When vendors compare the performance of disk backup systems with the performance of tape, they often will highlight the area in which their particular product has the biggest advantage and ignore situations where performance doesn’t matter. IT planners looking to optimize backup and recovery performance, as well as cost, should concentrate on designing an architecture that strikes the right balance between performance and cost. Creating a balanced architecture requires exploiting the advantages of each technology while compensating for its weak points.
Disk backup vendors tend to claim restore performance superiority, but that superiority is primarily for single file or folder-level restores, not for entire systems. Many disk backup vendors leverage technologies such as deduplication in order to compete with prices of tape based systems. However, deduplication impacts performance, especially in large restoration scenarios. In fact, ExaGrid a disk backup vendor known for first storing data to a non-deduplicated area of the disk system claims a significant restore performance advantage over deduplication-only systems, especially as restores get larger.
Tape-based libraries actually have an advantage in performance when doing large full system recoveries as are common in disaster recovery situations. It is also necessary to keep in mind that most organizations are protecting mission-critical systems with something other than backup, most commonly a high-availability solution like replication. For mission-critical applications, the speed of backup recovery is less of an issue since the primary recovery point, at least in a disaster, is not even part of the backup process. Additionally, the cost advantage of tape may enable the organization to leverage flash storage as the backup architecture’s initial ingestion point, with tape serving as an economical secondary storage tier for multiple onsite and offsite copies, making the total architecture significantly faster and more secure for both backups and recoveries.
It’s All About Architecture Design
The job of the IT planner is to design an architecture that best meets the organization’s current needs for capacity and performance while anticipating future needs – all while staying within budget. If organizations design their backup architectures as recommended in our first blog, the first tier is high performance which can be either flash or disk. In both cases though, deduplication is not as necessary since that data is tiered off to an affordable tape tier. Our recommended best practice is copying the data directly to the tape tier as soon as the backups to the ingest tier are complete. Copying data to tape immediately, enables the organization to have redundancy of backups faster. Also, when the time comes to release older data from the first tier, the backup software makes no additional copies. No deduplication on the ingest tier not only delivers faster ingestion it also delivers a faster secondary copy to tape. Regardless of the organization’s stance on tape, no IT professional should consider data protected until it is on two or three separate systems, preferably on two different media types and in different geographic locations with one being offsite and offline.
Conclusion
The short answer is that tape is faster than disk once it reaches its full speed. The problem is, and the reason why many disk backup vendors claim that disk is faster, is that it takes a while to get tape going at full speed and it also take a lot to saturate tape to its full potential. When the infrastructure can’t feed tape at its full potential, the technology has to slow down and wait for more data. Given the speed of the modern tape drive, many organizations have trouble creating an architecture that can maintain tape saturation.
The key for the IT planner is to realize that tape is faster and design a backup architecture to keep it properly fed. The architecture described in the last blog is an ideal method to do that. With the tape library attached directly to the ingestion system, the chances of keeping tape running at full speed increase greatly, especially if the ingestion system is not using deduplication.
George Crump is the Chief Marketing Officer at VergeIO, the leader in Ultraconverged Infrastructure. Prior to VergeIO he was Chief Product Strategist at StorONE. Before assuming roles with innovative technology vendors, George spent almost 14 years as the founder and lead analyst at Storage Switzerland. In his spare time, he continues to write blogs on Storage Switzerland to educate IT professionals on all aspects of data center storage. He is the primary contributor to Storage Switzerland and is a heavily sought-after public speaker. With over 30 years of experience designing storage solutions for data centers across the US, he has seen the birth of such technologies as RAID, NAS, SAN, Virtualization, Cloud, and Enterprise Flash. Before founding Storage Switzerland, he was CTO at one of the nation's largest storage integrators, where he was in charge of technology testing, integration, and product selection.
StorageSwiss.com - The Home of Storage Switzerland 
Tape rocks and don’t let anyone tell you or sell you otherwise. Density, data quality, archival quality and speed, when the infrastructure is architected correctly, cannot be beat.