We don't need Christmas presents all year!

I can be a big kid sometimes. I still get excited at the prospect of seeing Mickey at Disneyland Paris, at the thought of buying yet another guitar (my Achilles heel) and yes, even at the prospect of flying somewhere - although a certain budget airline is working hard to take the shine off that.

But what really gets me thrilled is the prospect of a new version of a favourite piece of software, including our own. It's like a Christmas present; I can kind of guess what it is from the shape, but what will it do? Will it be shinier than my old one? Cool features? Faster? Ooh, I just can't wait!

And while tearing off the virtual wrapping paper and installing my new thing is fine for personal use, perhaps for our businesses that's not the only way to get the same payoff. Maybe we can get deeper into what we already have, use it in innovative ways, be able to excite our customers with something they've been crying out for without the need to upgrade?

We know that users of cloud management have been desperate for a way to create automation policies which combine application performance data, VM performance metrics, cost information and date/time needs. Meanwhile, with Abiquo 4.4, launched a few months ago, we changed the way we delivered the control plane for cloud services to be available across the entire hybrid cloud (or multi cloud if you prefer).

You can read about the initial outcome in my last article, where I got all excited about cloud bursting. And then we started to realise that by combining existing Abiquo features, such as the ability to push custom metrics into our monitoring without any development needed, and then use these to drive our automation policies, we could do almost anything we could think of!

For example: wouldn't it be cool to tell your cloud platform to monitor the overall application performance and accumulated cost for a scaling group of VMs in AWS, and if it exceeds a certain value in a week AND the application performance is within a desired range, remove some VMs? And if it's the weekend, scale them down anyway?

Or to watch the performance of specific workloads and if they have a problem, add another one and when that's running, reboot the first one? Or even do it all across multiple clouds - my local application is slow, burst to AWS until the cost exceeds a defined amount, then email me and shut down AWS and burst to Azure instead?

And THEN how about we look at application availability in aggregate, and use that as a driver for cloud bursting, scaling, rebooting VMs... whatever! We just got multi-cloud application-level resilience!

Maybe I'll be able to wait for Christmas a little longer after all.

If you'd like to find out how all this stuff works, and that it's not just slideware, drop me a note.