 |
At present, I’m excited to announce enhanced native testing capabilities for AWS Step Capabilities by the TestState API, our testing API.
These enhancements can be found by the API, so you possibly can construct automated check suites that validate your workflow definitions regionally in your improvement machines, check error dealing with patterns, information transformations, and mock service integrations utilizing your most popular testing frameworks. This launch introduces an API-based method for native unit testing, offering programmatic entry to complete testing capabilities with out deploying to Amazon Net Companies (AWS).
There are three key capabilities launched on this enhanced TestState API:
Mocking assist – Mock state outputs and errors with out invoking downstream providers, enabling true unit testing of state machine logic. TestState validates mocked responses in opposition to AWS API fashions with three validation modes: STRICT (that is the default and validates all required fields), PRESENT (validates discipline sorts and names), and NONE (no validation), offering high-fidelity testing.
Assist for all state sorts – All state sorts, together with superior states equivalent to Map states (inline and distributed), Parallel states, activity-based Job states, .sync service integration patterns, and .waitForTaskToken service integration patterns, can now be examined. This implies you should use TestState API throughout your complete workflow definition and write unit exams to confirm management circulate logic, together with state transitions, error dealing with, and information transformations.
Testing particular person states – Take a look at particular states inside a full state machine definition utilizing the brand new stateName parameter. You possibly can present the entire state machine definition one time and check every state individually by title. You possibly can management execution context to check particular retry makes an attempt, Map iteration positions, and error eventualities.
Getting began with enhanced TestState
Let me stroll you thru these new capabilities in enhanced TestState.
Situation 1: Mock profitable outcomes
The primary functionality is mocking assist, which you should use to check your workflow logic with out invoking precise AWS providers and even exterior HTTP requests. You possibly can both mock service responses for quick unit testing or check with precise AWS providers for integration testing. When utilizing mocked responses, you don’t want AWS Id and Entry Administration (IAM) permissions.
Right here’s find out how to mock a profitable AWS Lambda perform response:
aws stepfunctions test-state --region us-east-1
--definition '{
"Kind": "Job",
"Useful resource": "arn:aws:states:::lambda:invoke",
"Parameters": {"FunctionName": "process-order"},
"Finish": true
}'
--mock '{"consequence":"{"orderId":"12345","standing":"processed"}"}'
--inspection-level DEBUG
This command exams a Lambda invocation state with out truly calling the perform. TestState validates your mock response in opposition to the Lambda service API mannequin so your check information matches what the actual service would return.
The response exhibits the profitable execution with detailed inspection information (when utilizing DEBUG inspection stage):
{
"output": "{"orderId":"12345","standing":"processed"}",
"inspectionData": {
"enter": "{}",
"afterInputPath": "{}",
"afterParameters": "{"FunctionName":"process-order"}",
"consequence": "{"orderId":"12345","standing":"processed"}",
"afterResultSelector": "{"orderId":"12345","standing":"processed"}",
"afterResultPath": "{"orderId":"12345","standing":"processed"}"
},
"standing": "SUCCEEDED"
}
Once you specify a mock response, TestState validates it in opposition to the AWS service’s API mannequin so your mocked information conforms to the anticipated schema, sustaining high-fidelity testing with out requiring precise AWS service calls.
Situation 2: Mock error situations
You can even mock error situations to check your error dealing with logic:
aws stepfunctions test-state --region us-east-1
--definition '{
"Kind": "Job",
"Useful resource": "arn:aws:states:::lambda:invoke",
"Parameters": {"FunctionName": "process-order"},
"Finish": true
}'
--mock '{"errorOutput":{"error":"Lambda.ServiceException","trigger":"Operate failed"}}'
--inspection-level DEBUG
This simulates a Lambda service exception so you possibly can confirm how your state machine handles failures with out triggering precise errors in your AWS atmosphere.
The response exhibits the failed execution with error particulars:
{
"error": "Lambda.ServiceException",
"trigger": "Operate failed",
"inspectionData": {
"enter": "{}",
"afterInputPath": "{}",
"afterParameters": "{"FunctionName":"process-order"}"
},
"standing": "FAILED"
}
Situation 3: Take a look at Map states
The second functionality provides assist for beforehand unsupported state sorts. Right here’s find out how to check a Distributed Map state:
aws stepfunctions test-state --region us-east-1
--definition '{
"Kind": "Map",
"ItemProcessor": {
"ProcessorConfig": {"Mode": "DISTRIBUTED", "ExecutionType": "STANDARD"},
"StartAt": "ProcessItem",
"States": {
"ProcessItem": {
"Kind": "Job",
"Useful resource": "arn:aws:states:::lambda:invoke",
"Parameters": {"FunctionName": "process-item"},
"Finish": true
}
}
},
"Finish": true
}'
--input '[{"itemId":1},{"itemId":2}]'
--mock '{"consequence":"[{"itemId":1,"status":"processed"},{"itemId":2,"status":"processed"}]"}'
--inspection-level DEBUG
The mock consequence represents the entire output from processing a number of gadgets. On this case, the mocked array should match the anticipated Map state output format.
The response exhibits profitable processing of the array enter:
{
"output": "[{"itemId":1,"status":"processed"},{"itemId":2,"status":"processed"}]",
"inspectionData": {
"enter": "[{"itemId":1},{"itemId":2}]",
"afterInputPath": "[{"itemId":1},{"itemId":2}]",
"afterResultSelector": "[{"itemId":1,"status":"processed"},{"itemId":2,"status":"processed"}]",
"afterResultPath": "[{"itemId":1,"status":"processed"},{"itemId":2,"status":"processed"}]"
},
"standing": "SUCCEEDED"
}
Situation 4: Take a look at Parallel states
Equally, you possibly can check Parallel states that execute a number of branches concurrently:
aws stepfunctions test-state --region us-east-1
--definition '{
"Kind": "Parallel",
"Branches": [
{"StartAt": "Branch1", "States": {"Branch1": {"Type": "Pass", "End": true}}},
{"StartAt": "Branch2", "States": {"Branch2": {"Type": "Pass", "End": true}}}
],
"Finish": true
}'
--mock '{"consequence":"[{"branch1":"data1"},{"branch2":"data2"}]"}'
--inspection-level DEBUG
The mock consequence have to be an array with one factor per department. Through the use of TestState, your mock information construction matches what an actual Parallel state execution would produce.
The response exhibits the parallel execution outcomes:
{
"output": "[{"branch1":"data1"},{"branch2":"data2"}]",
"inspectionData": {
"enter": "{}",
"afterResultSelector": "[{"branch1":"data1"},{"branch2":"data2"}]",
"afterResultPath": "[{"branch1":"data1"},{"branch2":"data2"}]"
},
"standing": "SUCCEEDED"
}
Situation 5: Take a look at particular person states inside full workflows
You possibly can check particular states inside a full state machine definition utilizing the stateName parameter. Right here’s an instance testing a single state, although you’ll sometimes present your full workflow definition and specify which state to check:
aws stepfunctions test-state --region us-east-1
--definition '{
"Kind": "Job",
"Useful resource": "arn:aws:states:::lambda:invoke",
"Parameters": {"FunctionName": "validate-order"},
"Finish": true
}'
--input '{"orderId":"12345","quantity":99.99}'
--mock '{"consequence":"{"orderId":"12345","validated":true}"}'
--inspection-level DEBUG
This exams a Lambda invocation state with particular enter information, exhibiting how TestState processes the enter and transforms it by the state execution.
The response exhibits detailed enter processing and validation:
{
"output": "{"orderId":"12345","validated":true}",
"inspectionData": {
"enter": "{"orderId":"12345","quantity":99.99}",
"afterInputPath": "{"orderId":"12345","quantity":99.99}",
"afterParameters": "{"FunctionName":"validate-order"}",
"consequence": "{"orderId":"12345","validated":true}",
"afterResultSelector": "{"orderId":"12345","validated":true}",
"afterResultPath": "{"orderId":"12345","validated":true}"
},
"standing": "SUCCEEDED"
}
These enhancements deliver the acquainted native improvement expertise to Step Capabilities workflows, serving to me to get prompt suggestions on modifications earlier than deploying to my AWS account. I can write automated check suites to validate all Step Capabilities options with the identical reliability as cloud execution, offering confidence that my workflows will work as anticipated when deployed.
Issues to know
Listed here are key factors to notice:
- Availability – Enhanced TestState capabilities can be found in all AWS Areas the place Step Capabilities is supported.
- Pricing – TestState API calls are included with AWS Step Capabilities at no extra cost.
- Framework compatibility – TestState works with any testing framework that may make HTTP requests, together with Jest, pytest, JUnit, and others. You possibly can write check suites that validate your workflows routinely in your steady integration and steady supply (CI/CD) pipeline earlier than deployment.
- Characteristic assist – Enhanced TestState helps all Step Capabilities options together with Distributed Map, Parallel states, error dealing with, and JSONata expressions.
- Documentation – For detailed choices for various configurations, consult with the TestState documentation and API reference for the up to date request and response mannequin.
Get began immediately with enhanced native testing by integrating TestState into your improvement workflow.
Joyful constructing!
— Donnie