Making unschedulable pods schedulable again after ResourceSlice cluster events
was accidentally left out when adding structured parameters to Kubernetes 1.30.
All E2E tests were defined so that a driver starts first. A new test with a
different order (create pod first, wait for unschedulable, start driver)
triggered the bug and now passes.
The event is only relevant when DRAControlPlaneController (= "classic DRA") is
enabled.
This change has no effect in practice because the only plugin using this event,
the dynamic resource plugin, also checks feature gates when asking for events
and correctly only asks for PodSchedulingContext events when
DRAControlPlaneController is enabled.
The structured parameter allocation logic was written from scratch in
staging/src/k8s.io/dynamic-resource-allocation/structured where it might be
useful for out-of-tree components.
Besides the new features (amount, admin access) and API it now supports
backtracking when the initial device selection doesn't lead to a complete
allocation of all claims.
Co-authored-by: Ed Bartosh <eduard.bartosh@intel.com>
Co-authored-by: John Belamaric <jbelamaric@google.com>
This is in preparation for revamping the resource.k8s.io completely. Because
there will be no support for transitioning from v1alpha2 to v1alpha3, the
roundtrip test data for that API in 1.29 and 1.30 gets removed.
Repeating the version in the import name of the API packages is not really
required. It was done for a while to support simpler grepping for usage of
alpha APIs, but there are better ways for that now. So during this transition,
"resourceapi" gets used instead of "resourcev1alpha3" and the version gets
dropped from informer and lister imports. The advantage is that the next bump
to v1beta1 will affect fewer source code lines.
Only source code where the version really matters (like API registration)
retains the versioned import.
This enables connecting the event handler for ResourceClaim to the assume
cache, which addresses a theoretic race condition.
It may also be useful for implementing the autoscaler support, because now
the autoscaler can modify the content of the cache.
When a claim uses structured parameters, as indicated by the resource class
flag, the scheduler is responsible for allocating it. To do this it needs to
gather information about available node resources by watching
NodeResourceSlices and then match the in-tree claim parameters against those
resources.
When filtering fails because a ResourceClass is missing, we can treat the pod
as "unschedulable" as long as we then also register a cluster event that wakes
up the pod. This is more efficient than periodically retrying.
Because of a misplaced `append` (should have been inside if clause, not after
it), some handler from a previous loop iteration was added again. This was
harmless because the resulting slice was only used for waiting for cache sync,
but should better get fixed anyway.
The name "PodScheduling" was unusual because in contrast to most other names,
it was impossible to put an article in front of it. Now PodSchedulingContext is
used instead.
When the client misses a delete event from the watcher, it will use the last state of the pod in the informer cache to produce a delete event. At that point, it's not clear if the pod was in the queues or the cache, so we should issue a deletion in both.
The pod could be assumed, so deletion of assumed pods from the cache should work.
Change-Id: I11ce9785de603924fc121fe2fa6ed5cb1e16922f
