12.2 Tactical Goals

Historical computing evolution and storage sector innovation influence new deployments. The first wave of storage networking adoption, starting in the late 1990s through 2003, helped shape the basic understanding of networked storage benefits and implementation practices. Coupling that understanding with historical computing patterns and sector innovation leads to identifiable trends useful in all types of storage planning.

12.2.1 Adapt to Infrastructure Fluidity

As we have seen throughout the book, implementing even the most basic storage network initiates a functional separation between servers and storage that in turn presents more choice and flexibility to storage architects . Once the primary server-to-storage connection traverses a network, the functional breakdown can continue by adding custom-tailored storage components . The increased performance of commodity servers coupled with specialized software means that many of these storage components can be produced and deployed rapidly and at a relatively low cost.

The introduction of function-specific storage nodes ” ranging from NAS heads to protocol converters to virtualization engines ”delivers infrastructure fluidity. For the storage domain, this flexibility will continue through the use of more specialized storage nodes across a wider functional spectrum.

IT project planners need to recognize this trend and plan accordingly . Including both custom storage hardware and more commodity-oriented platforms in new deployments guarantees a hedge on future hardware and software directions.

12.2.2 Use Distributed Resources and Intelligence

Tightly coupled with the emergence of infrastructure fluidity is the use of distributed resources and intelligence. Fluidity implies the ability to easily move within set boundaries, and with IP networks in the storage equation, those boundaries reach further than ever before. Much like water will follow the path of least resistance directly into the gravitational pull, storage processing will ultimately find a home in the nodes that can deliver the most functionality and reliability at the lowest cost. As more storage nodes become networked together, storage administrators will have the option to transfer tasks accordingly. Similarly, it will become easier to remove storage bottlenecks by adding scalable storage-processing power.

Nearly every large server and storage vendor now touts the promise of harnessing distributed resources and allocated IT services on demand. All of this functionality rides upon a coherent infrastructure, which in the case of storage comes from well-architected Fibre Channel and IP storage networks.

12.2.3 Prepare for Geographic Integration and Consolidation

Large corporations' dispersed offices have for years been linked together with IP networks servicing enterprise applications. Further, many corporations have linked upstream to suppliers or downstream to customers through electronic data interchange (EDI). That time- tested connectivity has removed many geographic barriers to everyday business.

Storage now enters that playing field through the emergence of IP storage. For internal and external applications, the ability for storage data to freely travel across secure IP networks radically changes the traditional notion of the storage footprint. Whereas it previously covered the confines of a data center, it now covers the globe.

Following the natural inclination of business solutions towards lower cost solutions, imagine factoring power consumption into the storage location decision. Assuming adequate bandwidth, it may make sense to locate large storage arrays requiring intense power and cooling close to power-generating faculties . While this example may seem farfetched to some, it highlights the unrestricted storage movement that is likely to influence future storage decisions.

12.2.4 Continually Monitor Organizational Structure and Utility

The earnings-driven corporate mantra of "do more with less" isn't likely to go away any time soon. As such, recurring high-cost components of technology infrastructure ” specifically labor ”will continue to be targeted for cost savings. Previously, storage administration has been a relatively labor- intensive process compared to other areas of IT deployment. Advances in hardware and software management are starting to change that, and now fewer individuals are needed to oversee larger storage domain operations.

Centralizing responsibility to a smaller storage professional team means more decision-making authority. For example, even though more functions may be automated, that small group will still determine storage rules and permissions for the same customer base. Theoretically, this seems like an ideal opportunity to streamline and concentrate responsibility within a highly skilled storage administration team.

However, rules are easily managed; exceptions are not. The possibility exists that exception management ”modifications to rules, one-time changes, and extraordinary circumstances ”can overwhelm reduced administration teams . Continually monitoring the structure and utility of the department ensures the storage administration team has the appropriate level of resources and the ability to accommodate unanticipated exception handling. A feedback process incorporating storage customers provides one manner to measure effectiveness. Rather than viewing this as a "complaint" structure, storage professionals should view it as a method to anticipate future requirements and should set the technical direction accordingly.