When your data center is constantly under pressure to address operational risks, try leveraging some ancient wisdom from King Leonidas and William Wallace
A Greek force of approximately 7,000 men marched north to block the pass in the summer of 480 BC. The Persian army, alleged by the ancient sources to have numbered in the millions but today considered to have been much smaller (various figures are given by scholars ranging between about 100,000 and 300,000), arrived at the pass in late August or early September. Vastly outnumbered, the Greeks held off the Persians for seven days in total (including three of battle), before the rear-guard was annihilated in one of history's most famous last stands. During two full days of battle, the small force led by King Leonidas I of Sparta blocked the only road by which the massive Persian army could pass. After the second day of battle, a local resident named Ephialtes betrayed the Greeks by revealing a small path that led behind the Greek lines. Aware that his force was being outflanked, Leonidas dismissed the bulk of the Greek army, and remained to guard the rear with 300 Spartans, 700 Thespians, 400 Thebans and perhaps a few hundred others, the vast majority of whom were killed.
-- Wikipedia, The Battle of Thermopylae [emphasis added]
Compare that to the Battle of Stirling Bridge, where William Wallace and his much smaller force of Scots prepared to make a stand against Edward I and his English forces. He chose a battleground that afforded him a view of the surrounding area for twenty miles, enabling him to not only see exactly what challenges he faced, but to make his plans accordingly. Leveraging the very narrow bridge at Stirling and some somewhat unconventional tactics at the time, he managed to direct his resources in a way that allowed him to not only control the flow of opponents but ensure victory for the Scottish forces.
What CIOs should take away from even a cursory study of these battles is this: strategic control can enable you to meet your goals with far fewer resources than expected. The choice of terrain and tools is commonly accepted as a force multiplier in military tactics. The difference between the two was in visibility; ultimately it was a lack of visibility that caused Leonidas' strategy to fail where Wallace was successful. Leonidas, unable to see sooner that he was being outflanked, could not provision resources or apply tactics in a way that enabled him to defeat the Persians. Wallace, on the other hand, had both visibility and control and ultimately succeeded.
What's needed in the data center is similar: finding strategic points of control and leverage them to achieve a positive operational posture that not only addresses implementation and architectural requirements but business requirements as well. IT has to align itself as a means to align with the business.
This strategic trifecta comprises business value, architecture and implementation and when identified, these strategic locations can be a powerful tool in realizing IT operational and business goals. Strategic points of control are almost always naturally aggregation points within an architecture; physical and topological locations at which traffic is forced for one reason or another to flow. The locations are ones within the data center in which all three strategic advantages can be achieved simultaneously.
Applications and data cannot be controlled nor policies enforced upon them to align with business goals on a per-instance basis. Applications and storage resources today are constructs, comprising multiple infrastructure and application services that cannot be managed effectively to meet business goals individually.Strategic points of control within the data center afford a unique opportunity to view, manage and enforce policies upon application and storage services as a holistic unit.
You'll note the similarity here with the battlegrounds chosen by Leonidas and Wallace: Thermopylae and Stirling. Thermopylae was a naturally occurring location that narrowed the path through which the invading army had to travel. Mountains on one side, cliffs on the other, Xerces had no choice but to send his army straight into the eager arms of the Spartans. Stirling is located within the folds of a river with a single, narrow bridge. Edward I had no choice but to send his men two by two across that bridge to form up on the chosen battleground, allowing Wallace and the Scots to control the flow and ultimately decide the moment of attack when it was most likely that the Scots could prevail.
As a data center technique, the strategy remains much the same: apply policies regarding security, performance, and reliability in those places where traffic and resources naturally converges. Use the right equipment in the right locations and the investment can multiply the efficiency of the entire data center just as both become force multipliers on the battlefield.
The policies implemented at each strategic point of control enable better management of resources, better direction of traffic, and improved control over access those resources. Each point essentially virtualizes resources, and policies that govern how those resources are access, distributed and consumed can be enforced. They optimize the end-to-end delivery of resources across vastly disparate conditions and environments. Such points of control, especially when collaborative in nature, provide a holistic view of and control over top-level business concerns: reliability, availability and performance. Leveraging strategic points of control also affords creates a more agile operational posture in which policies can be adjusted dynamically and rapidly to address a wide variety of data center concerns. All three foci are required; a lack of visibility by concentrating on individual performance, availability and capacity (operational risks) does not afford the opportunity to meet business goals. It is the performance of the application as a whole, not its individual components, that is of import to the business. It is the cost to deliver and secure the application as a whole that determines efficiency, not that of individual components.
These strategic points of control also offer the advantage of being contextually aware, which enables policies to be applied based on the resources, the network or the clients. Policies might be applied to all tablets or all applications of a specific type or they might be dynamic based on current operational – or business – parameters. Strategic points of control enable resources to be more effectively and efficiently managed by policies instead of people. This has the effect of tipping the imbalance of burden that currently lies primarily on the shoulders of people toward technology.
The goal of IT as a Service and a more dynamic data center is wholly supported by such a strategic trifecta, as it provides the means by which resources can be managed, provisioned, and secured without disruption. The virtualization of resources and their associated policies enables a more responsive IT organization by making it possible to manage resources in a very service-oriented fashion, applying and enforcing policies on an "application" rather than on individual servers, instances, or virtual images. A strategic point of control in the data center is the equivalent of a modern Thermopylae.
Like ancient but successful battles whose tactics and strategy have become standard templates for efficiently using resources by leveraging location and visibility, their modern equivalents in the data center can enable a CIO to align IT not only with the business, but its own operational and architectural goals as well.