We’ve all read about IAAS, PAAS, VAAS (all of these stand for “something as a service”), but the real question is what does it mean exactly and how does it benefit you and your enterprise? It’s easy for anyone to read a whitepaper, but if you don’t understand the underlying concept(s) behind what is being touted, it’s pretty much an exercise in futility. So what exactly does this all mean? This post will dispel the many myths behind what “aaS” means.
The concept of the “aaS” environment is basically to take outsourcing in-house. To offer the flexibility of dynamic service offerings previously only available to you through traditional service provider channels. Let me explain by example.
Let’s say you want the ability to “spin up” new servers as you need them without having to order, provision, and deploy new hardware. Or you just needed it immediately for a short-term project, to handle an increased workload (like a sale for your e-commerce site), or just for a new development team. It doesn’t make sense to buy a bunch of new hardware to use it on an ad-hoc basis. This is where “aaS” comes in. There are a few options, you can buy or you can build.
If you wanted to spin up 20 servers and didn’t have the time to wait for approval to purchase and provision this hardware, virtualization would be a great answer! Just a couple of clicks and you’ve got 20 new servers spinning up ready to go in about 5 min. (or less). Sure.. that makes a lot of sense, but there’s one problem. What happens if your development / qa-team needs to spin up these new environments on a very regular basis? It doesn’t make sense to have them constantly open tickets to do this.
Well, this is where aaS excels. Wouldn’t it be great if they could just go to an internal provisioning portal, spin up those instances by themselves, associate IP addresses, and just tear them down when they’re done? They could do this as much or as little as they want and you wouldn’t have to worry about dealing with the end-user. Everyone’s happy! Well.. That’s what “aaS” stands for. A service you can provide to your customers (internal or external).
You can offer them different ways to spin up these resources as they need them. An API, a Web-based Dashboard, and a few command line tools! Everyone’s happy! (Especially your support staff).
Does this exist today? Yes. Does it cost millions of dollars? No. In fact, it’s free. Just a matter of initial provisioning and you’re off and running! What if I told you it wouldn’t even cost you a penny in Virtualization Software licensing? Interested yet? Enter Openstack.
Openstack is a Platform that offers a middle layer to provide your internal or external customers with the ability to dynamically provision virtualized resources as they need. Everything from Virtual Servers, clusters, networking, storage, backups, and server imaging processes!
The great thing about it? It’s expandable. As your needs grow, you can easily add more compute nodes, storage clusters, and services without any interruption (if you plan it right).
You can always find out more about it at their website: http://openstack.org
While it may seem a bit intimidating at first, it’s actually a very easy system set up. (I decided to set one up last night across 3 servers. It took me 6 hours). Now, I have 3 servers set up and the ability to spin up about 80 virtualized servers in 5 minutes. Total cost of hardware to do this (Keep in mind this is my own personal lab, but it would cost you about $5,000.00 today in hardware). Not bad for 80 servers that are self-provisioning with all the goodies that any enterprise would need (including backups, imaging, and dns).
If I need more, I would simply just deploy more nodes or storage and just “add it” to the resource pool. I can do an initial configuration of a server, image it and launch as many of these as I want / need to quicker than most would take to do one! This is the power of an embedded self-provisioned services platform.
Welcome to the new IaaS / SaaS / PaaS paradigm.