Chapter 15. Automating infrastructure management


This chapter covers

  • Using CloudWatch alarms to trigger Lambda functions to solve infrastructure issues
  • Using Amazon SNS as a trigger for AWS Lambda
  • Using CloudWatch events and Lambda functions to synchronize a DNS or a service discovery tool
  • Using Lambda functions to process CloudWatch logs
  • Using CloudWatch events to schedule management activities
  • Designing a multiregion architecture using Amazon API Gateway, Lambda functions, and DynamoDB tables

In the previous chapter, you learned how to implement automatic deployment using Amazon S3 as the trigger of Lambda deploying functions, or using AWS CloudFormation to manage Lambda functions via a text file.

Now you’ll apply what you learned in this book to a different area: infrastructure management. With a pure serverless architecture, you don’t have much infrastructure to manage. You may still have to configure throughput provisioning to a DynamoDB table or shards for a Kinesis stream. But it’s likely you’ll have components that use virtual servers or load balancers that need to be managed.

Using AWS Lambda, you can automate how you manage alarms or process logs—steps that normally require human interactions. The idea here is to have a smarter architecture that can recover automatically in case of application or infrastructure errors.

15.1. Reacting to alarms

15.2. Reacting to events

15.3. Processing logs in near real-time

15.4. Scheduling recurring activities

15.5. Multiregion architectures and data synchronization