feat(cluster_healthcheck): add cluster health validation role

[stevefulme1]

Summary

Adds a new cluster_healthcheck role that validates the health of an OpenShift cluster for virtualization migration readiness. The role performs comprehensive checks across six categories and generates an HTML summary report with pass/fail/warning status and actionable recommendations.

Health checks included

• OCP Node Health - Node Ready status, MemoryPressure/DiskPressure/PIDPressure conditions, allocatable vs capacity ratios, kubevirt.io/schedulable label verification
• KubeVirt Health - HyperConverged CR conditions (Available/Degraded), virt-operator/controller/handler/api pods, CDI operator and deployment health
• MTV Health - ForkliftController CR status, MTV operator pods, Provider readiness, failed migration Plans
• Storage Health - StorageClass enumeration and default verification, CSI driver discovery, PV capacity, pending PVC detection
• Network Health - Multus pods, NetworkAttachmentDefinitions, OVN-Kubernetes/OpenShiftSDN health, migration network configuration
• Post-Migration VM - VirtualMachineInstance running state, guest agent reporting, network interface IPs, optional SSH connectivity

Files added

roles/cluster_healthcheck/
├── defaults/main.yml
├── meta/main.yml
├── README.md
├── tasks/
│   ├── main.yml
│   ├── ocp_node_health.yml
│   ├── kubevirt_health.yml
│   ├── mtv_health.yml
│   ├── storage_health.yml
│   ├── network_health.yml
│   ├── post_migration_vm.yml
│   └── report.yml
├── templates/
│   └── cluster_healthcheck_report.html.j2
├── tests/
│   ├── inventory
│   └── test.yml
└── vars/main.yml
playbooks/cluster_healthcheck.yml

Design decisions

• Follows existing validate_migration role patterns (task naming, k8s_info usage, variable prefixing)
• All variables prefixed with cluster_healthcheck_ per collection convention
• Private/internal variables use __cluster_healthcheck_ double-underscore prefix
• Uses FQCNs throughout (kubernetes.core.k8s_info, ansible.builtin.*)
• Check list is configurable via cluster_healthcheck_checks default
• Post-migration VM checks are opt-in via cluster_healthcheck_post_migration_vms
• HTML report includes per-category breakdown with recommendations

Testing

• ansible-lint --profile production passes with 0 errors on the role (playbook FQCN resolution matches existing collection behavior)

[sabre1041]

Review the issues that are being reported.

Also, please review conflicted files

[sabre1041]

@stevefulme1 looking better. still seeing misalignment on the CDI components. Here are the labels that are applied to the CDI pods

cdi-apiserver-76789fd699-7ft5j                         1/1     Running   0          6h      app.kubernetes.io/component=storage,app.kubernetes.io/managed-by=cdi-operator,app.kubernetes.io/part-of=hyperconverged-cluster,app.kubernetes.io/version=4.21.8,app=containerized-data-importer,cdi.kubevirt.io=cdi-apiserver,np.kubevirt.io/allow-access-cluster-services=true,operator.cdi.kubevirt.io/createVersion=4.21.3,pod-template-hash=76789fd699
cdi-deployment-78b4485977-ft8xh                        1/1     Running   0          6h      app.kubernetes.io/component=storage,app.kubernetes.io/managed-by=cdi-operator,app.kubernetes.io/part-of=hyperconverged-cluster,app.kubernetes.io/version=4.21.8,app=containerized-data-importer,cdi.kubevirt.io=cdi-deployment,np.kubevirt.io/allow-access-cluster-services=true,operator.cdi.kubevirt.io/createVersion=4.21.3,pod-template-hash=78b4485977,prometheus.cdi.kubevirt.io=true
cdi-operator-f9dcf5cf6-bfpjw                           1/1     Running   0          6h1m    app.kubernetes.io/component=storage,app.kubernetes.io/managed-by=olm,app.kubernetes.io/part-of=hyperconverged-cluster,app.kubernetes.io/version=4.21.8,cdi.kubevirt.io=cdi-operator,name=cdi-operator,np.kubevirt.io/allow-access-cluster-services=true,operator.cdi.kubevirt.io=,pod-template-hash=f9dcf5cf6,prometheus.cdi.kubevirt.io=true
cdi-uploadproxy-7dcfc947d7-jd5kb                       1/1     Running   0          6h      app.kubernetes.io/component=storage,app.kubernetes.io/managed-by=cdi-operator,app.kubernetes.io/part-of=hyperconverged-cluster,app.kubernetes.io/version=4.21.8,app=containerized-data-importer,cdi.kubevirt.io=cdi-uploadproxy,np.kubevirt.io/allow-access-cluster-services=true,operator.cdi.kubevirt.io/createVersion=4.21.3,pod-template-hash=7dcfc947d7

[stevefulme1]

Fixed the CDI label selectors and added the missing components:

• Changed CDI label selectors from app.kubernetes.io/component to cdi.kubevirt.io to match actual pod labels on OCP 4.21+
• Added health checks for cdi-apiserver and cdi-uploadproxy pods (were missing from the original implementation)
• Added both to the health report details section

All four CDI pods are now covered:

• cdi.kubevirt.io=cdi-operator
• cdi.kubevirt.io=cdi-deployment
• cdi.kubevirt.io=cdi-apiserver
• cdi.kubevirt.io=cdi-uploadproxy

[stevefulme1]

Fixed the provider readiness reporting issue (item 3 from Andrew's review).

Root cause: The original Jinja {% set not_ready = [] %} / {{ not_ready }} block inside a >- YAML scalar outputs the Python list's __repr__ as a string (e.g., "[]"). Downstream | length checks then evaluated the string length (2 characters for "[]"), not the list length (0 items) — so it always reported providers as not ready.

Fix: Replaced the inline Jinja loop with a two-step filter chain:

1. selectattr + contains to identify providers with Ready: True condition
2. reject('in', ready_list) to get the complement (not-ready providers)

This produces a proper Python list and also correctly handles providers missing status.conditions entirely (they're included in the not-ready list).

[sabre1041]

LGTM