Full backup provides the most complete method of recovery from a system outage or data loss. However, IT departments have been forced away from performing them at regular intervals as data size began sprawling beyond the capabilities of their backup systems. Incremental and Differential backups were created to fill in for the performance impact and timeliness of backup solutions. However, the real challenge that you’ll face during an outage will be the time that a restore takes, which will require a full backup, and then incremental backup restoration.
Another hurdle is corrupted backups that force you further back in time for your Recovery Point Objective (RPO). Virtual systems have added another wrinkle that has caught traditional backup vendors off guard and created a new category of backup solutions. However, even these newer backup solutions that target the virtualization marketplace are still doing a “backup” of the virtual machine. Thus, the RPO is still challenging.
Now we get to the real pain point of backup-based solutions: restoration. Restoration of incremental backups can easily take hours or days to complete before your end-users can continue working. Therefore, your Recovery Time Objective (RTO) is really the true enemy of productivity after an outage; you just don’t feel the restoration pain point daily like you feel the backup pain point.
The time alone to copy data off of backup media across the network has grown along with backup windows. When was the last time that you copied 100 GB across your network and timed it? How about 500 GB, 1 TB, 3 TB? Now what about multiple server’s worth of data competing for the same network? It takes hours to move the data that our servers are holding today, and your business is down until the copies complete at the very least.
As we continue to see further data sprawl and virtual server sprawl, it’s becoming readily apparent that the backup solutions that could barely meet our needs yesterday can’t scale to meet our appetite for growth now and in the future. Backup windows (RPO) and restoration (RTO) resemble a house of cards, with your company’s ability to do business sitting at the top and your reputation sitting right beside it.
IT departments need to get back to performing full backups on a regularly scheduled basis, rather than the exception. The only question is how to perform full backup fast enough in today’s data center so that we don’t have to settle for lackluster incremental backup and the time and risk associated with slow restoration. In-addition, it’s agreeable that any solution will have to be at a price point that’s somewhat competitive to your existing solution, otherwise management will likely hold back budget until the race to backup becomes a real problem. That’s why your disaster recovery project keeps getting pushed back year after year, right?
Nimble Storage takes a different approach to backup that leverages disk snapshot technology, providing near instant backup. Nimble’s patent-pending CASL file system combines primary storage with disk-based backup storage to provide a complete storage and protection solution. It includes inline compression and redirect-on-write snapshot technology that greatly reduces the amount of physical storage required to store data and its backups. Nimble snapshots eliminate backup windows and allow many more point-in-time backups (RPO) with much longer retention than existing methods. They also eliminate production impact and network bandwidth to copy backups to different storage media. Most importantly, recovery time (RTO) is greatly reduced because there is no longer a requirement to perform restoration from backup media to primary storage; you can simply mount and go.
So how are you planning to back up VMware, Hyper-V, SQL Server, Exchange, and the rest of your data center when you start exceeding your incremental backup window? We look forward to hearing your stories.