How do I Reduce Data Center Costs AND Get Ready for Cloud?

“Do more! But Spend less!  Over the past several months we have described many client interactions where financial pressures brought on by a stalled economy are butting up against enterprise needs to modernize and offer new services.  (see figure 1)  While many look to cloud to modernize AND cut costs, the view on the ground isn’t so simple. Quite simply, not every workload in the data Center is best served operating in the cloud at this time, and simply “lifting and shifting” to the cloud often does little to save costs.  It may actually cost more, if the data center space needs to continue to support the non-cloud ready remaining workloads. Infrastructure and Operations (I&O) professionals find themselves between a rock and a hard place, trying to continue service levels while cutting costs.  We can’t simply “shut down the data center”, but we NEED cost reductions today. So, what do we do?

My colleague David Cappuccio strikes a pragmatic and helpful tone in his recent research note, “Improve Data Center Efficiency While Reducing Costs”. In short, David describes a balancing act built on a base of workload placement, where I&O Pros work towards Cloud Readiness, but also optimize the existing Data Center assets to reduce operational costs, minimize new capital expense, while keeping service levels up.(see figure 2)  What is this magic?

In short, until financial restrictions subside, it’s a matter of “Don’t be cloud first – get cloud ready”, and “use what you’ve got, but do it better”, by increasing utilization and efficiency.

Simply stated, it’s a cost-conscious Data center strategy incorporating workload placement as a key aspect, where placement may ultimately include hosting or cloud, but in the short run, requires consolidation of data center-housed assets. Much like a crowded city bus during a rainstorm, we may need to pack more workloads into fewer systems until the rain stops (or financial restrictions are lifted).  We often hear that an organization is 85% to 90% virtualized, which is admirable, but if the systems those virtual workloads are running on are only 20% utilized, there is room for improvement through consolidation. From the cloud angle, we recommend going through an application assessment to determine which workloads are not only a strong fit for the cloud, but which will deliver what is most critical in the near term, whether cost savings, agility, or some other enterprise goal. When funding and time-to-benefit allow, I&O has a plan in place to begin migrating. Likewise, when renegotiating colocation contracts, we often see situations where customers are being charged a flat rate per rack assuming 5 kilowatts per rack, when in fact their power density is FAR lower. This is another opportunity for consolidation savings.  The bottom line is that while cloud may prove to offer cost savings over the longer term, near term cost reductions based on workload placement, increased density, and consolidation can offer obvious cost benefits today.