Most teams doing capacity management today still tend to be very infrastructure focussed. In fact, in many enterprises they are platform focussed – one team looking after storage, one virtualised compute, a third for network, and maybe another for mainframe. But in today’s complex, fast-paced digital economy this approach is exposing businesses to a high level of risk. Instead, capacity management teams need to have clear visibility of the inter-relationships at play between underlying infrastructure resources and the applications, business services and user demands that run across them. They need to be able to talk to the business in terms the business understands. So, in order to remain relevant, capacity management needs to move with the times and become service aware.
Yes, the capacity management team needs to have some domain knowledge (what are the relevant capacity metrics for different platforms, how do we access them, etc.), but after that the questions and requirements are generic: how much resource do we need; where do we need it; can we reclaim resource from over-provisioned areas; and so on.
To really deliver value they need to connect domain knowledge up with knowledge of the business services that are running on the infrastructure. Without this they will never be able to add real value to the business, and risk becoming viewed as irrelevant (or worse, as a problem).
There are many levels to this connection between services and infrastructure, but the starting point is building (and maintaining) a mapping of which services use which components. In a mature organisation this mapping might be discovered automatically and maintained in a CMDB. In many (most?) organisations it might reside in a number of spreadsheets or home-grown databases. Regardless of where it exists, this information is as valuable to the capacity management team as the utilisation and performance metrics that are normally focussed on.
Armed with this mapping information the team is now in a position to make its capacity analyses relevant to the business. It can show them which services are using the most (and least) capacity. It can identify components within those services that need additional resource – for example, a server that is consistently hitting high levels of utilisation – and so help improve service quality. And this makes it much easier to have a discussion about components that have been allocated too much resource, so that they can be reclaimed for use elsewhere.
Analytics, predictions and correlations
But this is just the beginning of the journey. If the team is applying analytics to its capacity data, it can predict future utilization levels and thereby future issues – and can tell service owners about future problems they may have, and what needs to be done, jointly, to prevent them. It can bring service data (number of users, types of transactions, etc.) into the capacity management tooling and report on that alongside the more ‘traditional’ infrastructure metrics. And with just a little more effort it can quantify correlations between these datasets, and use them to determine how many users/transactions can be supported by the existing infrastructure, and where additional capacity is required if the service needs more than this. These are the kinds of conversations that add real value and turn IT into a close partner with the business.
Integration and automation
Sumerian is at the forefront of delivering this type of service aware capacity management and has recently partnered with ServiceNow to integrate service aware capacity management into this leading ITSM platform.