Validate SR-IOV Operator Health Across Multiple Worker Nodes

💡 Quick Answer: Validate daemon pods, node states, VF counts, and allocatable resources on each worker; then schedule one SR-IOV test pod per node and confirm secondary interface attachment.

Use this checklist for nodes such as worker-a and worker-b.

1) Config Daemon on Each Worker

oc get pods -n openshift-sriov-network-operator -o wide | grep sriov-network-config-daemon

2) NodeState for Each Worker

oc get sriovnetworknodestate -n openshift-sriov-network-operator

3) Verify Target Interfaces in NodeState

oc get sriovnetworknodestate worker-a -n openshift-sriov-network-operator -o json | jq '.status.interfaces[] | {name,pciAddress,numVfs}'
oc get sriovnetworknodestate worker-b -n openshift-sriov-network-operator -o json | jq '.status.interfaces[] | {name,pciAddress,numVfs}'

4) Verify Scheduler Resources on Both Nodes

oc get node worker-a -o json | jq '.status.allocatable'
oc get node worker-b -o json | jq '.status.allocatable'

5) Run One Pod Per Node

Pin test pods with nodeSelector and request the SR-IOV resource (for example openshift.io/mellanoxnics: 1).

6) Confirm Secondary Interface in Pod

oc exec -n <ns> <pod-name> -- ip a

You should see an extra interface (for example net1) from the SR-IOV network attachment.

Success Criteria

• Both nodes expose expected SR-IOV allocatable resources.
• Pod networking succeeds on both nodes.
• No repeated CNI/IPAM errors in pod events.