How to architect your virtual infrastructure for private cloud success

• 1. In resource clusters or groupings in your virtual infrastructure, have you ensured a consistent computing platform (same processor/memory, same brand)?
• 2. Are all virtual servers created and run from network-attached storage?
• 3. Are your virtual networks and switches defined and made consistent across all hypervisors within a resource cluster or grouping?
• 4. Are your hypervisors within a resource cluster at the same revision? Differences in revision can cripple features necessary for availability and performance as well as business continuity.

Capacity: Capacity planning is important to managing a self-sustaining private cloud infrastructure. Without it, much of the resources that were architected to operate an efficient private cloud will be squandered on idle virtual servers – virtual servers that are no longer providing a useful service to the organization. When capacity is needed, it will be there.

Capacity is also the starting point for a chargeback or at the least, a showback cost model. This will be a mechanism that can be used inside an enterprise to ensure IT moves from a group described as overhead to one that is a business enabler. The showback model will ensure that the best decisions for business are made and that the proper costs are budgeted for resource usage to ensure a sustainable funding model in the organization. When growth in capacity is needed, it can be procured.

Provisioning: One of the largest benefits to come from the private cloud is the notion of self-service provisioning or business-driven provisioning. This is accomplished through automation built on top of your virtual infrastructure – from requests through approvals and including all components necessary to get operating systems and applications installed and configured to return an up-and-running service to the client. Automating key tasks within the process is paramount to success with your private cloud.

Availability and performance: Virtualization offers a host of features to improve on availability and performance. You should look at these features for your virtual infrastructure and implement them where possible. High availability features allow your virtual infrastructure to recover from failure of a compute resource (loss of a host, blade or physical server, within the virtual infrastructure) by restarting a lost virtual server on another host. Resource balancing features allow your virtual infrastructure to move workloads around to ensure the best performance possible within that area of your virtual infrastructure. Take a look at how your virtual infrastructure deals with these items and ensure that they are implemented to the fullest.

Business continuity: Does your virtual infrastructure have a formal DR strategy and plan in place? It’s important that all elements of the virtual infrastructure be examined from the hypervisor hosts themselves, to storage and networking, to the virtual servers they support. Are there sufficient resources at the DR site to support your critical applications? Are you satisfied with manual restoration or do you need a more automated approach? How often should you test your DR plan?

In our next installment, we will continue on this topic of IT maturity and look at where you are in terms of managing your virtual infrastructure, security, governance, development and the overall process.

Remember, "cloudification" isn’t just about me telling my story. We can only learn and move forward by sharing ideas and exchanging concerns and questions about cloud adoption. I’d like to hear your thoughts, experiences and best practices regarding how virtualization and the cloud have affected your enterprise and the way you deliver IT services. And, if you have any questions or want to start a discussion, I welcome that, too.