Managing ReplicaSets

1. Listing ReplicaSets

kubectl get rs
kubectl get rs -l app=web
kubectl get rs -o wide

Section	Info
Selector	matchLabels
Replicas	Desired/current/ready
Pods Status	Running/Waiting/Succeeded/Failed counts
Pod Template	Containers, labels
Events	Created/Deleted pod actions

kubectl scale rs web-7f8 --replicas=5

Warning: Scaling a Deployment-owned RS directly will be reverted by the Deployment controller. Scale the Deployment instead.

Command	Behavior
`kubectl delete rs NAME`	Delete RS + pods (cascading)
`kubectl delete rs NAME --cascade=orphan`	Keep pods orphaned

spec:
  selector:
    matchLabels: { app: web, pod-template-hash: 7f8abc }

Property	Detail
pod-template-hash	Added by Deployment to disambiguate RS
Immutable	Selector cannot change after creation

kubectl get rs web-7f8 -o jsonpath='{.metadata.ownerReferences}'
kubectl tree deploy web                # plugin: show ownership tree

Field	Detail
ownerReferences[].kind	Usually `Deployment`
controller: true	Marks owner that reconciles

kubectl get pods -l app=web,pod-template-hash=7f8abc

Approach	Use
label selector	Match RS selector
field selector	`--field-selector=spec.nodeName=node1`

--cascade	Effect
background (default)	Owner deleted; GC removes children
foreground	Owner stays Terminating until children deleted
orphan	Owner deleted; children remain

Symptom	Likely Cause
Pods stuck Pending	Unschedulable (resources, selector, taints)
RS stuck at 0 ready	Readiness probe failing
Image pull error	Wrong image/tag, missing imagePullSecret
Selector mismatch	RS selector doesn't match template labels
Stale RS persists	revisionHistoryLimit too high or paused