CLB AZ down

Last updated on Jun 5, 2026

CLB AZ down is an AWS chaos fault that disables one or more availability zones (ZONES) on a Classic Load Balancer (LOAD_BALANCER_NAME) for a configurable duration, then re-enables them. While a zone is disabled, the load balancer stops routing traffic to back-end instances in that zone, and the corresponding IP addresses are removed from the load balancer's DNS A records.

Use this fault to test how a multi-AZ workload behaves when a single AZ disappears from a Classic Load Balancer: whether DNS-based client failover works within the TTL budget, whether the remaining zones absorb the traffic without saturation, and whether autoscaling responds correctly to the concentrated load.

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:

AZ-level resilience: When one AZ disappears from the CLB, do clients fail over via DNS, and do remaining AZs absorb the traffic without breaching latency SLOs?
Capacity planning: Is the workload provisioned for N-1 AZ availability, or do the remaining instances saturate?
DNS TTL behaviour: Do clients with stale DNS cache continue to attempt the removed AZ until the TTL expires, and does the application handle the resulting failures?
Cross-zone load balancing: When cross-zone is enabled, do instances in the remaining AZs serve traffic that previously hit the removed zone?
Recovery and rebalance: When the AZ returns, do instances re-register cleanly and resume serving?

Prerequisites

Kubernetes version: 1.21 or later for the chaos infrastructure cluster. Go to What's supported to confirm distribution support.
Target CLB: LOAD_BALANCER_NAME exists in REGION and is configured for at least two AZs. The set ZONES must be a subset of the CLB's current AZ list (leaving at least one AZ enabled).
Instance health in other AZs: Instances in the remaining AZs are healthy and provisioned to absorb the additional traffic.
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
Classic Load Balancer (CLB)	Supported
Internet-facing or internal CLB	Supported
CLB in EC2-Classic	Not supported (deprecated by AWS)
Application Load Balancer (ALB)	Not supported (use ALB AZ down)
Network Load Balancer (NLB)	Not supported (use NLB AZ down)
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.

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "elasticloadbalancing:DescribeLoadBalancers",
        "elasticloadbalancing:DisableAvailabilityZonesForLoadBalancer",
        "elasticloadbalancing:EnableAvailabilityZonesForLoadBalancer",
        "elasticloadbalancing:DetachLoadBalancerFromSubnets",
        "elasticloadbalancing:AttachLoadBalancerToSubnets",
        "ec2:DescribeSubnets",
        "ec2:DescribeAvailabilityZones"
      ],
      "Resource": "*"
    }
  ]
}

DescribeLoadBalancers reads the CLB's current AZ list.
DisableAvailabilityZonesForLoadBalancer / EnableAvailabilityZonesForLoadBalancer apply to EC2-Classic CLBs.
DetachLoadBalancerFromSubnets / AttachLoadBalancerToSubnets apply to VPC CLBs.
ec2:DescribeSubnets and ec2:DescribeAvailabilityZones resolve ZONES to the matching subnet IDs.

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:

[default]
aws_access_key_id = REPLACE_WITH_ACCESS_KEY_ID
aws_secret_access_key = REPLACE_WITH_SECRET_ACCESS_KEY

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.

Go to AWS named profile for chaos to switch between profiles inside a single credentials file.

Fault tunables

Configure the following fault parameters when you add CLB AZ down to an experiment in Chaos Studio. Defaults are shown for reference.

Required parameters

Tunable	Description	Default
`LOAD_BALANCER_NAME`	Name of the target Classic Load Balancer.	(required)
`REGION`	AWS region that hosts the CLB (for example `us-east-1`).	(required)
`ZONES`	Comma-separated list of availability zone names to disable (for example `us-east-1a,us-east-1b`). The list must leave at least one AZ enabled on the CLB.	(required)

Chaos parameters

Tunable	Description	Default
`TOTAL_CHAOS_DURATION`	Duration of the fault in seconds. The AZs stay disabled for this period.	`60`
`CHAOS_INTERVAL`	Delay in seconds between successive iterations when running for more than one cycle.	`30`
`SEQUENCE`	Order in which multiple zones are disabled: `parallel` disables them all at once; `serial` disables them one at a time.	`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. AWS-specific shared tunables are documented in common AWS fault tunables.

Fault execution in brief

Reads the current AZ/subnet mapping of LOAD_BALANCER_NAME, calls the disable/detach API to remove the AZs in ZONES, waits for TOTAL_CHAOS_DURATION seconds, then calls the enable/attach API to restore the original AZ list.

Expected behavior during fault execution

The IP addresses of the disabled AZs are removed from the CLB's DNS A records. New DNS resolutions return only the surviving AZ IPs.
Existing clients with stale DNS cache continue to connect to the removed AZ IPs until the TTL expires; those connections fail or time out.
Back-end instances in the disabled AZs are removed from rotation. When cross-zone load balancing is enabled, the remaining AZs absorb the traffic.
Latency and throughput on the surviving AZs rise as they absorb the redirected traffic.
CloudWatch metrics (HealthyHostCount, RequestCount) reflect the AZ removal.

When the fault ends

The chaos pod restores the original AZ list. DNS records update within seconds and instances re-register.

Signals to watch

Attach resilience probes to assert each layer:

DNS resolution: Use a command probe that runs dig +short <clb-dns-name> and asserts that the number of A records drops by the number of disabled AZs.
Application availability: Use an HTTP probe and assert success rate stays above SLO.
Healthy host count: Use a Prometheus probe on the CLB's HealthyHostCount metric to confirm the surviving AZs continue to serve.
CPU saturation on surviving AZs: Use a Prometheus probe on aws_ec2_cpuutilization of the back-end instances in the surviving AZs.

Verify the fault execution effect

While the experiment is running, confirm the AZ is disabled and then re-enabled:

Inspect the CLB's current AZ list.

aws elb describe-load-balancers \
  --region <region> \
  --load-balancer-names <name> \
  --query "LoadBalancerDescriptions[].AvailabilityZones"

During the fault the listed zones should not include ZONES. After recovery the original list should be back.

Check DNS resolution.
```
dig +short <clb-dns-name>
```
Number of A records should drop during the chaos window.
Inspect instance health.
```
aws elb describe-instance-health \
  --region <region> \
  --load-balancer-name <name>
```
Instances in the disabled AZs should be removed from the listing during the fault.

Recovery and cleanup

End of duration: The chaos pod restores the original AZ list.
Abort the experiment: Stopping the experiment from Chaos Studio also triggers the restore call.
Manual recovery: If the fault exits before the restore call, run aws elb enable-availability-zones-for-load-balancer (for EC2-Classic) or aws elb attach-load-balancer-to-subnets (for VPC) with the original AZ/subnet list (recorded in the chaos pod logs).
Workload recovery: Instances re-register quickly; client DNS caches may take up to the configured TTL to pick up the restored IPs.

Limitations

At least one AZ must remain enabled: The fault refuses to disable every AZ (the CLB requires at least one AZ).
DNS TTL is honoured by clients: Even after the fault ends, clients with a stale DNS cache continue to attempt to reach the removed AZ until the TTL expires.
EC2-Classic CLB is deprecated: This fault works against VPC CLBs only.
Connection draining: Existing connections to instances in the removed AZ are not actively terminated; they continue until the client closes them or the connection idle timeout elapses.
Cross-region targeting: A single experiment targets one region (the value of REGION).

Troubleshooting

CLB AZ down fails with AccessDeniedException in Harness Chaos Engineering

The credentials supplied to the chaos pod do not have the required ELB or EC2 permissions. Confirm the IAM policy attached to the user, role, or IRSA service account includes elasticloadbalancing:DescribeLoadBalancers, elasticloadbalancing:DetachLoadBalancerFromSubnets, elasticloadbalancing:AttachLoadBalancerToSubnets, ec2:DescribeSubnets, and ec2:DescribeAvailabilityZones.

CLB AZ down fails because the resulting AZ list is empty

A Classic Load Balancer requires at least one AZ enabled. If the values in ZONES would leave the CLB with no AZ, the API call fails. Reduce ZONES to a subset that leaves at least one AZ enabled, or split the experiment into multiple iterations.

Clients keep hitting the disabled AZ for several minutes after the fault starts

Clients respect the DNS TTL of the CLB A records (default 60 seconds). Some applications cache DNS for longer than the TTL because of their HTTP client or JVM behaviour. Verify the cache duration in the client. CLB does not actively close existing connections when an AZ is disabled.

AZ not restored after the chaos window

If the enable/attach call failed (for example because a permission was revoked mid-experiment), the CLB is left without the original AZs. Run 'aws elb attach-load-balancer-to-subnets --load-balancer-name <name> --subnets <original-subnet-ids>' (or the EC2-Classic enable equivalent) with the original IDs (recorded in the chaos pod logs).

ALB AZ down: Detach AZs from an Application Load Balancer.
NLB AZ down: Detach AZs from a Network Load Balancer.
AZ blackhole: Blackhole all VPC traffic to an AZ instead of just LB rotation.
Common AWS fault tunables: Shared environment variables for AWS faults.

Use cases​

Prerequisites​

Supported environments​

Permissions required​

Authentication​

Fault tunables​

Fault execution in brief​

Expected behavior during fault execution​

Signals to watch​

Verify the fault execution effect​

Recovery and cleanup​

Limitations​

Troubleshooting​

Related faults​