manirajv06 commented on code in PR #1096:
URL: https://github.com/apache/yunikorn-core/pull/1096#discussion_r3628154821
##########
pkg/scheduler/objects/application.go:
##########
@@ -1085,8 +1085,29 @@ func (sa *Application) tryAllocate(headRoom
*resources.Resource, allowPreemption
request.setHeadroomCheckPassed(sa.queuePath)
requiredNode := request.GetRequiredNode()
+
+ // run predicates for this pod before in hand and fetch
feasible nodes
+ feasibleNodes, predicatesResult :=
request.preAllocateConditions(true)
Review Comment:
> If current pods are a problem we need to handle that via the preemption if
needed.
BTW, YUNIKORN-3330 is all about running preemption predicate checks during
preemption for daemon set pods through `tryReservedAllocate` cycle. If at all
for some reason check fails, then we don't have any other option waiting for
the situation to change on its own. Up until that point, pods would be in
"reserve" state only.
--
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]
