Disk-to-disk (D2D) backup/recovery is one of the hottest trends, and controversial topics, in the storage industry these days. Whether 'tis nobler to suffer the slings and arrows of outrageous tape backup, or to take arms against this sea of troubles and turn to cheap disks...
Fueled by inexpensive ATA (and, soon, Serial ATA) disk arrays and new software, the trend toward disk-based backup is in full swing. In a survey of InfoStor readers, more than 56% said they planned to implement disk-based backup.
The issue is only controversial if you buy into the assumption that D2D backup will replace tape. It won't. Usually, D2D backup implementations are actually D2D2T. Old habits die hard.
The real issue with D2D backup is not its effect on the tape market; it's how to determine which D2D methodology is best for your situation. And that's difficult because there are almost as many methods as there are players in the fledgling market.
This month's Special Report, by The Data Mobility Group's Dianne McAdam, explores the options for disk-based backup and restore. Most of the methods can be categorized into two bins: "disk as tape" (e.g., virtual tape) or "disk-as-disk," depending on whether or not the secondary disk masquerades as tape. But even within those two categories there is a confusing array of options. Dianne's report will help you get organized before you begin kicking tires.
Another controversial debate is Fibre Channel versus IP (e.g., iSCSI) SANs. But again, it's only controversial if you cast it as a "Will IP replace Fibre Channel?" debate.
Based on survey results, about half of our readers have no plans to implement iSCSI. However, a smattering have already implemented iSCSI storage area networks (SANs), 17% plan to do so within the next six months, and 22% will install an IP SAN in 6 to 12 months. That qualifies as a trend.
To see why—and how—early adopters are building iSCSI SANs, read senior editor Lisa Coleman's cover story, "Low cost lures users to IP SANs," which pre-sents a number of case studies.
Of course, the low-cost argument isn't going unheard in the Fibre Channel community, where vendors have finally started dropping prices. You can get low-end host bus adapters for about $800 per port. Not very good by Ethernet standards, but a step in the right direction if Fibre Channel SANs are going to take hold at the low-end of the market. (Now that's a debatable topic—whether Fibre Channel or IP SANs will catch on at the low-end.)
And intense competition in the Fibre Channel switch market has led to some significant drops in per-port pricing. It's all good news for end users on tight budgets.