You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@yunikorn.apache.org by GitBox <gi...@apache.org> on 2020/02/11 23:31:18 UTC

[GitHub] [incubator-yunikorn-core] yangwwei commented on issue #89: Core allocation/reservation logic renovation

yangwwei commented on issue #89: Core allocation/reservation logic renovation
URL: https://github.com/apache/incubator-yunikorn-core/pull/89#issuecomment-584913754
 
 
   Hi @wilfred-s 
   
   Thanks for the patch, overall the patch does a really good job to clean up things. Now the structure looks very clear. I have some high-level comments, most about reservation code:
   
   1. scheduling_application: 392-398 this internal unreserve -- would it cause some left-over things? compare to scheduling_partiton#unReserve. Ideally, it will be great if we can encapsulate the unreserve logic in functions and just call them when needed.
   
   2. We need to make sure in `tryAllocate`, when an allocation is made for a reserved request, we need to do unreserve properly.  Right now, it seems the unreserve did not happen for this case.
   
   3. For reserved apps, if allocation from reservation doesn't work, let's try to allocate on all nodes if possible.  Otherwise, such reservations may be "stuck" on some nodes which doesn't serve the reservation purpose.
   
   I will look into more details shortly.
   
   Thanks
   
   
   
   
   
   
   

----------------------------------------------------------------
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:
users@infra.apache.org


With regards,
Apache Git Services

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@yunikorn.apache.org
For additional commands, e-mail: dev-help@yunikorn.apache.org