3 Actionable Steps To Save Money on AWS
We all love using AWS for the breadth and depth of their cloud services. Combining these services with the ability to interact with virtually everything through their API’s, it has never been easier to deploy new services and applications.
Developers and infrastructure engineers often have a field day with their new found productivity and flexibility… all until they receive a tap on the shoulder from their red faced boss telling them they have gone way over budget.
We’ve put together a handy list of some easy wins for quick AWS Savings:
1. Power cycle instances and autoscale groups
Power cycling instances and autoscale groups is one of the safest ways to save money in AWS. If you have development and staging environments, they usually aren’t required 24/7 — so you should be able to switch them off when everybody packs up and goes home, effectively your servers sleep when your team does.
The most practical way to achieve this, is to tag your resources according to their role or environment and then run a scheduled job which turns off all at your dev and staging instances at say 8PM; and then schedule another job to fire them back up at 7AM in the morning so they are ready to go when they early birds come in.
With autoscale groups, it is a similar process but instead of turning your servers off, you set your minimum, maximum and desired instances to zero for the period you want to switch the servers off and then change it back when you want them to come back on.
2. Automatically terminate rogue, incorrectly tagged resources
The following is a great way to save money and ensure that your teammates tag resources correctly. Tagging is a key element of making your AWS account more manageable and also makes it much easier to understand where your dollars are being spent.
To achieve this, just run a scheduled job that will terminate resources in your AWS account which are missing the tags that you have defined as required, for example ‘Name of Project’.
Puppet labs have released an open source Python library that you can use if you don’t feel like writing up your own code — https://github.com/puppetlabs/py-awsaudit.
Another great open source tool for increasing your tag quality and cost accountability, is Auto Tag by GorillaStack which automatically adds a tag for the user who created the resource!
3. Utilise spot pricing (it’s never been easier)
Spot pricing can allow you to save up to 90% on some resources, yes that wasn’t a typo — 90%. Early on, many people were hesitant about using spot pricing because it was not well understood and difficult to utilise with many workloads, Amazon has been working hard to make spot pricing more accessible.
You can now use spot pricing when launching Elastic Map Reduce clusters, in EC2 Autoscaling groups, Elastic Beanstalk and AWS Cloudformation to name a few. If you are a cost conscious customer, it has never easier. With the launch of the spot fleet api, spot pricing is really starting to become more accessible and will deliver amazing savings.
AWS Savings Round up
By implementing one or two of these strategies you can stand to save a significant chunk from your bill and get your boss off your back about the budget.
If you need help getting a handle on your AWS bill today, without writing custom code and setting up jobs, GorillaStack can help. We currently have two key products to help you save, Power Cycle and Power Scale.
Out of the box, these tools give you intuitive scheduling, timezone support and the ability to snooze or cancel if you need to keep your servers or autoscale groups alive when you’re working late.
You can also receive flexible AWS billing alerts about your spend into tools you know and love like Slack and Hipchat, to make sure you never get caught out spending more than you expect.
GorillaStack is here to help your resources go to sleep when your developers do and to create tools that provide significant AWS Savings. Our aim is to make the Cloud a better place.
Originally published at blog.gorillastack.com on November 2, 2015.
