fix: correct order-service image name typo (ordr-service → order-service)

[github-actions]

Summary

Fixes image name typo in Act-3/argocd/apps/broken-aks-store-all-in-one.yaml that was introduced in commit e5c5ea5.

Closes DevExpGbb#17

Root Cause Analysis

The ArgoCD app agentic-platform-engineering-demo was marked Degraded due to cascading failures from two image name typos introduced in recent commits:

Bug 1 — RabbitMQ image typo (already fixed in e5c5ea5)

• Commit 38055fa introduced: mcr.microsoft.com/azurelinux/base/rabitmq-server:3.13 (single b)
• Commit e5c5ea5 corrected back to: mcr.microsoft.com/azurelinux/base/rabbitmq-server:3.13
• Cluster impact: rabbitmq-0 pod was created with the broken image, causing ImagePullBackOff. StatefulSet spec has been corrected; pod was manually deleted to trigger recreation with the correct image.

Bug 2 — order-service image typo (fixed by this PR)

• Commit e5c5ea5 introduced: ghcr.io/azure-samples/aks-store-demo/ordr-service:2.1.0 (missing e)
• This PR corrects to: ghcr.io/azure-samples/aks-store-demo/order-service:2.1.0
• Cluster impact: New order-service pod (9fc4bb856) stuck in Init:0/1 (init container waiting for RabbitMQ, then would fail image pull)

Cascading Failure

With RabbitMQ down, makeline-service loses its AMQP connection and returns HTTP 500 on /order/fetch, causing virtual-worker to crash with CrashLoopBackOff.

Changes

File	Change
Act-3/argocd/apps/broken-aks-store-all-in-one.yaml	ordr-service:2.1.0 → order-service:2.1.0 (line 196)

Test Plan

1. Merge PR → ArgoCD auto-syncs (automated sync with selfHeal: true)
2. Verify order-service deployment rolls out with new ReplicaSet
3. Validate all pods healthy:

   kubectl get pods -n default
   # Expected: rabbitmq-0 Running, order-service Running, virtual-worker Running

4. Check ArgoCD app health returns to Healthy

Rollback

# Revert the order-service image if needed
kubectl set image deployment/order-service order-service=ghcr.io/azure-samples/aks-store-demo/order-service:2.1.0 -n default