DON’T GO ALL-IN: MAKING A WISE INVESTMENT IN DATA CENTRES

By ZahlLimbuwala, CEO, Romonet

We are currently witnessing a transformation in the way in which the financial sector purchases and delivers IT. The commoditisation of computer hardware, virtualisation and the presence of cloud services has opened up a number of options for companies to reduce the costs of IT. Yet CIOs need to balance this against the costs and expertise that have already been sunk into existing IT infrastructure, particularly data centres.

Buy a house and then rent a flat?

Regardless of how it sources its IT services, any self-respecting IT department should be able to match capacity with requirements. The problem is that many firms’ data centre capacity far exceeds their needs, with little to no understanding of cost and how this unused capacity affects the business. For instance, in the effort to reduce the costs of running and operating IT, many organisations view the cloud as an alternative to in-house IT services. However, without proper controls and use policies in place, capacity (and thus cost) of Cloud services will grow exponentially and result in an even higher cost than the internal one; let alone the question of whether a service can make the security and compliance needs of your average financial institution.

Ultimately, moving to the cloud equates to renting capacity in someone else’s data centre as opposed to your own, which might not be as cost-effective as it first appears. To use an analogy, if you already own a house, it would be bad financial practice to rent a flat, split your family and belongings between the two, and keep paying the full bills for both while the house lay half-empty. Also if you can’t compare the costs (one being capital dominated by the mortgage and the other being operating cost dominated by the rent) on a ‘unit costing’ basis then an accurate comparison of which is cheaper in any given situation becomes very difficult.

End Game

Zahl Limbuwala
Zahl Limbuwala

Firms are currently placed in a Catch-22 situation. Demands for greater cost savings in IT make outsourcing and the public cloud an attractive proposition. Yet the more services are taken away, and the more IT infrastructure is reduced, the more costly in-house services become and any historic investment business case for in-house data centres becomes compromised. The ultimate end-game of this situation could be the complete winding-down of in-house data centres, with all IT services instead placed in the cloud or with a colocation provider. This would be unacceptable to the vast majority of firms for many reasons, not least the fact that doing so is essentially relinquishing total control of critical services and data.

More so a total move to any solution doesn’t make sense. With cloud, CIO’s have the opportunity to break the ‘service mono-culture’ that existed traditionally, with effectively a single cost/single SLA solution. Instead they can start to target services into appropriate cost and service level buckets delivering much more value for each invested and operational dollar to the business.

The most important thing however is to get a handle on how much of the available capacity is actually being utilised vs. sitting empty and still being paid for, then figure out a rationalisation that drives up the use of capacity that already has sunk and/or on-going contractual cost associated with it.

Energy efficiency is not financial efficiency

For this to happen, firms need to understand just what the different options cost them. While Power Usage Effectiveness (PUE) and other efficiency-focused metrics have been touted as the best way to measure a data centre’s value, ultimately the Total Cost of Ownership (TCO) and Unit Costing should be at the forefront of decisions. Once they have their TCO, firms should then make sure that they have the TCO breakdown for each IT service. With this information, CIOs can then begin to know what will happen to IT costs as data centre usage changes, and make decisions that include a real view of the financial impact for any given scenario.

Conclusion

So why aren’t firms doing this right now? The effort needed to understand the true costs of IT services, normalising and then comparing these costs in a highly complex environment, involves a great many variables and substantial human resource. Too often this complexity means that simplifying assumptions are made, and ‘fudge factors’ added, all in order to reach an expected result that is often incorrect or as best sub-optimal. Yet eventually CIOs will need to decide on the future of their data centre infrastructure. In order to gain a complete understanding of the costs involved, firms should undertake a comprehensive TCO analysis now, so that they can ensure that whatever data centre investments are made in future will be money well spent.