Ever since LTO-2, and definitely since LTO-3, one of the biggest challenges has been designing a backup environment that can stream data to these tape drives. In fact, as I commented on Marc Crespi's blog "Is Disk With Deduplication Really Faster than LTO4 Tape for Backups?," one of the big advantages of disk is how forgiving it is of slow backup jobs. If you can get tape streaming, it is fast. The challenge is getting it streaming. Today this requires either a caching of backup jobs to disk first or a direct local backup.
If the LTO-5 spec would have done its normal doubling effect, I'm not sure how many data centers would have been able to take advantage of the performance boost. The advantage of slowing down on the performance numbers means the pricing of the drives should be less expensive. It is more economical to use the slower components.
When LTO-6 comes to market, we see a return to larger performance upgrades. This again makes sense. By the time we see LTO-6 come to market, much of the shift to either 10GbE or 8Gb FC should be well underway. Depending on how the timing works we may also see 16Gb FC by that point. We will have the infrastructure to move the data. Despite all the gains we are making in deduplication and compression technologies, the amount of data to be protected and the size of those components will continue to increase. In short, we will need that performance.
So where does disk as a backup target, the thing I normally write about, fit into all of this? We'll cover that in our next entry.