Kit Wai Chan - Fotolia
Making an AWS Lambda function and microservices work together
Our developers want to switch to microservices to improve scalability and ease upgrades to our AWS-based applications. How does AWS Lambda support this type of architecture?
Enterprise application design is changing. The huge, complex, monolithic applications of years past are being replaced by applications that are assembled from a series of individual components called microservices. Applications based on microservices divide the greater application's functionality into discrete functions that are easier to combine and scale to meet changing user demands. As more enterprise applications deploy apps to the public cloud, providers like AWS design services that complement microservices architectures.
One of the core ideas of microservices is the "division or labor," which includes code segments developed and deployed in VMs and containers. But Public cloud providers like AWS are aggressively creating services that can facilitate microservices without VMs or containers.
AWS Lambda, for example, manages and runs code from other services in response to events. For example, changes made to Amazon S3 objects, database updates in Amazon DynamoDB and messages from Amazon Simple Notification Service and Amazon Kinesis Stream can all trigger an AWS Lambda function.
Events that occur from a Web-based or mobile device can also trigger an AWS Lambda function. For example, if a component of your microservices application requires more compute resources, a low-resource event could trigger Lambda to provision another instance of that component and apply load balancing. Similarly, Lambda could be used to service things like HTTP requests -- such as website clicks -- rather than creating a separate component to handle that activity.
AWS Lambda only runs when an event is triggered, unlike VMs or containers, which are always running and using resources until powered down or retired. Businesses pay only for the compute time that each AWS Lambda function actually uses -- in 100 millisecond increments. Functions built in Lambda are highly scalable and can be extremely cost-effective for low traffic tasks.
Developers can update or change the code for an AWS Lambda function at any time without affecting the operation of other related application components or Amazon services.
