kagent-oci-devops-demo/k8s/remediation

Declarative Remediation Manifests

Use these manifests with OCI DevOps Apply manifest to your Kubernetes cluster stages.

Business remediation

Apply in this order:

business-configmap.yaml
order-rollout-restart.yaml

business-configmap.yaml sets safe business thresholds.

order-rollout-restart.yaml reapplies the order-service deployment template so pods reload ConfigMap-backed environment variables.

Before applying order-rollout-restart.yaml, replace:

<restart-token>

with a new value, for example the OCI DevOps deployment ID, build number, or timestamp. If the annotation does not change, Kubernetes will not restart the pods.

ImagePull remediation

Use the manifest matching the affected deployment:

imagepull-all-services.yaml
imagepull-order.yaml
imagepull-payment.yaml
imagepull-inventory.yaml

For the demo, prefer imagepull-all-services.yaml. It restores ocir-secret across all three business deployments, so the same remediation pipeline works regardless of which service hit ImagePullBackOff.

Each manifest restores:

imagePullSecrets:
  - name: ocir-secret

Before applying an ImagePull remediation manifest, replace:

<region-key>
<tenancy-namespace>
<image-tag>
<restart-token>

with the same values used by the active deployment. The restart-token forces a new ReplicaSet after the secret is restored.

HPA remediation

Use:

order-hpa-capacity.yaml

It changes order-service HPA to:

minReplicas=3
maxReplicas=8
averageUtilization=65