Auto Scaling in AWS cloud

Auto scaling service is one of the most important features introduced by Amazon Web services to its cloud portfolio and still now it has been one of the most misunderstood features of AWS. I had helped many companies in the past to move their web apps into AWS with Auto Scaling capability and i have shared some of my experiences as insights in this article and the embedded slides.


Auto Scaling is not suitable for every application
Many customers have a misconception that by simply deploying their application in AWS makes it an automatic candidate for scaling, whereas they fail to understand that if we need to leverage Auto Scaling effectively our application’s technical components like Session data, Logs, In-memory Cache, File uploads/downloads etc should be properly designed to run on a dynamically scaled out /down server instances. Applications which are not architected keeping such technical fundamentals in mind will surely have problems when deployed on Auto Scaling mode.


Auto Scaling is an art
In the blogosphere we read articles about how easy and simple it is to configure AWS AutoScaling, but in reality Auto Scaling is more than configuration. Customer IT teams need to have in-depth and constant understanding of their Scaling needs , Load volatility patterns and AWS infrastructure expertise to closely align their dynamic infrastructure with their requirements. A badly configured Auto Scaling application can cause leakages and increase the cost of dynamic infrastructure rather than reducing them. In some cases leveraging AWS AutoScaling is not a one time configuration, but an ongoing activity with close monitoring of load trends and periodical tuning of the Auto Scaling parameters to keep the infrastructure cost and performance at optimum.


Dependencies with Other AWS technologies
Most of the time when customers migrate their Application to AWS infrastructure, they need to use an array of AWS technologies like S3, ELB, EBS, Regions/AZ, VPC, CloudWatch etc in addition to Amazon EC2 Auto Scaling in their system. These dependencies pose the need for holistic AWS expertise and in-depth understanding of intricacies in AWS and how well these technologies behave in coordination with AWS Auto Scaling services. Customer IT teams which satisfies the above needs will have better success ratio in migrating and sustaining their applications in AWS.


In the embedded slides I have explored AWS AutoScaling in following perspectives

  • Need for Auto Scaling
  • How AWS Auto scaling can help us?
  • How to configure an Auto scaling policy in AWS
  • Things to remember before Scaling out and down
  • Understand the Intricacies while integrating Auto scaling with other Amazon Web Services
  • Risks involved in AWS Auto Scaling

Big thanks to this blog author

Latest blog

Web Analytics