Prepare and test your skills
Prepare and test your skills
Free sample question, answered and explained below. The Professional Cloud DevOps Engineer practice bank has 1,455 more like it. Try 10 free: instant score, no account.
A DevOps engineer deploys a new version of a microservice to a Google Kubernetes Engine (GKE) cluster integrated with Cloud Service Mesh and Google Cloud Observability. After deployment, automated canary verification fails due to missing service performance telemetry and elevated upstream errors.
The engineer observes the following symptoms:
app and istio-proxy).traceId with proxy logs shows traffic being handled as opaque TCP streams rather than HTTP/gRPC protocol traffic.Which root cause and diagnostic action explain this deployment failure?
http or http- prefix; inspect the Kubernetes Service definition and rename the port to restore HTTP layer telemetry and protocol handling.In Cloud Service Mesh and GKE, the service mesh relies on Kubernetes Service port definitions to determine the application protocol (e.g., HTTP, HTTP/2, gRPC). When port names do not begin with an accepted protocol prefix such as http, http2, or grpc (or use the standard appProtocol field), the Envoy proxy defaults to treating traffic as raw, unparsed TCP streams.
http or http- naming convention, Cloud Service Mesh treats incoming connections as generic TCP.http or http- enables Envoy to parse Layer 7 HTTP request headers, inject distributed tracing context headers, and generate appropriate HTTP-level telemetry and metrics.traceparent, x-cloud-trace-context) and attach trace spans automatically.Service manifest without altering application binaries.Other possibilities (such as proxy injection failure or API quota exhaustion) are ruled out by the telemetry evidence: pods already possess running istio-proxy containers, and raw TCP logs are successfully ingested. Modifying the Service port name directly fixes the root protocol mismatch.
istio-proxy sidecars were not injected into the application pods; update the namespace labels to enable automatic sidecar injection.Automatic sidecar injection in Cloud Service Mesh uses a mutating webhook to automatically inject the Envoy proxy (istio-proxy) container and init container into newly scheduled pods within namespaces labeled with istio-injection=enabled or the mesh revision tag.
app and istio-proxy).The diagnostic output had shown pods running 1/1 containers and istioctl analyze had warned that the namespace was missing injection labels (Warn [IST0102]).
When troubleshooting observability issues in Cloud Service Mesh, verifying container count is step one. Once verified, troubleshooting should move upstream to telemetry filters, port configurations, and API quotas.
monitoring.googleapis.com.The Cloud Monitoring API enforces rate limits on metric data ingestion via quotas. When a project exceeds its metric ingestion quota, API requests to monitoring.googleapis.com return HTTP 429 Too Many Requests error responses.
429 errors in the Traffic By Response Code dashboard for the Cloud Monitoring API.The API & Services dashboard displayed spikes in 429 Rate Limit Exceeded responses for monitoring.googleapis.com, and all existing services across the mesh simultaneously lost metric reporting while maintaining proper HTTP log parsing.
While checking Monitoring API quotas is a valid step when telemetry is globally missing, isolated protocol parsing errors point directly to workload or service configuration mismatches.
stdout and increase memory limits to avoid OOMKilled states.An application runtime failure or OutOfMemory (OOMKilled) error occurs when the application container crashes, enters a CrashLoopBackOff state, or terminates unexpectedly due to unhandled exceptions or resource exhaustion.
Running state with both containers active and serving requests (albeit over TCP streams).kubectl get pods.137 in Kubernetes events.The pod status indicated CrashLoopBackOff, the lastState.terminated.reason was OOMKilled, and container logs revealed fatal application exceptions during initialization.
Distinguishing between application runtime crashes and configuration mismatches requires correlating pod lifecycle statuses with proxy telemetry in Cloud Logging and Cloud Trace.
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 Validated