Define AWS Step Functions Concepts with Example

AWS Step Functions were one of the most applauded announcements in 2016 which lets you to coordinate multiple AWS services into serverless workflows to build and update apps quickly. With Step Functions, you may design workflows that stitch together services like AWS Lambda or Amazon ECS into feature-rich applications. Workflows are made up of a series of steps where the output of one step becomes input for the next.

• Step Functions make application development easy and more intuitive by translating workflows into state machine diagrams that are easy to understand and debug.
• It is easy to monitor different steps of execution as it happens that makes identifications and fixing problems quick.
• Step Functions tracks and triggers each step and tries again if there are errors. In this way, applications execute in the same order as expected.

Moving ahead, let us discuss the AWS Step Functions in detail.

What are AWS Step Functions?

Step Functions are the last application service released by the AWS to solve complex problems that are frequently experienced by people. They are mostly used for orchestrating complex workflows using Lambda functions. In use cases, there are multiple processes composed of different tasks. To run the entire application in a serverless way, you should create a Lambda function for each task and run those functions using an orchestrator.

Building code to orchestrate those functions is not an easy task and really hard to debug. AWS step functions remove this problem by using easy design and implementing the complex flow for functions or tasks. In easy words, AWS Steps Functions makes it easy to coordinate components of distributed apps and microservices using visual workflows.

Let us understand the concept in depth with the help of some examples.

Example 1 - Game Hosted on the AWS

A simple use case is handling one of the users who completes a level in your game hosted on the AWS. Here, you should perform many different tasks as given below.

Read: How to Prepare AWS Certified Solutions Architect Associate Certification Exam?

• You should update different DynamoDB tables.
• You should store reports on S3.
• You should put the metric on CloudWatch for further analysis.

Finally, running each of these three Lambda functions is quite tough either sequentially or parallelly. AWS Step Functions lets you run these functions in parallel and handle a different kind of exceptions for each function and process the final output without any complications. This is how to implement the step functions in AWS for the given example.

Example 2- Serverless Handler for Libraries

In most examples, there is a need for human interaction to produce the final output. For instance, a library wants to keep track of each item borrowed by customers and it wants to help customers returning items before the deadline. As soon as, the student checks out a book, the employee puts the action in the system and machine orchestrates all necessary actions ahead for bringing back the book from student to the library.

Thanks to AWS Step Functions here that allows to run a Lambda function and send an email to the customer that confirms the check out with a renewal link. Another Lambda link is created in conjunction with the Amazon API gateway that shows that loan is completed. After a few days, the system machine sends the automated message to inform customers about the renewal or return of the book.

This example is a little more complicated than the first one where a simple game was hosted on the AWS. Here, it is more difficult designing and implementing the step functions. The advantage of using step functions for this example is that you can manage really long tasks and handle human interactions that modify the flow of execution.

Components of AWS Step Functions

How AWS Step Functions are powered? As you can imagine, there are several affinities between this service and SWF (Amazon Simple Workflows) but Step Functions are less complicated and easy to use when compared to other AWS Services. With the depth understanding of its components, you can immediately start using AWS Step Functions for your next project.

1). State Machine

The major component of step functions is the state machine that represents the flow you need to put in place to achieve the final objectives. For example, if you wanted to manage lending resources for a library you should create a state machine first that coordinates each task to give a better experience to customers.

Read: Difference between AWS Inspector and AWS Trusted Advisor

It is easy to create a State Machine with JSON services. Using APIs or AWS console, it is easy starting as many executions as you need. The JSON templates are based on the Amazon States Language. While it is not easy composing a JSON template, AWS console has a real-time graph that shows what to do.

Here are the JSON templates for implementing the State Machines for example number 1 and the number 2 above.

2). State

A State Machine is made of boxes, and each one represents a State. States are referred to by their name inside the State Machine template. Each name must be unique and there are many different State types. Currently, the available states (based on the publish date for this post) are:

A state machine is composed of boxes and each box represents one state. States are referred by their name within the state machine template. Each name should be unique and there are different state types too. Currently, the available states for a state machine are given as:

• Choice State: It is used to branch the execution.
• Fail or Succeed State: It is used to stop the execution with failure or success.
• Pass State: It passes the input to the output and injects some prefixed data too.
• Wait State: It provides the delay for a certain time period or till a specified date.
• Parallel State: It helps to execute tasks in parallel.
• Task State: It executes some code in the state machine.

3). Task

Mostly work in the state machine is accomplished through tasks. A task can be defined as”

• A Lambda Function where you need to specify its ARN.
• An Activity that is defined as the piece of code and hosted wherever you want. It allows to include human tasks in your state machine or those to be hosted in a Lambda function.

How to use the AWS Step Functions?

AWS Step Functions are used in three ways – AWS Console, AWS Cloud Formation, and API/SDK.

Read: What is AWS Lightsail? Amazon EC2 vs. Amazon Lightsail

• AWS Console: AWS offers a really small but a smart set of blueprints to start out. For example, if you need a state machine by a parallel step, you just have to click the related blueprint, change the name and ARN details and your state machine is ready for the production.
• AWS Cloud Formation: AWS Step Functions are easy to implement. You just need to specify state machine template and service role ARN.
• AWS SPI/SDK: The biggest advantage of step functions is that everything is an API and it is applied with SFN. There are lot more APIs available and you can find documentation for each of them as per the requirement.

Here are several great things about AWS step functions:

• AWS Step Functions provide something that is named as state-as-a-service. Usually, it is tough keeping the state of execution up to date. With AWS Step Function, you may orchestrate each of them and run in the right order when needed.
• Another wonderful feature is that you are able to build really long tasks. The maximum duration for the single execution is one year. For long tasks, you can specify date and time too.
• AWS Step functions allow you to implement a retry strategy for different kinds of errors that are incurred by Lambda functions. The concept becomes clear with the help of an example below.
• With AWS Step Functions, it is easy providing a wonderful experience with a lot of blueprints and UI where all the needed information is available.

Final Words:

One big difference between step functions and simple workflows is pricing. AWS step functions are billed for each state transition of your execution. For example, if the state machine has three steps in series, each execution consists of four state transitions. For every account, first 4K transitions are included in the free tier and it will last forever.

The free tier is a wonderful thing to have but for other state transitions, you will be charged if some activity is hosted there. According to expert, AWS Step Functions can be expensive with a lot of execution that incur heavy expenses but it removes the pain of orchestrating different tasks.