HoustonPutman commented on PR #1730:
URL: https://github.com/apache/solr/pull/1730#issuecomment-1615147207

   Ok the tests now pass (fixed a weird bug where splitShard can send a 
placementRequest with no actual replicas to add... but anyways)
   
   The "hacky" fix for tied-nodes, is now expanded and no longer "hacky". We 
only loop once to retry replicas skipped because of tied nodes, but that should 
be ok. We can improve this later by looping more if we need to. However, the 
code is now generalized for any amount of tied-nodes, and will act accordingly 
given the number of replicas to place and the number of tied nodes.
   
   I wish that this fix could have been separate from this PR, but there is 
little point having this MigrateReplicas API if it isn't going to do a good 
migration.
   
   @tflobbe Amazingly, `AffinityPlacementFactoryTest.testScalability` is now 
20% faster than before this PR (which is roughly 60-70% faster than a few weeks 
ago). I haven't taken the time to see how this code sped it up so much, but 
I'll take the win.


-- 
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