AWS Step Function
Introduction
What is Step Function?
AWS Step Function is a serverless orchestration service that allows integrating multiple AWS services to collate & design an enterprise-critical application or workflow with advance conditional branching and error handling
Step Functions provides sequencing, error handling, retry logic, state, and offers a visualization-driven method to organize and represent the series of event-driven steps of your application through its graphical console.
And makes it easier and simpler to build and execute multiple-state event-driven critical applications. Step Functions can be invoked automatically via events or via APIGateway and simplifies tracking each step, and retries in case of errors, so the application executes in a defined sequence.
What is Amazon State Language (ASL)?
ASL is a JSON-based, structured language leveraged to design the state machine, a collection of states, that can do work (Task states), determine which states to transition to next (Choice states), stop execution with an error (Fail states), and so on.
ASL consist of three things
State Machine Structure
Intrinsic functions
Common State Fields
State Machine Structure: State machines are declared using JSON text and represents a structure consists of the following fields
Comment: description of state machine
StartAt: specify the state name from where the state machine to start the execution, and it is case sensitive
TimeoutSeconds: maximum time in seconds a state machine can execute
Version: specify the version of ASL (default 1.0)
States: a comma-delimited set of states
Note: From the above fields Comment, TimeoutSeconds, Version are optional fields
Intrinsic functions: Intrinsics are constructs like in programming languages, and can be leveraged to manipulate the data going to and from Task Resources
States.Format: string construction from literal and interpolated values, and takes one or more arguments
States.StringToJson: input is a single argument, a reference path to an escaped JSON string
States.JsonToString: input is a single argument, a reference path, and the interpreter returns a string which is a JSON text
States.Array: input is a zero or more argument, and the interpreter returns a JSON array containing the Values of the arguments, in the order provided
Common State Fields
Type: state’s type
Next: name of the next state that is run when the current state finishes
End: name of the state that is run at the end
Comment: human-readable description of the state
InputPath: state’s input to be passed to the state’s task for processing
OutputPath: state’s input to be passed to the state’s output
Note: From the above fields Comment, InputPath, OutputPath are optional fields
What is Workflow Studio?
Workflow Studio for AWS Step Functions is a visual workflow designer to create serverless workflows by orchestrating AWS services.
With easy drag and drop capability you can design, edit, control how input and output is filtered or transformed for each state and configure error handling. It not helps to create a workflow but also validates the ASL and auto-generates code.
Retrieved code can be reviewed, extracted for local development or AWS CloudFormation. Once finished, we can save the workflow, run it, and observe the outcome
Build a workflow using Workflow Studio
In this tutorial, we learn how to work with Workflow Studio for designing the Step Functions.
This tutorial will be broken into three phases
Creating a state machine
Start a new execution
Step 1: Creating a state machine
A state machine is a visual representation of the workflow. We will leverage Workflow Studio to design, configure and examine each step of the workflow.
Traverse to the AWS Step Functions and Click on the Get Started option on the right
Next we will select the authoring method, for this tutorial select Design your workflow visually and type as Standard. Post that click Next
Note: To understand the difference between Standard and Express under Type section expand Help me decide option
Now, we will in the workflow designer studio
i. Click on the Flow panel from the left pane, search for a Pass state, drag it to the empty state and label it as Start state
ii. Search for a Choice state, drag and place it below the Start state, name it as Hello Step Function
iii. Search for a Pass state again, drag and place it under one branch of the Choice state, name it as Pass state
iv. Similarly place one more pass state as a Fail state to the other branch of the Choice state.
v. Now we will add rules in the Choice state to leverage the boolean variable $.Validate. If $.Validate is False, then the flow will be pointed towards the Fail state or it will continue its normal execution flow.
vi. Configure the default for the Choice state to be Yes, and the cause to be Not Hello World for the No Fail state.
vii. Search for a Wait state, drag and place it below the Pass state, name it as Wait State, set the wait time to 5 sec.
viii. This will complete the workflow and it looks as below, post that click Next
ix. Now we will on the review section, under Definition section, review the visual design as well as the ASL code which is auto-generated. Once the review is completed, click Next at the button
x. Now we will provide below details which are required to deploy the state machine
Name: name for the state machine
Permissions: IAM role to execute the state machine
Logging: enable logging level like All, Error, Fatal, and Off
Tracing: enable Xray tracing
Tags: define tags to the state machine
xi. Choose to Create state machine
Step 2: Start a new execution
State machine executions are instances where we run the workflow to perform tasks.
To start the execution, select Start execution from the job detail console
Next we will provide the input i.e name and Input. Post that click Start execution
Note: name and Input both are optional.
Once the execution is started we can verify the output as shown below
Output when Validate flag is True
Output when Validate flag is False
