Many organizations are implementing infrastructure automation to save time, eliminate variability, and reduce costs. John Ray explains why you should consider infrastructure automation and how to get started.
Infrastructure automation is a term that gets thrown around a lot. While it is touted as being a solution for everything from the mumps to inconsistent application deployments, it really is about one thing: going fast.
In order to move quickly, you have to decrease the amount of time it takes to provision and configure new infrastructure. All of the clouds in the world aren’t going to help if you still have to manually execute this process.
Automation opens the door for developers to rapidly provision new environments. For example, let’s say you’re a developer, and you need a new web server. Normally, you’d have to send out a request, and three days later you’d finally get a new VM-ware server. Who has time for that? Time to automate and get your server up in minutes not days.
Another reason you should consider using infrastructure automation is to decrease or eliminate variability in new infrastructure. Let’s go back to that web server example. Imagine you finally have your new server, but it’s not to the same spec as the rest of the servers. And, it might not have the tools you need on it. Infrastructure automation eliminates those kinds of errors. Servers are configured the same way, every single time. Even better – if someone does change something, the automation will put it back where it belongs.
Besides increasing velocity and decreasing variability, infrastructure automation is the answer to your boss’ demands to do more with less. Automation can significantly decrease the management overhead required to maintain complex environments. Instead of staffing up to remediate a security vulnerability, you can simply write the requisite code and then execute it. Now your small team can manage many thousands of nodes with no more effort than managing 100.
Getting started with infrastructure automation
There is some upfront work, but the important thing to remember is to take it step by step. A lot of organizations want to automate everything at once. They’ll get Puppet, or Chef, and try to automate across the board. Unfortunately, that’s not a really effective implementation strategy. Systems are too complicated to automate everything all at once. To correctly implement automation, organizations need to start small.
But finding the starting point can be difficult. You might think, “Let’s automate users, just on the web server.” While it’s a small part of your system, that’s an ineffective place to start. Automating users just on the web server means you’re managing users in other parts of your system differently. This adds complexity to your system, instead of simplifying it.
What you’re looking for is something that is common between 90% of your nodes, like an SSH key. That’s what you automate first. Something that is mundane, easy to do, and applies to the majority of your infrastructure. Then, go to something that affects 80% of your nodes, and so on. If you get these big-ticket items done first, you’ll find that you’ve automated most of your base system pretty quickly, and making other changes is simple. When you need to implement a new web server, for example, you’ve already got the base. That’s a good start.
Need technical help?
The concept “start small” may seem simple, but when it comes to implementation, most people have trouble really understanding what that means. Shadow-Soft can help you figure out exactly where to start – both in terms of step-by-step automation, and how to conquer big-ticket items.
We understand that all organizations are unique, and have individualized needs and systems to match. We can work within your structure to create immediate, impactful change.
Ready to learn more? Click here to schedule a call with our team.