Selling Network Uptime
As end users' dependence on networks increases, solutions that protect against network downtime are an easy add-on for systems resellers.
Business Solutions, January 1998
A RAID system stores data traditionally kept on the disk drive of a network server. This data includes computer files and software applications accessed by network users. For example, in an insurance agency, this data would include computer files with information on clients and claims. It would also include insurance software programs used to organize and access these files.
In a traditional network server, data is stored on a single disk drive. "Drives failing are the most common cause of server failure, and subsequent network failure," says Whyte. "Failure of the drive in a network server can bring a business to a standstill."
Creating Access To More Than One Copy Of Network Data
A RAID system uses more than one drive for the storage of network data. Depending on the level of RAID being deployed, a copy of the data stored on any single RAID drive, is also stored either on another drive in the RAID system, or spread out across multiple drives. So, when a drive fails, that failed drive's data can be reconstructed. If a RAID system rebuilds data while the network is running, its drives are "hot-swappable."
Half-Day Of Downtime Equals Cost Of RAID Installation
"The number of users on a network, the number of applications running on a network, and the availability of data from other sources, are some variables that determine the amount of money a network crash can cost a business," says Whyte. "A reseller offering RAID systems can do 'cost-of-network' failure calculations for customers by plugging figures in for these variables.
"A 40 gigabyte RAID system (enough storage for a network of about 20-30 users) sells to an end user for about $15,000. A half-day of downtime at most end-user sites costs more than a RAID installation would at that site."
When A Customer Can't Say No
John Frassel of RAID vendor Antrone Research, Inc. (Anaheim, CA), says his company recently began a relationship with a reseller whose main business is helping end users recover data after their networks crash. "This reseller's services are expensive," Frassel says. "By the time he reaches the end-user site, the customers are usually on their knees crying about their lost data. At that point, the reseller will explain to them that a RAID system will help prevent their system from crashing again. The end user is usually not in a position to say no to a RAID installation."
RAID Complements Tape Backup
Frassel adds that although RAID protects against network failure, it is not a substitute for traditional tape backup. A tape backup makes a copy of a network's data at the end of a given period of time, typically a workday.
If a user accidentally deletes vital information from a file or a computer virus corrupts a file by entering a tape backup, the user can retrieve the version of the file that existed at the end of the previous day. The user can then copy that version and start over.
"In a RAID system, if a user accidentally hits the delete key and loses an important file, that process is copied in the RAID system's backup. Corruption from a computer virus will also be copied," says Frassel. "While RAID protects against network downtime, it does not protect against manual errors or viruses. A RAID system and a tape backup system are complementary sales. This is good news for resellers, because there are still good margins on both."
Sorting Through A Sea Of Vendors
Once you decide to offer RAID, you will be faced with choosing a vendor, not an easy task in today's crowded market. "There are a lot of Johnny-come-latelys in the RAID industry. If there are currently 150 RAID vendors, maybe 10 of them are the same ones that were selling RAID last year," says Whyte.
Whyte says that resellers should test RAID systems before installing them, to make sure that they do what vendors claim they do. "For example, a lot of RAID vendors claim they have upgraded their systems to operate through Ultra SCSI (small computer system interface) ports," says Whyte. "Maybe only one in 10 offers true Ultra SCSI capabilities, though. Although the other nine may connect into an Ultra SCSI port, most of their components are still designed to operate on slower Fast and Wide SCSI levels. As a result, they do not operate at maximum Ultra SCSI rates."