dlmarion commented on PR #2790:
URL: https://github.com/apache/accumulo/pull/2790#issuecomment-1171440586

   > Wrt the handling of session.queries, does re-adding the unfinishedRange 
make sense? Should we be re-trying the same range immediately or should we be 
using some sort of FIFO queue and adding the unfinished range to the back.
   
   I think you need to add the unfinishedRange to the top of the queue or else 
the check that you have 
[here](https://github.com/apache/accumulo/pull/2790/files#diff-df2b4a76126e09d824ca0ac903abd6563c72db9793844b59e58d82c8a206d695R139)
 won't work I think. I don't believe there is a down-side to immediately 
retrying where you left off, the yield gives the tablet server an ability to 
run a different scan on the scan thread as a means of time-sharing.
   
   > Or are unfinished ranges supposed to be taken care of in the 
TabletServerBatchReaderIterator?
   No, it does not look like the client scan code knows anything about yielding 
iterators. The client continues to look for results while the scan result 
`more` flag is `true`.
   
   >What is the design to get the benefit of yielding?
   It's been a while. I would point you to 
https://issues.apache.org/jira/browse/ACCUMULO-4643 and its related pull 
requests. 


-- 
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