Skip to main content

Pod application function exception

Last updated on

Pod application function exception is a Kubernetes pod-level chaos fault that causes a specific function in an instrumented application to throw a configurable exception for a configurable duration. Only the named function is affected; other application code paths run normally. When the fault ends, the function returns to its normal behavior immediately.

Use this fault to validate how callers and dependents behave when a specific business function starts throwing: an unchecked exception from a library wrapper, a domain-specific exception from a validation routine, or any failure path that propagates as a thrown exception rather than a returned error.

Run your first experiment

If you have not configured the chaos infrastructure yet, go to Quickstart to install the chaos infrastructure and run an experiment end to end.

Use cases​

Run this fault when you want to answer concrete questions like:

  • Unchecked exception propagation: When a deep function throws, does the request boundary catch the exception, log it, and return a clean response, or does the stack trace leak to the client?
  • Exception-aware fallback paths: Does a wrapper that catches a specific exception type route to a fallback implementation, or does it rethrow and break the caller?
  • Retry filter correctness: Do retry policies treat the injected exception as retryable or non-retryable as intended, and does the policy match the framework's behavior?
  • Circuit breaker behavior: Does a circuit breaker that counts exceptions open after the configured failure threshold and short-circuit subsequent calls?
  • Observability coverage: Does the exception surface in traces, logs, and alerts with the right error type and message?

Prerequisites​

  • Kubernetes version: 1.21 or later. Go to What's supported to confirm distribution support.
  • Target pod is Running: The application pod is in the Running state.
  • Application is instrumented: The application registers a name and exposes the target function to the chaos infrastructure. Without instrumentation, the chaos pod cannot reach the function.
  • Function is identifiable: The function to fail is reachable by the name set in TARGET_APPLICATION_FUNCTION.
  • Workload selector defined: The chaos experiment knows the target application by name.

Supported environments​

PlatformSupport status
Amazon EKSSupported
Azure AKSSupported
Google GKESupported
Red Hat OpenShiftSupported
RancherSupported
VMware TanzuSupported
Self-managed Kubernetes (CNCF-certified)Supported
GKE AutopilotSupported (no privileged access required)
EKS Fargate, ACI virtual nodesSupported (no container runtime socket required)

Permissions required​

The fault runs under the chaos infrastructure's service account.

Resource (apiGroup)VerbsWhy it is needed
pods ("")get, list, create, delete, deletecollection, patch, updateDiscover the target pod and run the chaos pod
pods/log ("")get, list, watchStream chaos pod logs for status and debugging
deployments, statefulsets, replicasets, daemonsets (apps)get, listResolve the target workload to the pods it owns
events ("")get, list, create, patch, updateRecord fault progress as Kubernetes events
jobs (batch)get, list, create, delete, deletecollectionRun the chaos job that drives the fault

The default Harness chaos infrastructure service account already includes these permissions.

Fault tunables​

Configure the following fault parameters when you add Pod application function exception to an experiment in Chaos Studio. Defaults are shown for reference.

Required parameters

TunableDescriptionDefault
TARGET_APPLICATION_NAMEName of the target application as registered with the chaos infrastructure.(required)
TARGET_APPLICATION_FUNCTIONName of the function inside the target application to throw the exception from.(required)

Chaos parameters

TunableDescriptionDefault
MESSAGEException message attached to the injected throw. Empty uses a default message.""
TOTAL_CHAOS_DURATIONDuration of the fault in seconds.60
RAMP_TIMEWait period in seconds before and after the fault. Go to ramp time to read how it is applied.0

Tunables that apply to every fault are documented in common tunables for all faults.

Use a recognizable exception message

Set MESSAGE to a unique string so you can grep logs and traces to confirm the injected exception during analysis.

Fault execution in brief​

Signals the instrumented application to make the function named in TARGET_APPLICATION_FUNCTION throw an exception containing MESSAGE for TOTAL_CHAOS_DURATION seconds.

Expected behavior during fault execution​

  • Calls to the named function throw the configured exception. Other functions in the same application run normally.
  • Direct callers see the exception propagate up the stack unless they catch it. Frameworks may surface it as a 5xx response, a queued message rollback, or a propagated failure depending on the runtime.
  • Downstream services may see reduced or absent traffic if the throwing function fronted upstream calls.
  • Error dashboards and traces should show the injected exception type and message alongside any cascading failures.
When the fault ends

The function returns to its normal behavior immediately. In-flight calls that already threw stay failed; new calls succeed as before.

Signals to watch​

Attach resilience probes to assert each layer:

  • Application error rate: Use an HTTP probe against endpoints that exercise the function to detect 4xx/5xx spikes.
  • Function-level metrics: Use a Prometheus probe on the function's exception counter or success rate to confirm the injection.
  • Application logs: Use a command probe to grep container logs for the configured MESSAGE.

Verify the fault execution effect​

While the experiment is running, confirm the function is throwing:

  1. Exercise the function from a client.

    kubectl run -n <namespace> tester --image=nicolaka/netshoot --rm -it -- \
      curl -s http://<service>:<port>/<endpoint-that-calls-the-function>

    The response should reflect the failure, either as an HTTP error or an error payload that mentions the injected exception.

  2. Confirm the exception surfaces in logs.

    kubectl logs -n <namespace> <target-pod> --tail=200 | grep "<MESSAGE>"

    The configured MESSAGE should appear in stack traces or error logs for each thrown invocation.

Recovery and cleanup​

  • End of duration: The function returns to its normal behavior automatically.
  • Abort the experiment: Stopping the experiment from Chaos Studio triggers the same cleanup path.
  • Stuck state: If the application caches the failure (for example by tripping a circuit breaker that does not reset on its own), restart the pod to clear the cached state.

Limitations​

  • Instrumentation required: The fault only affects applications that have registered themselves and their functions with the chaos infrastructure. Uninstrumented applications cannot be targeted.
  • Function-name granularity: Only one function at a time is targeted. Use multiple experiments in sequence for multi-function scenarios.
  • Exception type is fixed by the instrumentation: The runtime exception type thrown is determined by the application's instrumentation layer; only MESSAGE is configurable. Use Pod JVM method exception for JVM-specific exception-type control.

Troubleshooting​

Pod application function exception experiment stays Pending or never starts in Harness Chaos Engineering

Inspect the chaos pods in the experiment namespace with kubectl describe pod -n <chaos-namespace>. The most common causes are taints on the target node that the chaos pods do not tolerate, insufficient resources, or a PodSecurity admission policy. Add the required tolerations to the experiment or adjust the namespace's Pod Security level.

No exceptions observed during pod-application-function-exception

The most common causes are: TARGET_APPLICATION_NAME does not match the registered application name; TARGET_APPLICATION_FUNCTION does not match a registered function; the application image does not include the chaos instrumentation; or the call path under test never invokes the named function. Verify registration by listing instrumented applications and confirm by exercising the function with a known client.

Function appears to throw intermittently after pod-application-function-exception ends

Check whether the application has a circuit breaker or cooldown window that keeps the function in a degraded state after the injection ends. Adjust the circuit breaker reset interval or restart the pod to clear it.