Two things are important after your disaster recovery plan is in place: First, do testing. Then afterwards, do external audits so someone can point out any holes. How often would you recommend people perform testing on their disaster recovery plans?
We usually say testing should be done once a quarter, and I'd say audits once a year. On a disaster recovery plan, testing is particularly important. You might find your UPS can run the computers, but it doesn't cool the room when everything else shuts down. Or, you can't get your tapes back because they are at a remote site. What other steps would you advise to prepare for a disaster?
Companies can certainly go out and try to get an external expert to help them, if they don't have the internal expertise. They could find a partner to work with them, and just about every hardware company could come in and help you. There's also companies like Glasshouse, who may be a little more independent than inviting in a StorageTek, for example. How do IT managers justify to upper-level management the need for an external audit of their disaster recovery or backup plans?
The average downtime costs a company about a million dollars an hour in [lost] revenue. In the energy industry, an hour of downtime costs $2.8 million an hour. For manufacturing companies, it costs $1.6 million For chemical companies, the loss is
A lot of companies feel like they have a plan in place, but they haven't done the testing and they haven't done the auditing. I used to run a small data center, and we thought we had a pretty good disaster recovery plan in place, [until] we tested and cut power to the building. Then, we realized the HVAC needs to be on the UPS as well, or else the boxes would shut down in 20 minutes from overheating. How do you view the blackouts that recently impacted the Northeastern United States?
The power outage was sort of a wake-up call for a lot of businesses to double-check their disaster recovery plans and make sure everything was working. Your survey shows a lot of people still feel vulnerable about their DR and backup strategies.
That's right. About 65-70% feel they're vulnerable in some way. That number seemed high to me. It might have to do with who we were talking to -- people working right on the systems. CEOs might give you a better number. But, these are pretty honest results. What do you advise people for where tape and disk are best suited?
We tell people that each application is different. You have to assess your environment, your application, and see if disk is appropriate. If you have open files, you may need to make a snapshot copy of them first. But, just because you've done that, doesn't mean you're protected. You have to secure that data still and get it offsite.
We have this council with 40 end users. The group we [brought] in were high-end users. When we talked to them about disk-to-disk backup and not continuing on to tape, they looked at us as if we're nuts.They ask us, 'Why would you move it from expensive RAID disk to inexpensive ATA-type disk and not ever back it up to tape?'
On the disk versus tape debate, what do you say to the Gartner Group's recent statements that tape will be obsolete by 2008?
Results of this survey confirmed once again that tape continues to be vital, and that disk and tape are complementary. We said all along that [disk and tape] are really complementary technologies. People have been saying tape's been dead for 20 years. But, there are a lot of companies out there like IBM and StorageTek that are still investing in tape technology.
Having your data on a spindle, rotating media, without securing it in some way, is going to do you little good in the event of a fire or other disaster. It needs to be removable and able to be secured off-site.
Why don't more companies have a disaster recovery plan in place yet -- in the wake of 09/11?
This is probably skewed to the smaller companies. At home, for instance, I don't back up my personal computer every day. If I lose my hard drive and I can't get my data back, I guarantee you I'd back up every day. What we're saying here is, be proactive. Don't wait for a disaster to happen. What plans do you have for any future tape products?
First thing, we need to continue to educate end users on the importance of data recovery and planning. As far as tape technology goes, if you look at the migration path, the cost per gigabyte will continue to be as aggressive as ever in regards to competing technologies. We see tape being around for a long time -- as do HP, IBM and StorageTek. What do you say about products that emulate tape?
A good example is the Quantum DX30. There may be a role for that, but I think Quantum would even agree, that it's not supposed to replace all tape. You [might use that to] back up to your incrementals, but still use tape. Some information we've seen is that the DX30 has not been that well accepted in the market.
Even with virtual tape, from a StorageTek or an IBM, the whole idea there is to more efficiently store on tape, and take the burden off the main CPU and hard drive.