dlmarion commented on PR #3801:
URL: https://github.com/apache/accumulo/pull/3801#issuecomment-1750784663

   > > If we could leverage the non-primary managers, then this PR could likely 
be closed leaving the compactors the way that they are today.
   > 
   > I think this PR is still, useful but maybe scoped down. Thinking of this 
at a really high level and seeing the following. What I am trying to puzzle out 
is where actual code links in to these high level concepts. Also, is the 
following the complete high level picture?
   > 
   > Generalizing distributed work in Accumulo
   > 
   > 1. Work must be found, currently TGW+TabletMgtmIterator for log sort and 
compaction
   > 2. Work must partitioned and prioritized (multiple in memory bounded prio 
queues that can replace stuff)
   > 3. Workers need to find/request work (thrift task RPCs in this PR)
   > 4. Work needs to be done (may need to emit status and/or metrics about 
work).
   > 5. Results of work needs to be committed (this is highly dependent on the 
work)
   
   There is also the issue of checking for dead workers so that the task can be 
queued up again. We have this today in the Manager with the 
DeadCompactionDetector.


-- 
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