fix: add retry and caching to cert-manager manifest download in e2e-nightly (#324)

Align the nightly workflow's cert-manager install with the composite
action's pattern: download with curl -fsSL (3-attempt retry), cache
the manifest across runs, and apply from the local file. A transient
GitHub CDN failure (504, rate limit) during the download would
otherwise kill the E2E job.

Closes #322

Signed-off-by: Sebastien Tardif <sebtardif@ncf.ca>

Author: SebTardif

.github/workflows/e2e-nightly.yaml

@@ -131,6 +131,7 @@ jobs:
             /tmp/cert-manager-controller.tar
             /tmp/cert-manager-webhook.tar
             /tmp/cert-manager-cainjector.tar
+            /tmp/cert-manager.yaml
             /tmp/prometheus.tar
             /tmp/stress-ng.tar
             /tmp/busybox.tar
@@ -317,10 +318,30 @@ jobs:
             /tmp/cert-manager-cainjector.tar \
             -c "$CLUSTER_NAME"
 
+      - name: Download cert-manager manifest
+        shell: bash -Eeuo pipefail -x {0}
+        run: |
+          if [[ -f /tmp/cert-manager.yaml ]]; then
+            echo "Cached: /tmp/cert-manager.yaml"
+          else
+            for attempt in 1 2 3; do
+              if curl -fsSL -o /tmp/cert-manager.yaml \
+                "https://github.com/cert-manager/cert-manager/releases/download/${{ env.CERT_MANAGER_VERSION }}/cert-manager.yaml"; then
+                break
+              fi
+              echo "::warning::cert-manager.yaml download attempt $attempt failed, retrying in 10s..."
+              sleep 10
+              if (( attempt == 3 )); then
+                echo "::error::Failed to download cert-manager.yaml after 3 attempts"
+                exit 1
+              fi
+            done
+          fi
+
       - name: Install cert-manager
         shell: bash -Eeuo pipefail -x {0}
         run: |
-          kubectl apply -f https://github.com/cert-manager/cert-manager/releases/download/${{ env.CERT_MANAGER_VERSION }}/cert-manager.yaml
+          kubectl apply -f /tmp/cert-manager.yaml
           kubectl wait --for=condition=Available deployment/cert-manager -n cert-manager --timeout=120s
           kubectl wait --for=condition=Available deployment/cert-manager-webhook -n cert-manager --timeout=120s
           kubectl wait --for=condition=Available deployment/cert-manager-cainjector -n cert-manager --timeout=120s