Run a Backup—Now.

Yes, it sounds trite, but we’re often surprised at how many companies don’t do their backups often—or as completely—as they should. We’re talking primarily about long-term tape backup. Despite what the hard drive vendors tell you, tape backup isn’t dead at all; it’s by far the most efficient way to store data for the long term.

Most of these downtime events are minor and easily or quickly corrected. But all events provides valuable lessons. A record needs to be kept to help assess the status quo and to enable a fact-based discussion with management.

Effective archiving amounts to sorting out the “storage junk drawer” Ridding production disks of unused, but important to keep data can free up as much as 40 percent of every disk you own! With a little effort, you can also purge orphan data and eliminate contraband data, recovering another 15 percent of disk space.

When implemented properly, virtualization centralizes capacity management, performance management and data protection management. It also breaks vendor lock-ins, reduces storage costs while improving Input/Output Operations Per Second (IOPS) and throughput. It also provides defense in depth for data.

When implemented properly, virtualization centralizes capacity management, performance management and data protection management. It also breaks vendor lock-ins, reduces storage costs while improving Input/Output Operations Per Second (IOPS) and throughput. It also provides defense in depth for data.

Many companies use software or hardware data mirroring—both synchronous (local/MAN) and asynchronous (distance/WAN)—but they don’t test to see whether data is, in fact, being replicated. The only way to be sure is to break the mirror and perform a comparison. This should be done as soon as possible.

Everybody wants to replicate across the WAN, but they fail to appreciate issues involved. It’s not the link speed, but the latency that matters; adding to distance-induced latency are network factors that can kill your recovery process. Suggestions: Try before you buy; break mirrors and compare data states; Remember, not all apps require High Availability failover or quick restore following an outage. Keep in mind the cost and, even if you do WAN-based replication, perform a tape backup anyway.

The main beef with tape backup is restoration speed. Most disasters could be stamped out with fast restoration of discrete files or data sets. That’s the appeal of Virtual Tape Libraries (VTLs). Used wisely, VTLs deliver fast restoration of deleted or corrupted files. Add in de-duplication for reducing capacity requirements on VTL disk—if your governance people are okay with it.

IT such as VDI (virtual desktops), VoIP (voice over IP) and G4 have potential roles to play in user recovery and disaster recovery planning. For example, virtual desktops could simplify user workstation restorations; IP telephony could eliminate Plain Old Telephone Service (POTS) redirection hassles; and G4 and G5 networks could simplify user connectivity to re-hosted applications and services.

All of these nifty ideas are pointless if you can’t get management to buy in to them. Management may be circumspect about spending money on continuity planning. So include data recover provisions in normal acquisition budgets. Just call it “enhanced value architecture”. To virtualize storage and to position value-add data protection services on the abstraction layer, call it “internal cloud storage with cost-efficient service scaling and high resiliency.” And keep tape in play to use it for data archives. Just call it “e-litigation and compliancy mitigation.”