Reputation: 29
I am a DevOps engineer in a company and we want to build a solution for our developers where they can specify in a common calendar that from a certain time to a certain time they want a particular EC2 instance to be up and running and using that schedule EC2 instances can start and stop on their own. Also if there is some overlap of times for the same instance then it is kept running until all the developers have stopped using it. This would be extremely helpful for optimizing the usage by making sure that some instance is not kept accidentally running even when it is not required. Is there any tool in AWS to do the same i.e. automating the start/stop process based on calendar event/schedule?
I came across few solutions : https://zapier.com/zapbook/amazon-ec2/google-calendar/ Integration of Google Calendar with amazon EC2
One of them is a third party solution(Zapier) and another one is something that we have to develop on our own by working with the Calender APIs and then polling them continuously to determine if a particular instance needs to start/stop. Is there any other way of doing the same, specially if there is some solution which AWS only provides?
Hope somebody is able to help me.
Thanks!
Upvotes: 1
Views: 385
Reputation: 5012
If you want an AWS-only solution, you can create an auto-scaling group that is based on a schedule (M-F 9AM - 6PM), but it won't have Google Calendar integration.
Your auto-scaling group can also use Spot Instances if you'd like to save money.
Upvotes: 2