Prepare and test your skills
Prepare and test your skills
Free sample question, answered and explained below. The Professional Cloud Architect practice bank has 2,335 more like it. Try 10 free: instant score, no account.
An e-commerce platform running a microservices architecture on Google Kubernetes Engine (GKE) with Cloud Service Mesh experiences intermittent HTTP 500 errors and latency spikes during peak shopping periods.
The operations team must determine whether the failures originate from proxy routing issues or unhandled application code exceptions, while also identifying latency bottlenecks across downstream microservice calls.
Which systematic diagnostic procedure should the team use to isolate the root cause?
Restarting the istio-ingressgateway deployment tears down active gateway pods and terminates existing TCP connections and client pools. Inspecting Kubernetes node system logs (such as kubelet, containerd, or OS journal logs) provides information regarding operating system events, node resource pressure, and container lifecycle events.
kubelet, dockerd/containerd). They do not capture container-level stdout/stderr application exceptions, HTTP response codes, or proxy routing metadata.The cluster were suffering from an unresponsive Kubernetes node operating system or node-level container runtime failure preventing pods from scheduling.
Troubleshooting best practices dictate gathering observability data (logs, traces, error reports) before executing remediation actions like pod restarts, which destroy ephemeral diagnostic state.
Google Cloud Observability—encompassing Cloud Logging, Cloud Trace, and Error Reporting—provides a unified telemetry framework designed to diagnose distributed systems. In service mesh environments, sidecar and ingress proxies generate structured access logs, application runtimes report exceptions, and distributed tracing tracks RPC latencies across microservice boundaries.
RESPONSE_FLAGS (e.g., UF for upstream connection failure, UT for upstream timeout) and RESPONSE_CODE. When RESPONSE_FLAGS is empty or unset (-) while RESPONSE_CODE is 500, the error originates directly from the backend application rather than proxy-level routing.
This multi-tiered observability approach systematically addresses every level of the failure domain—infrastructure and networking at the proxy layer, application logic via error reporting, and inter-service performance via distributed tracing.
istioctl analyze is a static configuration diagnostic tool that scans Kubernetes cluster resources and Istio custom resource definitions (CRDs) for structural or semantic errors, such as missing sidecar injection labels, unresolvable gateway hosts, or mismatched destination rules. Cloud Monitoring alerting policies evaluate aggregated numerical metrics against predefined thresholds to send notifications when anomalies occur.
istioctl analyze evaluates static configuration definitions. It cannot monitor live application runtimes, intercept dynamic traffic, or capture application-level unhandled exceptions and crash stack traces.RESPONSE_FLAGS to determine if a specific 500 error was generated by the proxy layer or the backend pod.istioctl analyze finds misconfigurations before or during deployments (e.g., misconfigured route rules).istioctl.The goal was to validate whether recent Kubernetes YAML deployment manifests contained syntax errors or mismatched Istio CRD configurations prior to routing production traffic.
While istioctl analyze is useful during initial deployment troubleshooting, live troubleshooting of 5xx errors requires dynamic telemetry (logs, traces, and exception aggregations).
VPC Flow Logs record network flow samples for TCP and UDP traffic traversing network interfaces within Google Cloud Virtual Private Cloud (VPC) subnets, capturing metadata such as source IP, destination IP, port, and packet counts. The Cloud Monitoring Metrics Management page is a cost-governance tool designed to analyze ingestion volumes, metric cardinality, and billing costs for user-defined and system metrics.
RESPONSE_FLAGS, making it impossible to detect HTTP 500 application errors.The primary requirement was auditing network egress volume, debugging subnetwork routing tables, or reducing Cloud Monitoring ingestion costs.
Tracing distributed microservices requires context propagation (like W3C traceparent headers) evaluated in Cloud Trace, which operates at Layer 7 rather than the VPC subnetwork packet level.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated