MAX_AGE_DAYS: 0

Author: yerzhan7

.github/workflows/e2e-rosa-tests.yaml

@@ -114,7 +114,7 @@ jobs:
         id: check_cluster_age
         working-directory: tests/e2e-kubernetes/rosa/
         env:
-          MAX_AGE_DAYS: 3
+          MAX_AGE_DAYS: 0
         run: |
           # Check if cluster creation timestamp exists in Terraform state
           CLUSTER_CREATED=$(terraform show -json | jq -r '
@@ -144,7 +144,7 @@ jobs:
             echo "destroy_cluster=false" >> $GITHUB_OUTPUT
           fi
       - name: Destroy old cluster
-        if: steps.check_cluster_age.outputs.destroy_cluster == 'true'
+        if: steps.check_cluster_age.outputs.destroy_cluster != 'false'
         working-directory: tests/e2e-kubernetes/rosa/
         env:
           TF_VAR_aws_region: "${{ vars.AWS_REGION }}"

tests/e2e-kubernetes/testsuites/credentials.go

@@ -139,7 +139,8 @@ func (t *s3CSICredentialsTestSuite) DefineTests(driver storageframework.TestDriv
 	// we're trying to set a more restricted role (e.g. with "AmazonS3ReadOnlyAccess" policy),
 	// in these test cases to ensure it does not fallback to Driver-level 4) credentials.
 	// NOTE: In OpenShift clusters as IMDS is not available we rely on Driver-level IRSA 2)
-	// credentials instead of IAM instance profile.
+	// credentials instead of IAM instance profile. EKS Pod Identity and IAM Instance Profile
+	// are not tested on OpenShift because they are not supported.
 
 	f := framework.NewFrameworkWithCustomTimeouts(NamespacePrefix+"credentials", storageframework.GetDriverTimeouts(driver))
 	f.NamespacePodSecurityLevel = admissionapi.LevelPrivileged