Realising the Full Potential of Converged Infrastructure

Part of our CTO Series of articles aimed at technology decision makers.

Adopting a converged compute platform has many benefits. Unfortunately it’s not as simple as wheeling in the new technology, throwing workload at it, and sitting back to reap the rewards. Our experience has shown there are 3 critical areas to work through to ensure your move to converged infrastructure delivers the optimum benefits that it’s more than capable of.

Service Design

First of all it’s important to have a clear idea of what you’re going to run on your new platform. Categorising the types of services that will be run and creating Service Level Packages to define and describe these have a number of benefits. It gives the business clarity of what’s being offered in the service. Having a properly documented service with clear and concise terms and conditions reduces the chances of misunderstandings or unrealistic expectations. From this basis your Systems Integrators and Application Teams have the information they need to interconnect other systems and to correctly consume the service. This helps drive consistency and standardisation, which along with clearly defined service levels, gives the operations team a solid foundation from which to manage the service. In terms of the services themselves, our experience has shown that significant savings in terms of time and cost can be obtained by not only running IaaS, such as VDI or Server virtual machines, but by moving into the PaaS layer and offering Database as a Service and other Middleware applications.

On-boarding Process

Understanding the sorts of workloads that will be run on the new platform is just as important as designing the actual services. Moving disparate workloads into a new service can lead to all sorts of issues no matter how well defined the new service happens to be. The best way to understand the workload is to talk to the Application Owner and to observe the application over a period of time. The sorts of metrics typically gathered include Disk I/O, CPU, storage and memory usage. This information allows the on-boarding team to ‘right-size’ the new environment, and to have a better understanding where the busy and quiet workload periods tend to be. Our experience has shown that examining each workload and ensuring it conforms to the service guidelines, as well as having an operational baseline, makes subsequent support so much easier.

Operating Model

Perhaps the most significant aspect of moving to a converged platform is the opportunity to adopt a new approach to operations. Many organisations have their operational teams and processes based along traditional technology silos. This means having a specific team to support storage, another to support compute, another to support databases, and so on. While this is a solid support methodology, it tends to instil a more ‘reactive’ approach to operations. Under this model, each team will work through incident, request and problem tickets via a queue, the focus being on dealing with each of these as they arise. Because staff are usually measured on ticket closure, there is little incentive or opportunity to work on incident prevention or service improvement initiatives. Other issues can arise where a particular ticket crosses a technology boundary or it’s unclear which team is best placed to deal with the item, it may ‘bounce’ between teams causing delays in resolution and closure.

With converged computing, the technologies of compute, network, storage and the hypervisor have been brought together. It stands to reason that the support teams and processes should also converge.

It stands to reason that the support teams and processes should also converge.

Our experience has shown that creating a support team that owns and manages the converged platform and its associated services completely changes the operating dynamic. Specifically the approach to operations shifts from being ‘reactive’ to ‘proactive’. Providing you have clearly defined Service Level Packages in place and followed an on-boarding process, the operations team have the tools and knowledge to manage the service in its entirety. They have intimate knowledge of the customers and workloads running on the platform, and the customers have a ‘one stop shop’ within IT to interact with. The focus becomes that of prevention and maintenance, heading off any issue or trend before it has a service impact.

In summary, moving to a converged platform can be the catalyst to introduce change. The operating model for some organisations may present the biggest challenge, however the benefits can reach far beyond what intrinsically comes with the converged infrastructure itself.