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

Lambda delete function concurrency

Delete the reserved concurrency configuration on an AWS Lambda function for a configurable duration and restore it afterwards so you can test how the workload behaves when the function has to share ac

Lambda delete function concurrency is an AWS chaos fault that deletes the reserved concurrency configuration on a target Lambda function (FUNCTION_NAME in REGION, version FUNCTION_VERSION) for TOTAL_CHAOS_DURATION seconds, then restores it. While the reserved concurrency is removed, the function shares the account-level unreserved concurrency pool with every other function in the same region.

Use this fault to test how a critical Lambda workload behaves when its concurrency guarantee disappears: whether the function is throttled by competing workloads, whether monitoring detects the throttling, and whether downstream consumers degrade gracefully under reduced throughput.

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:

  • Throttling exposure: When the reservation disappears, does the function start receiving TooManyRequestsException from competing workloads in the same account?

  • Downstream impact: Do consumers downstream of the Lambda function degrade gracefully when throughput drops?

  • Monitoring fidelity: Do CloudWatch alarms on Throttles fire within the SLA?

  • Capacity planning: Does the rest of the account have enough unreserved concurrency to absorb this function's traffic?

  • Recovery: When the reservation is restored, does the function return to its previous throughput?

Prerequisites

  • Kubernetes version: 1.21 or later for the chaos infrastructure cluster. Go to What's supported to confirm distribution support.

  • Target Lambda function: FUNCTION_NAME exists in REGION and has a non-zero reserved concurrency configured.

  • AWS credentials available: Either an AWS credentials file uploaded as a File Secret in Harness Secret Manager (see Authentication below) or an IAM role for service accounts (IRSA) bound to the chaos infrastructure service account.

  • IAM permissions granted: The credentials or role include the permissions listed below.

Supported environments

Platform
Support status

AWS Lambda (zip and container image)

Supported

Lambda functions with reserved concurrency configured

Supported

Lambda functions without reserved concurrency

Not applicable (nothing to delete)

AWS regions

Supported in every commercial region; pass the region in REGION

Permissions required

The IAM principal that the chaos pod uses (the credentials mounted from the Harness Secret Manager file secret, the IRSA role on the chaos service account, or the role assumed via ASSUME_ROLE_ARN) needs the following AWS actions.

Go to common policy for all AWS faults to use a single superset IAM policy across every AWS fault.

Authentication

The fault supports three credential delivery models. Pick one based on how your chaos infrastructure is deployed.

Method
When to use it
How to configure

Harness Secret Manager file secret

Chaos infrastructure runs outside EKS, or you want explicit static credentials

Upload the AWS credentials file as a File Secret in Harness Secret Manager and reference its identifier via AWS_AUTHENTICATION_SECRET

IAM Roles for Service Accounts (IRSA)

Chaos infrastructure runs in EKS and uses an OIDC-bound service account

No tunable changes; the chaos pod inherits the role automatically. Go to AWS IAM integration to set it up

Assume role

The fault needs to act in a different account or with elevated permissions

Set ASSUME_ROLE_ARN to the role ARN; the chaos pod assumes the role on top of its base credentials

When using the Harness Secret Manager method, the contents of the File Secret should be the AWS credentials file in the standard ~/.aws/credentials format:

Upload this file as a File Secret in Harness Secret Manager (Project Setup → Secrets → New File Secret), and pass the secret identifier in AWS_AUTHENTICATION_SECRET when configuring the fault.

Fault tunables

Configure the following fault parameters when you add Lambda delete function concurrency to an experiment in Chaos Studio. Defaults are shown for reference.

Required parameters

Tunable
Description
Default

FUNCTION_NAME

Name of the target Lambda function.

(required)

REGION

AWS region where the Lambda function is deployed.

(required)

FUNCTION_VERSION

Version of the Lambda function to target. Use $LATEST for the latest version.

$LATEST

Chaos parameters

Tunable
Description
Default

TOTAL_CHAOS_DURATION

Duration of the fault in seconds.

30

CHAOS_INTERVAL

Delay in seconds between successive iterations when running for more than one cycle.

30

SEQUENCE

Order in which multiple iterations are executed: parallel or serial.

parallel

RAMP_TIME

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

0

Authentication

Tunable
Description
Default

ASSUME_ROLE_ARN

ARN of an IAM role to assume on top of the base credentials. Leave empty to use the base credentials directly.

""

AWS_AUTHENTICATION_SECRET

Identifier of the File Secret in Harness Secret Manager that contains the AWS credentials file. Not required when using IRSA.

""

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

Fault execution in brief

Reads the current reserved concurrency on FUNCTION_NAME in REGION, deletes it, waits for TOTAL_CHAOS_DURATION seconds, then restores the original reserved concurrency value.

Expected behavior during fault execution

  • The function loses its reserved concurrency and shares the account-level unreserved pool with every other function in the same region.

  • Under sustained load (or when other workloads are also active), the function may be throttled with TooManyRequestsException.

  • CloudWatch metrics Throttles and Errors may rise; ConcurrentExecutions is no longer guaranteed.

  • Synchronous invokers (API Gateway, ALB) surface throttling errors directly; asynchronous invokers retry per their policy.

WHEN THE FAULT ENDS

The chaos pod restores the original reserved concurrency value. The function regains its concurrency guarantee.

Signals to watch

Attach resilience probes to assert each layer:

  • Throttles: Use a Prometheus probe on aws_lambda_throttles_sum.

  • Concurrent executions: Use a Prometheus probe on aws_lambda_concurrent_executions_maximum.

  • Invocation errors: Use a Prometheus probe on aws_lambda_errors_sum.

  • End-to-end availability: Use an HTTP probe on the user-visible endpoint backed by the function.

Verify the fault execution effect

While the experiment is running, confirm the concurrency was deleted and then restored:

  1. Check function concurrency.

    During the fault this returns no ReservedConcurrentExecutions; after recovery it returns the original value.

  2. Check throttles.

    Throttles may rise if the function is under load and the account-level pool is contested.

Recovery and cleanup

  • End of duration: The chaos pod restores the original reserved concurrency value.

  • Abort the experiment: Stopping the experiment from Chaos Studio also triggers the restore call.

  • Manual recovery: If the fault exits before restore, run aws lambda put-function-concurrency --function-name <name> --reserved-concurrent-executions <N> --region <region> with the original value (recorded in the chaos pod logs).

  • Workload recovery: Throttling stops as soon as the reservation is restored; in-flight retries by asynchronous invokers continue per their retry policy.

Limitations

  • Account-level pool dependency: The visible impact depends on how loaded the account-level pool is. On a quiet account the deletion may produce no measurable effect.

  • Synchronous vs asynchronous: Synchronous invokers see throttling immediately; asynchronous invokers retry per their policy and may not surface visible failures.

  • Provisioned concurrency unaffected: Provisioned concurrency is a separate configuration and is not deleted by this fault.

  • Cross-region targeting: A single experiment targets one region (the value of REGION).

Troubleshooting

Lambda delete function concurrency fails with AccessDeniedException in Harness Chaos Engineering

The credentials supplied to the chaos pod do not have the required Lambda permissions. Confirm the IAM policy attached to the user, role, or IRSA service account includes lambda:GetFunctionConcurrency, lambda:PutFunctionConcurrency, and lambda:DeleteFunctionConcurrency.

No throttling observed during the chaos window

If the account-level unreserved concurrency pool is large enough to absorb the function's load (or no other functions in the region are running), the deletion has no visible effect. Run the experiment when other workloads are active, or run a load generator against the function to drive concurrency above the unreserved pool size.

Concurrency not restored after the chaos window

If the restore call failed, the function may be left without reserved concurrency. Run 'aws lambda put-function-concurrency --function-name --reserved-concurrent-executions --region ' with the original value (recorded in the chaos pod logs).

The function did not have reserved concurrency to begin with

Lambda delete function concurrency assumes the function has reserved concurrency configured. If GetFunctionConcurrency returns no value, set a reservation with 'aws lambda put-function-concurrency --function-name --reserved-concurrent-executions ' before running the experiment.

Related faults

Last updated

Was this helpful?