Chaos testing without being Netflix: Resilience Hub's 4 ready-made tests

AWS update: Aug 2026 · By Anmol Handa · The reel is 30 seconds. This is the long version.

Chaos engineering ke liye Netflix hona zaroori nahi. AWS Resilience Hub now gives you pre-built resilience tests, so your first failure experiment doesn't start with a blank page.

What's new

On August 3, 2026, AWS announced recommended resilience tests in the next generation of AWS Resilience Hub. You get four pre-configured test templates. Under the hood they use AWS Fault Injection Service (FIS) to inject controlled faults. Each test targets your service's resources automatically, checks your alarms and recovery objectives, returns a pass or fail result and generates a report.

Test template	What it validates
Availability Zone: recovery	Your service can detect and recover from an Availability Zone impairment.
Dependency validation	How your service responds when its dependencies within the Region are impaired.
Multi-Region: isolation	Your service can keep operating when connectivity to another Region is unavailable.
Multi-Region: recovery	Your service can recover and serve clients from another Region within your recovery objectives when a Region is impaired.

At launch it's available in 15 Regions, including US East (N. Virginia and Ohio), US West (Oregon), Europe (Ireland, London, Frankfurt, Paris, Stockholm), Asia Pacific (Mumbai, Singapore, Sydney, Tokyo, Seoul), Canada (Central) and South America (São Paulo).

Why it matters

The hardest part of a first chaos experiment usually isn't the fault. It's designing the experiment: which resources to hit, what "healthy" means, and how you'll know you passed. These templates hand you that structure, tied to the recovery objectives you already set.

And a pass/fail result based on your alarms answers the question that matters: when an AZ goes bad, does anyone get paged, and does the service actually recover?

Seedhi baatPehla chaos test aaj, bina scratch se experiment likhe. Bas staging se shuru karo.

Steps

1. List the templates in your Region

Use a recent AWS CLI v2 and a Region from the list above.

aws resiliencehubv2 list-test-templates --region us-east-1

2. Inspect one template before you run anything

Take a template ARN from the list. This one is the AZ recovery example from the AWS docs; swap in your Region.

aws resiliencehubv2 get-test-template \
  --test-template-arn "arn:aws:resiliencehub:us-east-1:aws:test-template/aws-az-recovery:rtaz001" \
  --region us-east-1

3. Read the actions and parameters

The response lists the template's actions (each with an actionId, description and resourceType) and its parameters (name, description, type, required, defaultValue, maxValues). This tells you exactly what will be disrupted and what you have to supply.

4. Create the test with your parameter values

You pass values with the --parameters option when you create a test. Follow the full example in AWS's getting started guide rather than guessing the shape. The console works too.

Caveats

These are real faults, injected by FIS. A template is pre-configured, not risk-free.
Pass/fail is only as good as your alarms and recovery objectives. If your alarms are weak, a "pass" can be misleading.
This is the next generation of Resilience Hub and only in the launch Regions listed above.
FIS usage is billed. Check pricing before you schedule tests on repeat.
Multi-Region templates assume you actually have a second Region set up. Start with AZ recovery.

Try it safely

Run it in a staging or sandbox account first, never prod on day one.
Start with Availability Zone: recovery, the smallest blast radius of the four.
Write your hypothesis down: "We lose one AZ, the alarm fires within N minutes, and errors recover within our RTO."
Tell on-call and the service owner the time window. Pick a quiet hour.
Know how to stop the experiment before you start it.
Afterwards, fix the top finding and run the same test again.