Storage tone: what networked storage can learn from consumer product goods - Storage Networking

Computer Technology Review, Jan, 2003 by Alan Kessler

* Begin the required maintenance--an additional two to eight hours.

In dynamic storage environments where changes are typically easier to implement than in static environments, extensive planning and some downtime is also required. In a dynamic network, the customer's data must be physically replicated (typically at 100GB per hour for backup or replication), either to other disks or to backup tapes, before making the configuration changes. Once the data has been safely moved off the storage volumes, the structures can be deleted and recreated to reflect the new configuration.

Once the new data structure is configured, the following steps must then be performed to prepare the new volume for restoring of the previously backed-up data:

* Update host configuration device table files--two hours.

* Update network storage configurations--two hours.

* Layout new OS specific file system--two hours.

* Restore original data at 100GB per hour.

These steps are in addition to the required planning, scheduling, notifying and maintenance. As you can see, even this relatively minor change requires a typical storage unit be taken offline and made unavailable to users for a period of time. Since downtime adversely affects business continuity, the goal should be to decrease the amount of time spent offline.

With a smarter storage network, the difference in the time spent on planning, downtime and implementation pale in comparison. Imagine a smart network storage solution that allows an administrator to simply increase volume size through a drag-and-drop function that instantly creates the new storage space required. At the same time, it automatically creates the required structures and extends volumes into the newly allocated space. A user can perform a storage management function without taking the application offline, or disrupting end-users. No more weekends spent reconfiguring arrays because what used to take several days can now be done in a matter of seconds.

Intuitive Storage

A smarter network storage solution allows the product to be much more intuitive as well. It can be argued that lack of intuitiveness among storage products prevents widespread adoption. In fact, one could suppose that even if storage solutions were free, adoption rates would still be low because existing solutions are not easy to use. In the near future, with solutions based on consumer goods philosophies, users will no longer be intimidated by network storage solutions, but rather understand how to use them instinctively. Like handheld PDAs and other consumer devices, network storage should be easy to understand, self-healing, dynamic (without user intervention) and enable customers to use it to its fullest capabilities without ever having seen it work before. In other words, storage solutions shouldn't require a Ph.D. to operate them.

Intuitive network storage solutions move beyond wizard-driven GUIs that take the user step-by-step through each management process to offer a single-step drag-and-drop action for even complex management tasks (such as expanding volumes or moving data from one disk to another). They also offer a command line interface that enables more sophisticated storage administrators to build scripts for common tasks and to access advanced capabilities. In addition, they'll offer time savings through consolidation of network resources. For example, a typical storage network consists of up to 50 different servers all supporting various applications, such as email, databases, ERP programs and the like. Intuitive storage solutions will tie those various servers together to perform all storage and data management processes from a single location. Leveraging these intuitive systems can result in significant improvements in overall management when the disparate technologies that manage the environment work together, to deliver performance results throughout the network.