Users Describe DR Detriments

SAN DIEGO -- Storage Networking World -- Lack of testing, lack of CIO support, and interoperability issues are reducing disaster recovery to mostly talk in many organizations, according to users at a presentation here today.

During an address this morning, William Peldzus, director of storage architecture at consulting firm Glasshouse Technologies, warned that there is a worrying lack of focus on disaster recovery testing from users today. Only a handful of around 50 execs attending the presentation would admit that they have a "detailed" disaster recovery plan in place.

Against this backdrop, Peldzus urged firms to "test, test, test," or risk serious problems further down the line. "Probably the biggest issue [for firms] is that disaster recovery doesn't work after they have implemented it," he said.

Users in the audience agreed. "You have to do your testing, but it's always a challenge to get there... the research, setting everything up, and getting everything together," explained Bernard Rueschoff, IT director of San Miguel County in Colorado.

"It's absolutely the number one reason why people fail in their disaster recovery -- because they fail to test their systems," explained John Careccia, computer services manager of Lane County in Oregon, who admitted that he has concerns about his own organization's Omniback system from HP. "The tapes are not impervious to dying because of use. My biggest push is to move to disk-to-disk-to-tape."But getting the support of high-level execs to implement a detailed disaster recovery plan was identified as a challenge by a number of users attending the event. "In many companies, CEOs do not give a high priority to disaster recovery, maybe because it doesn't make money," said Tak Chikaraishi, leader of the special project team at Tokyo-based CLC Corp.

These sentiments were echoed by Careccia: "Its getting your business owners to understand the importance that they have to put on the disaster recovery of their data."

Even with executive support, there are still plenty of technology hurdles. Peldzus warned that, in the current climate, users looking to array-based replication face vendor lock-in. "It needs the same storage array on each side," he said, adding that this limits the options available to users. "There are some intricacies that you have to be aware of."

To make matters even worse, users will need specialized software from their array vendor, such as EMC's SRDF or PPRC-XD from IBM. Peldzus added that HDS is working hardest to address interoperability issues.

Users, certainly, are keen to see interoperability addressed. "I want a more heterogeneous environment," said CLC's Chikaraishi. "Maybe in the future there will be software to link hardware from different vendors."San Miguel County's Rueschoff told Byte and Switch that he will be keeping a close eye on replication standards such as the various versions of the Lightweight Directory Access Protocol (LDAP) in an attempt to get around this problem. "Then you can play with the crossover of different architectures," he added.

— James Rogers, Senior Editor, Byte and Switch

Organizations mentioned in this article: