From e2cd157d3112b0cf2a37baaaa2b492017ae6b000 Mon Sep 17 00:00:00 2001 From: "Bobby (Babak) Salamat" Date: Fri, 20 Oct 2017 18:13:51 -0700 Subject: Addressed reviewer comments --- contributors/design-proposals/scheduling/pod-preemption.md | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/contributors/design-proposals/scheduling/pod-preemption.md b/contributors/design-proposals/scheduling/pod-preemption.md index 2b271d9a..707708eb 100644 --- a/contributors/design-proposals/scheduling/pod-preemption.md +++ b/contributors/design-proposals/scheduling/pod-preemption.md @@ -150,10 +150,13 @@ The head of the queue will always be the highest priority pending pod. 1. Move the pod to the list of unschedulable pods. 1. If a node was chosen to preempt pods, set the node name as an annotation with the "NomindatedNodeName" key to the pod. -1. When any pod is scheduled, or terminated, a node is added/removed, or when -labels of pods or nodes change, remove all the pods from the unschedulable pods -list and add them to the pending queue. (Scheduler should keep its existing rate -limiting.) +1. When any pod is terminated, a node is added/removed, or when +pods or nodes updated, remove all the pods from the unschedulable pods +list and add them to the scheduling queue. (Scheduler should keep its existing rate +limiting.) We should also put the pending pods with inter-pod affinity back to +the scheduling queue when a new pod is scheduled. To be more efficient, we may check +if the newly scheduled pod matches any of the pending pods affinity rules before +putting the pending pods back into the scheduling queue. #### Changes to predicate processing -- cgit v1.2.3