One small step for tape...One giant leap for VTL

Tom, I can see your

Tom,
I can see your point(s), as many see things in the same light on this topic.

In a nutshell, there will always be many factors which come into play within the organization when considering VTL and each and every one will be different.

Outside of death and taxes, there are two things that are certain in our chosen field...there will always be 24 hours in a day and data will always continue to grow.

SO, to respond to your closing line, I would say most people look to VTL as a solution so they can drop their backup window(by as much as 70%) and add other functionality(such as replication and PiT) without too much investment, nor a forklift, nor retraining the backup team.

You can retool if you like, but I don't see media errors, media changes and a zillion small files smoking down the pipeline being fixed by the newer libraries anytime soon.

How expensive are these VTLs

How expensive are these VTLs (including licensing) compared to purchase of a functionally equivalent size non-vtl backup disk array (or just more disk for disk backups [snapshot, or any other disk to disk backup method] on the same array, if one is then immediately going to tape)?

It seems like VTLs are mostly a bandaid for shops who are unable (or unwilling) to retool their backup processes. A bandaid that leaves shops yet again at the mercy of vendors - in order to properly do DR, the VTL must be offsite, VTL disks must go offsite, or tapes must be made of the VTs. In the first two cases, much faster and simpler DR is facilitated with a conventional array. Making real tape copies of VTs means increasing the complexity of one's tape librarian processes, and probably requires support of (and another license payout to) backup software vendors - if one is relying on one of the big commercial packages.

I guess I'm missing something because, after my initial interest, I can't think of a compelling reason to use a VTL.

My experience with VTL

My experience with VTL licensing is not with the number of tapes but rather the nunber of jukebox slots. I've been using EMC's CDL (CLARiiON Disk Library) for 8 months. We have the ability of creating virtual tapes that are 500GB each by emulating Sony SAIT-1 drives. We're licensed for unlimited slots which, although pricey, gives us the flexibility of creating as many tapes as necessary.

We also clone from virtual to physical daily and that is our pain point. It's getting increasing harder to move 9 plus TB from virtual to physical in 22 hours (we want a 2 hour maintenance window for our environment).