Setting up the Storage Center is a simple matter of configuring the IP address of the controller through a serial connection, which is usually performed by the Compellent tech during on-site installation. The wizards that help you automatically configure the system for use can safely be left in their default configurations, and virtually any setting you make can be changed later without having to reconfigure the volumes you create. This includes RAID levels, spare drives, number of tiers of storage, volume size, and more. The system automatically defines storage tiers based on the types of drives available; if more storage is added, additional tiers become available.

[ CompellentStorage Center 3.5 was selected for an InfoWorld Technology of the Year award. See the slideshow to view all winners in the storage category. ]

Click for larger view.

Compellent has clearly put considerable development effort into usability. Most features have setup wizards that make configuration a simple matter of entering the necessary information, or more often, selecting it from a list of available choices, and taking the default choices for most items. Thus, setting up and cloning boot from SAN configurations using the Server Instant Replay function, or setting up remote replication over IP to a second Storage Center in another location, are both simple operations. Remote replication includes deduplication and encryption, although not dedicated compression.

The automated data migration feature is unlike many ILM systems in that the administrator does not have to manually designate files to be moved or even create policies that move all files of a certain type or location. Instead, data that is infrequently accessed is moved automatically from the first-tier FC drives (10K or 15K) or second-tier FC drives (7.2K) to third-tier SATA-based storage, without requiring manual configuration or intervention by the administrator. Frequently accessed data is automatically kept in first-tier storage, and data that has been migrated is automatically moved back to first-tier when it begins to be used. Of course, the admin can designate partitions or even folders that will not be automatically migrated, if desired.

The instant replay function is another extremely useful feature of the Storage Center. An instant replay is a snapshot of the system at a given point in time. Because an instant replay uses only pointers to data, two replays use only as much storage as is required to store the pointers and copies of any files that have changed between the two snapshots. This amounts to data deduplication, and allows for virtually unlimited snapshots (unlimited by the software and limited physically only by disk capacity). Instant replay also supports VSS (Volume Shadow Copy Service) in Windows Server 2003, so Windows users can take advantage of the deduplication feature as well.

The same instant replay functionality can be used for servers. Once a boot volume is set up and the operating system is installed for a server, a second bootable volume can be created, and the only additional space used is that which stores the files that are different. For Linux this might amount to a few configuration files, and even for Windows it doesn’t amount to much except for the Windows swap file. This means that you could have several hundred servers booting from one main OS image, and each would maintain only the files that are different. It also means that rolling out patches and updates can be greatly simplified, since only the main image needs to be updated, and the others recloned. Unfortunately, there isn’t yet a way to push changes in the main image out to the clones.