In the last post, I likened the virtualized application model to an organism of sorts, where each individual component can work together virtually as a dynamic pool of resources. But what happens when something isn’t working as it should?
Well, that’s really where the next logical step in your organization’s virtualization evolution comes in, one central management console that understands all the pieces plugged in – in essence, one single point to look at where you can automate and resolve issues and reduce routine maintenance of your environment, what we call: dynamic provisioning and automation.
For instance, say a server turns off in the middle of the night. Where you are already in your virtualization evolution, you’ve written the required rules to ensure that your applications move through or are taken over by other functioning parts of your organism, other servers, and they continue to run like before. This sort of automation ensures that you won’t have to do anything immediately. You may get notified so that first thing in the morning you can have a look at the physical environment and do some troubleshooting via your central console, but you essentially bypass the usual human remediation process and let the software handle the issue for you.
The same goes for provisioning the amount of resources needed for any given application to run on a server. You’ll always want your application to have an appropriate amount of memory, processor, storage and network connectivity. So you’ll set rules in your software to deal with deficiencies that may occur or when thresholds are reached. You might decide, for instance that you need a certain application to have 1 Tb of storage but if the system ever approaches 800 Gb, space available for the application will automatically increase to 1.5 Tb and an email will be automatically sent to you letting you know when it’s happening. And if for any reason there isn’t enough storage space to expand, you’ll get an email telling you potentially weeks in advance, so you can call your systems partner and get another tray of disks for your storage array.
Incidentally, it’s important to keep in mind that this isn’t a one size fits all proposition. The amount of dynamic provisioning and automation will depend on what’s appropriate for your IT investment and can be configured to your needs. After all, there’s no point buying off on a sophisticated software package designed for 1,000 servers if your environment only needs 20, right?