This article is part two of a series. The first post, CIO: Are you a data hoarder?, can be found here. That post reviewed the need to implement a more cost-effective, reliable form of data storage–one that doesn’t involve excessive maintenance costs (outdated servers) or use valuable facility space (again, outdated servers) all the while being unreliable and difficult to access.
Although that prospect sounds desirable and is arguably the better alternative, very rarely is it easy to replace an entire storage system. This post will address some of the IT issues one may face when transitioning to a new method of data storage.
Largely due to data growth, storage management has become more difficult for IT.
Finance, budget, and other constraints don’t help. Storage now takes up a larger piece of overall IT, Watts said. “Storage infrastructure is growing, but the [number of] people to manage the storage is not growing.”
Although the cost of storage hardware is continually falling, another hard truth is that the price of software for managing storage isn’t dropping as fast. Power, cooling, and real estate costs are also rising. Such realities can make it necessary to use data management techniques such as information life cycle management even at the file level in order to reduce actual required storage volumes. According to Watts, a business perception that storage is cheap can hinder the implementation of efficiency strategies. “Asking businesses to come up with solid policies on data taxonomy and classification, along with retention schedules, isn’t easy,” he said. “It’s far easier for them to just say, ‘buy more storage and store everything.’”
Storing everything, though, slows data analysis and hinders getting the right data to the right person at the right time, making it more difficult and expensive to make informed decisions.
Meanwhile, for organizations saddled with older storage arrays, speed and energy can
pose problems. Also, spinning discs can have a “nasty end-of-life habit,” according to Watts. Drive failures in an array can prove expensive at the business level. IT can cascade arrays—move a previously high-performance array to a lower tier when a new model is adopted—though it’s key that IT ensure business-critical data is in the right place on the right type of storage system to ensure data availability and business continuity.
Be careful of the worst-case scenario—a “fork lift.”
The worst-case scenario is a proverbial “fork-lift” replacement of a storage system, Watts said, due to the expenses, lead time, and downtime involved. Positively, some options, such as storage virtualization, provide the means to use existing equipment and add new equipment the virtualization system manages, mitigating headaches, he said.
“Storage managers need to understand how virtualization and mobile computing impact the storage environment, and how the storage environment may need to change to meet these new requirements,” Watts said. This could require cross-training and even collapsing the organization to create teams that understand more of the virtual stack and virtual machine storage relationships. Additionally, storage managers might need to work more closely with mobility and security teams, which they traditionally haven’t done.
Next wednesday, we’ll continue the conversation with the final installment–how to really make the transition.