Maintaining High Availability with Auto Scaling (for Linux)Go to Lab
ran out of time. Need more time.
Starting on step 49, the instructions are missing to direct students to open new PuTTy session into a newly created instance since the previous instance was terminated.
I liked it
It was frustrating to get the autoscaling policies work. There were some fields in CloudWatch screens that were insufficiently documented. CPU utilization is not a very real world thing on which to trigger autoscaling. We should have used a canary (a lambda) to trigger autoscaling when response time becomes unmanageable. You need a real web application to test.