EMC World Boston: The Death of Tape and What's More Important than Technology
I'm convinced most technology providers don't fully understand how slowly large companies move to adapt new technology and eliminate old. Back in the early '90s, we talked about eliminating tape for backup, and here we are two decades later at a storage event and the customer panel is talking about the recent elimination of tape drives.
I'm at EMC World Boston this week. EMC's first panel for us analysts was made up of a large transportation company, a university, a manufacturing company and a health care facility. Back in the '90s we talked about this strategically, that the major driver to removing tape libraries and moving to EMC's storage solutions was recovery. Let's chat about this.
It's the Recovery, Stupid
One company -- we were asked not to name companies and executives who presented there -- had a major outage and needed to recover. Its representatives went to its offsite tape storage company and requested its tapes back. They didn't arrive in a timely manner, and someone called, likely in a panic, wondering where the tapes were. The response was the delivery truck had broken down and was in the shop, that the tapes were safe in the truck and would be delivered when the truck was repaired.
That company cited the speed -- or lack of -- of both the backup, and particularly the recovery, as the primary reasons behind replacing its aging tape offerings and moving to more modern disk-based backup solutions.
Virtualization is relatively new, yet there's some variation in rates of adoption. Our own Mike Vizard recently wrote a post called "Application Testing in the Cloud." The range of conversion ranged from 50 percent to 80 percent, and a combination of line funding and regulatory compliance was the reason it wasn't higher. The first required line organizations, who owned the server budgets, to realize that virtualization would reduce their operating costs and they hadn't been sold on it. This seemed to point to a growing problem that relates to the distribution of profit-and-loss responsibilities to the line organizations and away from IT. Most sales efforts focus on IT as the buyer, but increasingly the line organization being served must pay for major technology projects and, if it isn't sold, it won't supply the funding.
This was an example out of the health care company: Some systems require specific hosting technology and some of these requirements do not yet allow for virtualization. Changes to the specifications are in the works, but in the meantime, the company cannot move.
Each of the companies cited examples in which the move to virtualized servers produced strong financial returns, better uptime, and generally happier customers, which is why each of the executives was advocating the technology internally.
One Throat to Choke
It was interesting the reasons why the panelists chose EMC. Apparently, it had less to do with EMC's specific technology and more to do with the breadth of its offerings. Though features and performance are certainly nice, the panelists made the point that storage is such a conservative decision that the technology was less important than the vendor's ability to assure that the entire solution would work flawlessly during a recovery event.
These companies had abandoned their tape-based backup-and-recovery products because of problems during recovery. Some of these problems evidently (reading between the lines) grew out of complexity in the solution because of the number of companies involved. So fixing the problem wasn't just replacing the tape-based offerings, but moving to one vendor who could handle the whole solution.
Public vs. Private Cloud
Generally these firms weren't fans of the public cloud, citing security concerns. However, several used it cost-effectively for low-security information. The university employed Microsoft's Live services for students' personal use, but placed anything of a private nature on private storage resources.
The panelists talked about the need to properly classify information and to make sure that only information with the lowest classification be available for the public cloud. Some believed that, over time, they might become more comfortable with the public cloud. They appeared attracted to the financial benefits, but were not yet convinced it would be secure enough for anything confidential.
Proper classification also is vital to make sure the highest classifications of data are encrypted and protected during transport and storage. Some of these companies' biggest problems appeared to be creating and enforcing policies surrounding security classification and related data handling.
It is interesting when at one vendor's conference, another vendor's product is cited by name. These executives were eagerly anticipating using Exchange 2010 because it has archival storage built in. Not having this capability in earlier versions of Exchange was clearly painful in terms of capacity management and user satisfaction. Based on what these executives were saying, this feature might be the biggest driver behind deployment of this offering.
EMC's strategic advantage isn't really its products, but the way it embraces its customers. At the end, executives from each of the companies said they were there because of how well EMC had taken care of them. It's pretty hard to get companies to publically praise a product, but EMC, of all the companies I cover, focuses the deepest on customer satisfaction. This was one of the things that most impressed CEO Pat Gelsinger when he left Intel and joined EMC.
Overall, I got the sense that these executives most appreciated that in their time of need, EMC was there and that the products worked as promised. In storage, the killer feature might be the trustworthiness of the vendor. I guess that should be obvious. Surprisingly, I don't think it is.