is in, there is no need to check 
them all.
   
   Not in this case.  I observed the problem happening in the tablet location 
cache which checks the lock in zoo cache prio to returning a location.  This 
was causing every scan/write in accumulo to make a read to zookeeper.  Not sure 
if this is workable or worthwhile but wondering about adding the RG to the 
tsever location stored in the metadata table.  Then the tablet location cache 
would know of it.  Also scans of the metadata table would know of it and it 
could be exposed via TabletInformation.  The RG is invariant for the lifetime 
of the tserver process so storing it there seems like it would be safe.  The 
downside is its just reduntant data.  If we did this the location entry in the 
metadata table might look like `<RG>:<host>:<port>[<session>]`
   
   If there are places in the code that are doing a lock check and the RG is 
known then it would be nice to pass it along.  Need to survey the code and see 
if there are any places like this.


-- 
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]

Reply via email to