Ctrlk
Sign UpGet a DemoContact Us
For the complete documentation index, see llms.txt. This page is also available as Markdown.

Pod HTTP status code

Override the HTTP response status code returned by a target Kubernetes pod to test client error handling, retry classification, and circuit-breaker behavior on specific HTTP status codes.

Pod HTTP status code is a Kubernetes pod-level chaos fault that overrides the HTTP response status code returned by the target pod for a configurable duration. You can also overwrite the response body to match the new status (for example, return a 500 with an error payload), or leave the body untouched. When the fault ends, status codes return to whatever the application produces naturally.

Use this fault to test how a client handles specific HTTP error codes: a 429 from a rate-limited dependency, a 503 from an overloaded backend, a 404 from a missing resource, or a deliberate 200 returned over a "successful" but broken response.

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:

  • Retry classification: Does the client retry on 503 but not on 400? Does it honor Retry-After on 429?

  • Error budget consumption: Forcing a percentage of 500s reveals whether your SLO alerting and error-budget burn rules fire as expected.

  • Cache invalidation: Returning 404 for resources the client previously cached as 200 exposes stale-cache handling.

  • Auth failure handling: A targeted 401 or 403 reveals whether the client refreshes tokens, prompts the user, or simply loops.

  • Rate-limit handling: Returning 429 with a Retry-After body verifies whether the client backs off and resumes rather than thrashing.

Prerequisites

  • Kubernetes version: 1.21 or later. Go to What's supported to confirm distribution support.

  • Target pods are Running: The application pods you intend to target are in the Running state before the fault is launched.

  • Privileged pods allowed: The cluster lets you schedule privileged pods in the chaos namespace. GKE Autopilot supports this fault but requires the one-time setup in Chaos on GKE Autopilot; other locked-down distributions may need similar exemptions.

  • Container runtime access: The chaos pod can reach the container runtime socket on the target node (/run/containerd/containerd.sock, /var/run/docker.sock, or /var/run/crio/crio.sock).

  • HTTP service on a known port: The target container serves HTTP, HTTPS, or gRPC traffic on a port you can specify with TARGET_SERVICE_PORT.

  • Workload selector defined: The chaos experiment knows the target workload by kind, namespace, and either names or labels.

Supported environments

Platform
Support status

Amazon EKS

Supported

Azure AKS

Supported

Google GKE

Supported

Red Hat OpenShift

Supported

Rancher

Supported

VMware Tanzu

Supported

Self-managed Kubernetes (CNCF-certified)

Supported

GKE Autopilot

Supported with Autopilot setup

EKS Fargate, ACI virtual nodes

Not supported (no access to container runtime sockets)

Permissions required

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

Resource (apiGroup)

Verbs

Why it is needed

pods ("")

get, list, create, delete, deletecollection, patch, update

Discover target pods and run the chaos pod on the same node

pods/log ("")

get, list, watch

Stream chaos pod logs for status and debugging

deployments, statefulsets, replicasets, daemonsets (apps)

get, list

Resolve the target workload to the pods it owns

events ("")

get, list, create, patch, update

Record fault progress as Kubernetes events

jobs (batch)

get, list, create, delete, deletecollection

Run 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 HTTP status code to an experiment in Chaos Studio. Defaults are shown for reference.

Chaos parameters

Tunable
Description
Default

STATUS_CODE

HTTP status code to return. Supported values: 200, 201, 202, 204, 300, 301, 302, 304, 307, 400, 401, 403, 404, 500, 501, 502, 503, 504. 0 picks one at random from the supported list.

0

MODIFY_RESPONSE_BODY

When true, replaces the response body with RESPONSE_BODY (or a default body matching the status code if RESPONSE_BODY is empty). When false, the original body is returned alongside the new status code.

true

RESPONSE_BODY

Body string to overwrite the response when MODIFY_RESPONSE_BODY is true. Empty falls back to a default body matching the status code.

""

CONTENT_TYPE

Content-Type header set on the modified response.

text/plain

CONTENT_ENCODING

Encoding applied to the body. Supported values: gzip, deflate, or empty for no encoding.

""

TARGET_SERVICE_PORT

Port the target container listens on for HTTP traffic.

80

TOXICITY

Percentage of intercepted responses whose status is overridden, between 0 and 100. 100 modifies every response.

100

TOTAL_CHAOS_DURATION

Duration of the fault in seconds.

60

Proxy and interface

Tunable
Description
Default

PROXY_PORT

Port the chaos proxy listens on inside the container's network namespace. Must not conflict with any port already in use on the target container.

20000

NETWORK_INTERFACE

Network interface inside the target container's namespace. Almost always eth0 for standard CNI plugins.

eth0

Targeting

Tunable
Description
Default

TARGET_PODS

Comma-separated list of pod names to target. Empty selects from the workload's pods using POD_AFFECTED_PERCENTAGE.

""

TARGET_CONTAINER

Container in the pod whose network namespace to enter. Empty targets the first container in the pod spec.

""

NODE_LABEL

Label selector to filter target pods by the node they run on. Empty disables node-based filtering.

""

POD_AFFECTED_PERCENTAGE

Percentage of the workload's pods to target. 0 means one pod.

0

SEQUENCE

When multiple pods are targeted, inject parallel (all at once) or serial (one after another).

parallel

Runtime and helper

Tunable
Description
Default

CONTAINER_RUNTIME

Container runtime on the target nodes. One of containerd, docker, crio.

containerd

SOCKET_PATH

Path to the container runtime socket on the target node. Set to match CONTAINER_RUNTIME.

/run/containerd/containerd.sock

RAMP_TIME

Wait period in seconds before and after the fault. Go to ramp time to read how it is applied.

0

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

TEST A CLASS OF FAILURES WITH RANDOM CODES

Setting STATUS_CODE=0 cycles through the supported codes at random. This is useful for fuzz-style testing of a client's error handling across the full spectrum of HTTP failures.

Configure for your container runtime

Set CONTAINER_RUNTIME and SOCKET_PATH to match the runtime on the target node:

CONTAINER_RUNTIME

SOCKET_PATH

containerd (default)

/run/containerd/containerd.sock

docker

/var/run/docker.sock

crio

/var/run/crio/crio.sock

Fault execution in brief

Intercepts HTTP responses on TARGET_SERVICE_PORT inside the container's network namespace and replaces the status code with STATUS_CODE (and optionally rewrites the body), optionally limited to a configurable percentage of responses so other traffic is unaffected.

Expected behavior during fault execution

  • HTTP and HTTPS responses on TARGET_SERVICE_PORT return the configured STATUS_CODE instead of the application's original status.

  • When MODIFY_RESPONSE_BODY=true, the body is replaced with RESPONSE_BODY (or a default error body matching the status code). Content-Type and Content-Length are updated accordingly. When false, only the status line changes.

  • Only responses selected by TOXICITY are modified; the rest pass through unchanged.

  • Traffic on other ports of the same container is not affected. gRPC responses (which carry status in trailers) are not currently overridden; use Pod API status code for gRPC-style status injection.

  • Clients that branch on status codes execute their corresponding error-handling paths: retry, fail-fast, prompt for auth, back off, and so on.

WHEN THE FAULT ENDS

After TOTAL_CHAOS_DURATION, the proxy is torn down and HTTP status codes return to whatever the application produces. In-flight responses already buffered for modification are released as soon as cleanup runs.

Signals to watch

Attach resilience probes to assert each layer:

  • Status-code distribution: Use a Prometheus probe on the service's HTTP status counter (http_requests_total{status="503"}) to confirm the injected codes match STATUS_CODE.

  • Client retry counter: Use a Prometheus probe on the calling service's retry counter to verify it backs off correctly.

  • End-user error rate: Use an HTTP probe against the top-level API to detect whether the injected errors leak through to end users.

Verify the fault execution effect

While the experiment is running, confirm the status code is being overridden:

  1. Make a request and inspect the status line.

    The status line should equal HTTP/1.1 <STATUS_CODE> ... on at least TOXICITY percent of repeated calls. The body matches RESPONSE_BODY (or the default for the status code) when MODIFY_RESPONSE_BODY=true.

  2. Confirm the proxy is intercepting the right port.

    The proxy listens on PROXY_PORT and reroutes the configured TARGET_SERVICE_PORT through it for the fault's duration.

Recovery and cleanup

  • End of duration: The proxy is removed automatically and HTTP status codes return to baseline.

  • Abort the experiment: Stopping the experiment from Chaos Studio triggers the same cleanup path.

  • Failed cleanup: If automated cleanup did not complete, restart the target pod to reset its network state.

Limitations

  • Serverless Kubernetes (EKS Fargate, ACI virtual nodes): These platforms do not expose container runtime sockets and reject the privileged access the fault needs. GKE Autopilot is supported once the one-time setup in Chaos on GKE Autopilot is in place.

  • Windows containers: This fault is supported on Linux pods only.

  • HTTPS without supplied certificates: This fault does not terminate TLS. If the target serves HTTPS and you need to modify encrypted responses, use Pod API status code, which accepts CA, server, and client certificates as TLS inputs.

  • gRPC status: gRPC carries status in HTTP/2 trailers (grpc-status), not the HTTP status line. Use Pod API status code when you need to override gRPC status codes.

  • Status codes outside the supported set: Only the codes listed under STATUS_CODE are supported. Arbitrary three-digit codes are not accepted.

  • Port already bound: If PROXY_PORT collides with a port the target container is already using, the fault fails to start. Pick a port number outside the application's range.

Troubleshooting

Pod HTTP status code experiment stays Pending or never starts in Harness Chaos Engineering

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

Status code is unchanged during pod-http-status-code

The most common causes are: TARGET_SERVICE_PORT does not match the port the application actually listens on (verify with kubectl exec -- ss -tlnp); the traffic is HTTPS and the chaos proxy is not terminating TLS (use Pod API status code instead); STATUS_CODE is set to a value outside the supported list; or TOXICITY is set lower than expected. Re-run with TOXICITY=100 and a known status code like 503 to confirm the path is working.

Connection to container runtime fails for pod-http-status-code in Harness Chaos Engineering

The default SOCKET_PATH is /run/containerd/containerd.sock. For Docker, set CONTAINER_RUNTIME=docker and SOCKET_PATH=/var/run/docker.sock. For CRI-O, set CONTAINER_RUNTIME=crio and SOCKET_PATH=/var/run/crio/crio.sock.

Status code overrides persist after pod-http-status-code ends

Automated cleanup did not complete. Restart the target pod to reset its network state. If the issue recurs, capture the chaos pod logs from the experiment namespace before the next run and share them with Harness support.

Related faults

Last updated

Was this helpful?