Pod Annotator only sets the first container in the list and ignores init containers.

[kannon92]

The code below patches the containers of the first pod. There are two bugs and one question I see here:

1. An init container can have access to a GPU so one may not be able to get the MIG slice for the init container.

2. Pods can take a list of containers and this only edits the first container in the list.

Question: Ephemeral containers are also ignored but not sure if they need access to the MIG slice.

func (a *PodAnnotator) Handle(ctx context.Context, req admission.Request) admission.Response {

...
pod.Spec.Containers[0].EnvFrom = append(pod.Spec.Containers[0].EnvFrom, v1.EnvFromSource{
ConfigMapRef: &v1.ConfigMapEnvSource{
LocalObjectReference: v1.LocalObjectReference{Name: configMapName},
},
})

// Add extended resource to resource limits
if pod.Spec.Containers[0].Resources.Limits == nil {
	pod.Spec.Containers[0].Resources.Limits = make(v1.ResourceList)
}
pod.Spec.Containers[0].Resources.Limits[v1.ResourceName(extendedResourceName)] = resource.MustParse("1")

// Marshal the updated pod object back to JSON
marshaledPod, err := json.Marshal(pod)
if err != nil {
	return admission.Errored(500, fmt.Errorf("could not marshal pod: %v", err))
}

// Return the patch response
return admission.PatchResponseFromRaw(req.Object.Raw, marshaledPod)

}

[asm582]

Thanks for bringing this up, this is by design.

• Can the init container use the same slice as the actual "workload" container?
• Do pods that take the list of containers use the same slice or need different slices?

[asm582]

This came up in a different PR review, comment: https://github.com/openshift/instaslice-operator/pull/141/files/b36341967d2a863d38d08993917af4d8ecc0aa75#r1792193024

[openshift-bot]

Issues go stale after 90d of inactivity.

Mark the issue as fresh by commenting /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.
Exclude this issue from closing by commenting /lifecycle frozen.

If this issue is safe to close now please do so with /close.

/lifecycle stale

[kannon92]

/remove-lifecycle stale