The whole point of backup is being able to recover the data quickly. Zetta’s Cloud Storage and Data Protection 2010 survey found that 10% of organizations had a Recovery Time Objective (RTO) of less than an hour and another 34% had RTOs between one and four hours.
Of course, having a tight RTO and meeting it are two different issues. The same survey found that only 35% of the companies were meeting their RTO consistently (75-100% of the time). 38% were failing to meet their RTO even half the time.
However, there is a huge difference between meeting RTOs for companies that use cloud-based disaster recovery vs. those who do not. In October 2010, Aberdeen Group surveyed more than 100 organizations, 45% of which were using cloud storage and 55% which were not. In its report on the survey — Small and Mid-Sized Organizations Gain Disaster Recovery Advantages Using Cloud Storage — analyst Dick Csaplar wrote that while both groups reported similar RTOs — 12 hours for those with cloud storage and 13 hours for those without — those using cloud storage hit their RTOs 100% of the time, while the others only met their slightly looser objective 80% of the time. The gap was even greater when looking at the average length of downtime per event: 8.0 hours for those without cloud storage vs. just 2.1 hours for those using cloud storage.
“The number of reported DR events was very similar between Cloud users [2.5 events per year] and non-users [3.1 events per year], indicating that while the cloud architectures might offer better resiliency and fail safes against minor technical glitches, they can no more protect against widespread disaster and the machinations of fate than any other service,” wrote Csaplar. “However, the time required to recover from such downtime events for Cloud users was almost four-times faster than for non-users. The eight hour recovery times for non-users suggest a heavy reliance on traditional forms of emergency tape backup, as this eight hour time frame is consistent with the amount of time required to recall an archived tape and recover the missing data from the streaming media.”
Disaster recovery, though, is just one use of backups. Restoring accidentally deleted or altered files is a routine daily occurrence, and tape is too slow for that as well. While disk storage, whether in house or from the cloud, allows instant file restoration, restoring from tape requires determining which tapes contain the data, retrieving them from the tape archive and loading each of the required full and subsequent incremental backup tapes needed to ensure the correct file version is restored.
A better approach is to use enterprise online backup and disaster recovery such as provided by Zetta Data Protect. With the Zetta solution, a small piece of agent software called ZettaMirror replicates and keeps synchronized local files and databases to the Zetta site. Through a familiar web browser or mapped drive interface, users can find and instantly restore their own deleted files without having to contact IT. And, since Zetta stores file history though snapshots, it is easy to restore any earlier version of the file, even if it was from years ago.
By cutting the restoration process from hours to seconds, Zetta Data Protect makes it easy to meet even the strictest RTOs.
The science of backup has been with us for decades, though perhaps it seems at times to be more art than science. Modern business environments, new technologies and sheer data volumes have made the task of backup (and recovery) ever more challenging.
I’m in my seventh year working on cloud-delivered backup solutions, and I’m still struck by the number of times I hear potential customers asking if they can use cloud backup solution to do a Bare Metal Restore (BMR).
According to 
RSS