?
According to a recent?report from Forrester Group," Business Continuity and Disaster Recovery are top IT priorities for 2010 and 2011." This should come as no surprise as extreme weather conditions are at an all time record high around the world.? Yesterday, the East Coast was rocked by a surprise 5.9 earthquake. ? This year alone, the world has seen record snow falls in Europe and the eastern seaboard, flooding in the midwest, hurricanes and tornadoes.?? Extreme weather produces extreme disasters.? One IBM customer who has literally weathered the storms and recovered from a natural disaster is?Richard Cosby, Systems Administrator for Estes Express Lines. In 2004, Estes was hit hard by an unforseen natural disaster, Hurricane Gaston.? The city where Estes is based, Richmond Virginia, received fourteen inches of rain in just a few hours.? The result was massive flooding, which seeped into the Estes datacenter.? "Every piece of hardware, including our seven IBM System i servers, all of the SAN [storage-area network] storage, the fabric switches, routers, network equipment, phone system, was flooded, " explains Mr. Cosby.
So how did Cosby recover from the Hurricane and rebuild?? Fortunately, he had been protecting his data along.? Recently I sat down with Mr.Cosby to find out how he protected his data. He very graciously answered by questions, and offered some best practices that may be of interest to Storage Community members.? Following is my interview with him.
Q) Tell us a little about your position at Estes Trucking? What are your responsibilities?
A. I'm the IBM i & Storage System Admin. and datacenter manager. I'm responsible for our line-of-business systems and DS8000 SAN storage servers.
?Q You really had to learn about Disaster Recovery the hard way. Were you in the habit of regularly backing up your data nightly before the Hurricane Gaston Hit in 2004?
A. Before the hurricane disaster, we backed up production date nightly, using a combination of DS8000 FlashCopy and then a save to LTO tape (disk to disk to tape). This process allowed a full copy of production data to be point-in-time flashcopied and then made available to our backup LPAR. This process does not interrupt our production LPAR jobs.
?Q I read that you spoke at a user group and that you mentioned that it only takes you about 18 minutes now to do a complete system back-up using your Power HA System. Is that true? Has Power HA dramatically cut your back-up time? How many people are involved in doing your data back-up?
A. The nightly DS8000 Flashcopy process takes less than three (3) seconds to complete. Then, the data volumes are made available to our backup LPARs. Without this process (disk to disk to tape), we could not perform a point-in-time nightly backup, since the save to tape on one LPAR executed for seven (7) hours for the tape save, 8TB.
No staff is involved. Job schedule entries execute and monitor the nightly saves to tape.
Q. Do you still use Tape Back-up? Have you read all the controversy about whether or not Tape is dead? I know Google uses Tape to back up all the Gmail data and it has saved them.
A. We use LTO tape backup for all backup tasks. LTO is very inexpensive, requires no power for data cartridge storage, not subject to malware corruption, hardware encryption, transportable. Data Dedup is far more complex and expensive for backup needs. Our datacenter was flooded and all hardware was destroyed. A Dedup process would not have provided us a DR backup solution, as implemented, without a second site and networking cost.
?Q. How important was Flash Copy to your back-up and recovery efforts?
A. Flashcopy allows us to make point-in-time backups without any interruption to production, so very important.
?Q. Can you outline Best Practices that your have implemented for Back-up that doesn't interrupt your current production?
A.Yes, that's very simple:
- Make a Flashcopy
- Make the copy available to our backup LPARs
- Copy to tape, does not interrupt production
?Q. Did you divide your data into Tiers, ie. Tiered storage so you know what to back-up or prioritize first?
A. The Flashcopy target is reusable nightly. Tier 2 is LTO tape
?Q. For users who haven't yet experienced a disaster, any advice on how to put a DR plan in place?
Understand what an hour of outage costs, both direct loss of business and long-term customer confidence
Once Disaster Recovery is in place, TEST.
Remember, IT?Disaster Recovery is only one element of a corporate Business Continuity plan.
?Q. Finally, how do you estimate your storage costs? Do you use any packaged Storage Calculators?or homegrown tools?
A. We use homegrown tools, since we consider all benefits of SAN storage, not just direct costs compared to traditional internal HDD costs.
?To learn more about Disaster Recovery, and how to get prepared, listen to the IBM Webcast on Sept. 21st entitled, "Planning for Disaster Recovery in a Power Environment: Best Practices to Protect Your Data."
Posted 08-24-2011 12:49 AM by Mary Hall
salmonella bst miranda kerr d.b. cooper d.b. cooper chocolat haiti
No comments:
Post a Comment