- Replicate your infrastructure across two regions. You will harden the application to a regional failure and you will double your capacity.
- Take an AMI image of a front-end server to save your configuration and then add more servers to your cluster pror to the conference. Remove the servers from the cluster after the spike from the conference.
- Test to determine your throughput and how many users you can support. Develop a scaling plan for your front end, microservices, and database based on CloudWatch metrics that align with the tested bottlenecks.
- Use Auto Scaling groups to create more front-end servers when the CloudWatch metrics for CPU usage on a single instance rise above 80% for five minutes.
Subscribe to our mailing list and get interesting stuff and updates to your email inbox.