After updating the Instana agent DaemonSet image tag to a newer version in a Kubernetes cluster, multiple agent pods enter CrashLoopBackOff status. The administrator needs to identify the root cause quickly. Which action will most directly reveal why the pods are crashing?
Select an answer to reveal the explanation.
Short Explanation and Infographic
When a pod is in CrashLoopBackOff, the container process is starting and immediately exiting. The kubectl logs command retrieves the container stdout and stderr, which for the Instana agent will contain the specific startup error such as a missing required environment variable like INSTANA_AGENT_KEY or backend endpoint, an invalid configuration value, or an unhandled startup exception. This is always the most direct and fastest diagnostic step for a crashing container.
Full explanation below image
Full Explanation
When a pod is in CrashLoopBackOff, the container process is starting and immediately exiting. The kubectl logs command retrieves the container stdout and stderr, which for the Instana agent will contain the specific startup error such as a missing required environment variable like INSTANA_AGENT_KEY or backend endpoint, an invalid configuration value, or an unhandled startup exception. This is always the most direct and fastest diagnostic step for a crashing container. The correct answer is 'Run kubectl logs on one of the CrashLoopBackOff agent pods to read the agent process stdout and stderr, which will contain startup errors such as missing required environment variables or configuration validation failures'. The incorrect options — 'Run kubectl describe node to check whether nodes have sufficient allocatable CPU and memory for the agent pods', 'Inspect the DaemonSet manifest for incorrect imagePullSecrets referencing a nonexistent registry credential', 'Examine Kubernetes control-plane audit logs for API server rejection events directed at the agent ServiceAccount' — are wrong because they do not align with IBM Instana's architecture or recommended practices for this scenario. Understanding this concept is essential for the Domain 6: Troubleshooting domain of the IBM Instana Observability certification.