There are some common initial questions people have when they are considering building a cloud, questions like:
• Is the cloud right for me?
• Should I go private, public, or hybrid?
• Is it best to keep the cloud separate from the rest of my IT, or should I merge them?
• How much will it cost?
Once these questions have been answered and the cloud is up and running, a whole other set of questions start to come up about how the cloud is going to grow and evolve to support the business:
• How do I manage the cloud?
• What is involved in keeping a cloud up and running?
• Can it scale?
• How can it integrate with the tools I use to manage the rest of my IT, like Remedy?
• What are the tangible benefits, and how can I quantify them in a business case?
Work your cloud
As with the rest of IT, you can’t just wind up a cloud and set it going; it’s going to need a certain amount of work to get the most value out of your cloud. There are actually two levels of maintenance involved: maintaining the infrastructure platform (assuming that you didn’t opt for a pure public cloud, of course) and maintaining the user-facing parts of the cloud, mainly the service catalog. The offered services need to be kept up to date and aligned with users’ expectations and needs, even as those change over time.
The good news is that maintaining the in-house part of the cloud’s infrastructure is not much different from maintaining any large data center. The difference is that there are a large number of servers that are pretty much identical at the level that the cloud admins need to worry about. This means that the maintenance is a perfect candidate for very extensive automation, with only occasional manual work being required.
Keeping the service catalog up to date is another matter, and might require a new role to mediate between the users and IT. Given the pace of change, it is very important to choose a technology that will make this process easy and painless, and not one that requires a lot of manual work to create new catalog entries, update existing ones, or spawn variants.
But does it scale?
Scalability is one of the main reasons that people move to cloud infrastructure. The idea is to get more efficient use of in-house infrastructure and also involve public infrastructure to deal with peak loads or use cases that are best suited to that cost model. The caveat is that the applications needs to be able to take advantage of this capability. An extra server doesn’t help your overloaded cluster until it’s actually added to the cluster, so you need a cloud platform that is capable of delivering a full application stack and adapting it on the fly.
Since the cloud is the same as IT, just harder, better, faster, stronger, it will also need to be managed much the same as the rest of IT. The same processes, audit trails, and approvals still apply, even if the resources concerned live in the cloud. The integration has to be completely automatic, however; manual processes simply cannot keep up with the speed of cloud, and will either cause huge bottlenecks or, more probably, be bypassed as quickly as possible.
Make it pay
Finally, the tangible benefits of cloud can be hard to quantify up front, but one of the benefits of automated delivery is automated information gathering. The KPIs and trends coming out of the cloud platform are invaluable to document volume and usage patterns. These data can then feed ongoing alignment of IT expenditure and effort to business value.
by Dominic Wellington