Q&A: Squeezing More Out of Your IT Environment
Where should IT start when trying to do more within new (and tight) budget constraints?
In tough economic times, IT is under enormous pressure to maintain current service levels and do so with a smaller budget and often a smaller staff. What should your IT department do when you must “economize”?
For recommendations, we spoke with Mark Kremer, chief executive officer of Precise, a transaction performance management software vendor.
Enterprise Strategies: In order to do more with less, where should IT start?
Mark Kremer: First, sort your workload by priority, identify your top, must have projects and focus on them while putting the rest on the back burner. It is not easy; as such a process is usually put off by IT departments until economic realities, and then there is a lot to cut. Once you have reduced the number of balls you are juggling you have significantly reduced the risk of all of them falling down. To reduce the risk of cost overruns, a CIO would do better to avoid deployments of new applications; all new deployments tend to push costs beyond budgets, especially reduced ones.
One thing a CIO cannot do is compromise the service of a business-critical application, the workload will probably go up by itself. I would recommend taking a close look at the performance traits of such applications, they are likely to have been over provisioned or by that token ill utilized. That would be a good opportunity to improve service where it counts, save on infrastructure, but most of all cut down the ever ticking Cost of Ownership meter
What criteria should be used when trying to squeeze the most from your current IT environment?
It may sound obvious, but you might be surprised by the insight gained from a simple triage process. Triage your applications into the following classes:
- Top business applications -- the mission-critical applications a business needs to stay alive
- Bottom business applications – rarely used, legacy apps that have readily available alternatives (a combination of e-mail and spreadsheets can replace almost anything)
- Everything in between
The best strategy is to make the bottom applications obsolete, redeploy the "in-between" applications on virtual environments, and use the resources saved to optimize the top applications. CIOs are measured against many criteria, but the service level of top business applications is what is critical (and, critically, most visible). It is good business and makes good career planning sense for CIOs to invest in optimizing the major business applications by cutting out the minor ones. Starving the turkeys to feed the eagles is a tried-and-true strategy.
What’s driving the push to get more out of what you already have?
Mostly the recognition of the new economical reality -- IT departments don’t have the luxury to do more with more, and in most cases are asked to do even more with what is already in place. IT budgets also suffer in a way from Moore’s law of exponential CPU growth and the abundance of computing resources. In addition, the ever-dropping costs of computing and networking power create an environment of IT over-provisioning within the enterprise infrastructure.
Executives believe, and they are probably right, that their IT can serve the business for an extended period by better utilizing their current IT investments. They think that there is enough IT fat to live on when IT budgets are starved.
What are the typical targets and approaches IT departments take?
Most IT departments will adopt logical processes for prioritizing their investments—similar to the one we have discussed above. We have also seen approaches that take significant risks like forgoing security, monitoring, testing, or performance management even for critical business applications in hope that “it will not happen to them.” Hope is not a good strategy, not with IT.
What’s the impact of this movement?
IT is adapting fast. In better times, IT slowly gets lolled into “lazy purchasing” (that is, putting the convenience of the purchase ahead of its value). Considerations such as single-vendor purchase, vendor size (and therefore vendor credibility), and incumbent vendor preference took precedence over inherent product value, ROI, and cost effectiveness. IT is adapting fast to an ROI-value-driven process which in a sense rocks the technology boat by driving IT to consider more vendors and thus providing vendors with access to more customers. In short, this movement stirs competition which is healthy for all.
What mistakes has IT made in the past in squeezing more from systems?
Clinging for too long to the systems already in place has proven to be costly in the past. Delaying the transitions from mainframes to open systems, or from open systems to Intel-based blade platforms, has exacted monumental TCO costs from the laggards. Similarly, late adopters of packaged applications have suffered that same pain. Timing is everything when it comes to jumping on board IT trends.
Where is this “economization” headed?
Economization will squeeze existing inefficiencies out of IT infrastructure, cynically enough, to introduce new ones as most IT trends do. Paradoxically, data center computing platforms today are surprisingly underutilized (at 15 to 20 percent some analysts say) yet are still often breached by performance SLA violations. Virtualization decouples the conceptual server from the physical hardware allowing IT to squeeze more virtual servers into fewer physical servers to increase platform utilization.
Proper performance management addresses inherent application performance bottlenecks optimizing transaction response time allowing more transactions to flow through the same infrastructure at a better SLA. I believe that economizing will combine both schools of thought to better utilize infrastructure and apply it to where it will most benefit the business. Because there are no free lunches, this seemingly ultimate solution will introduce non-trivial resource sharing challenges of virtual machines and virtual environments monitoring, calibration, packaging for performance management, and, of course, licensing headaches for all.
What products or services does Precise offer to help IT squeeze more from their environment?
Precise provides Transaction Performance Management solutions for production systems.
IT is organized by tier silos -- Web, application servers, database, storage -- to optimize service costs through service aggregation. Business applications transactions flow across IT tiers in the course of their execution. The inherent conflict between cross-tier business transactions and IT silos lacking in application awareness is the root cause of most transaction performance inefficiencies.
Precise allows users to track business transaction performance across all IT tiers, detect performance loss, and qualify it by business impact and resolve it in transaction context at an unprecedented MTTR (Mean Time to Repair). Precise proactively addresses transaction performance bottlenecks often ahead of SLA breach incidents.
Precise customers do more with less by flowing more transactions through the same infrastructure at a lower TCO, what’s not to like about that?