Correct the failure duration for unavailable checks

[johscheuer]

Description

Fix the error reporting for the unavailability checker in cases where the cluster gets available again, e.g.:

  2024/11/29 01:48:16 invariant InvariantClusterStatusAvailableWithThreshold failed for the first time
  2024/11/29 01:48:16 invariant InvariantClusterStatusAvailableWithThreshold failed: cluster is not available
  2024/11/29 01:48:31 invariant InvariantClusterStatusAvailableWithThreshold failed after: 15.223519039s
  2024/11/29 01:48:37 invariant InvariantClusterStatusAvailableWithThreshold failed after: 20.467108341s
  2024/11/29 01:48:44 invariant InvariantClusterStatusAvailableWithThreshold true again

...

Operator Upgrades [AfterEach] upgrading a cluster with a pending pod Upgrade from 7.1.57 to 7.1.63 with a pending pod [e2e, pr]
  [AfterEach] /codebuild/output/src3856629688/src/github.com/FoundationDB/fdb-kubernetes-operator/e2e/test_operator_upgrades/operator_upgrades_test.go:91
  [It] /codebuild/output/src3856629688/src/github.com/FoundationDB/fdb-kubernetes-operator/e2e/fixtures/upgrade_test_configuration.go:115

  [FAILED] Unexpected error:
      <*errors.errorString | 0xc00af12930>: 
      invariant InvariantClusterStatusAvailableWithThreshold failed for 235ns
      {
          s: "invariant InvariantClusterStatusAvailableWithThreshold failed for 235ns",
      }
  occurred 

invariant InvariantClusterStatusAvailableWithThreshold failed for 235ns <- Is the wrong duration.

Type of change

Please select one of the options below.

• Bug fix (non-breaking change which fixes an issue)

Discussion

Testing

Documentation

Follow-up

[foundationdb-ci]

Result of fdb-kubernetes-operator-pr on Linux CentOS 7

• Commit ID: a2a7d94
• Duration 2:44:12
• Result: ✅ SUCCEEDED
• Error: N/A
• Build Log terminal output (available for 30 days)
• Build Workspace zip file of the working directory (available for 30 days)