Read the Original Article at http://www.informationweek.com/news/showArticle.jhtml?articleID=10100204
Draft Inc. has had a business-continuity plan for some time, but it wasn't quite good enough. The largest direct-marketing and promotions agency in the United States, which counts Coca-Cola, Kellogg, and Saab among its clients, is no longer willing to wait 72 hours to get access to more than a terabyte of customer data that resides in a remote archiving facility run by a services provider. In a real disaster that results in the loss of one of its offices, its practice of backing up data from every system to its own backup server located at the same site wouldn't help the company, either.
To keep business going in an emergency, the $3.6 billion-a-year company is beginning a project to upgrade its business-continuity architecture at its New York and Chicago offices so it can retrieve critical data within two to four hours after an outage. Draft wouldn't disclose what it's spending, but the company says its approach will be cost-effective and provide a significant return on investment. "We need to eliminate any downtime as quickly as possible, because every hour has some impact," says Julian Morris, senior VP and director of IT at Draft.
Over the next two months, Draft will deploy a cluster of Hewlett-Packard ProLiant servers running Windows 2000 and network-attached storage appliances. NAS appliances in New York and Chicago, each with 1.4 terabytes of capacity, will constantly mirror each other's data. If a long-term outage occurs in either office, the company will continue to run its business from the unaffected site, more than 1,000 miles away.
"We'll have continuous online backup, with mission-critical data on a NAS device, replicated back and forth," Morris says.
The company is using Double-Take replication software from NSI Software Inc. It provides real-time replication by constantly monitoring any changes made to open files. Morris found out about the technology a couple of years ago and thought it promising enough that he and his team have been working with NSI developers since last year to make Double-Take more automated and intuitive. The software now does a better job at discerning a real outage, whereas before it would send alarms if transmissions were merely delayed.
The technology will be a key addition to Draft's 12-step process for surviving outages. The company has in place templates that outline how to conduct a situation assessment, ensure employees are out of harm's way, activate a command center, and recover data from backup sites. In each of Draft's 60 offices in 29 countries, an employee is assigned to take charge in case of an emergency; they're all given copies of the 12-step templates to follow. Now, Morris says, "NSI is fully integrated with our disaster recovery process."
NSI's replication technology has certain advantages over comparable offerings from leading competing vendors, such as Legato Systems Inc. and Veritas Software Corp., says Steve Kenniston, an industry analyst at the Enterprise Storage Group. "NSI is more robust and easier to implement for Windows," he says. It offers a point-and-click interface and the ability to move information faster than some competitors. "And right now, NSI charges half of what Veritas charges." He expects NSI to some day also support Linux.
Some analysts say that NSI can take downtime on Windows systems from almost one week a year in many implementations down to 50 minutes annually. Dell, HP, and IBM all resell the software, which is priced starting at $2,495 per Windows server, and scales up for customers using the Advanced or Data Center versions of the operating system.