I have gone on and on about this topic in particular for the better part of 5 years and see so many different arguments on both sides that made me want to talk about this again on this newly revised blog. First I am a proponent of doing the right thing and that being said when I talk to different companies about what their focus is around DR it is not focused on tape. When I talk to companies and CIOs about long-term retention of data, tape comes up for storage of data offsite but then gets pushed to the side when mentioning things like object storage with WORM technology and geo-redundancy at pennies per GB.
Now I know that there are awesome stories about how data was recovered from tapes from the 60s and 70s like the Apollo 11, 12 and 14 missions from a donated IBM 729 Mark V tape drive but still think that all the money, research and testing that went into how these tapes were written to and cared for are the keys to how this recovery was possible. Now flip to current date and the cost of doing this in today’s IT environment is more hassle than it’s worth. New technology does not always trump the old way of doing things but there is a point at which you need to evaluate how we treat and store data. If tape is an absolute must, then I ask why and is it because of the safety of storing this data in an offsite vault for long-term archival or is it more compliance?
If I store data on disk and it is enterprise disk packed into an appliance that uses a battery backed up RAID controller with sufficient cache that can replicate to an offsite location that is also then replicated again for a triple down on how many times the data is stored and in different regions, would you feel safe about storing your data on disk? If I then have the ability to fail over and leverage this solution in an HA (High Availability) fashion with automation of where the data is being backed up to (in other words, if one site appliance fails the other picks up with little to know configuration change), would you consider moving away from tape? How about cost… Triple down on data sounds costly but with the right solution it can be very cost-effective.
Take a hybrid approach to managing this data where you own ½ the solution fully on site with offsite public cloud options where I can leverage economies of scale at a low monthly cost. Or even build your own private cloud with cost effective appliances getting the cost down below tape and offsite storage of tape while giving you the speed of disk and ultrafast recovery times improving IT efficiencies. Couple either of these options and more with object based geo replicated storage and we now have a way to store long-term data at pennies per GB and have your data safe for decades.
Ok, ok… I’ve gone on and on about disk based solutions but what about tape. Well tape is not dead, not yet at least. I know I use one right now as a coaster… Well all joking aside, tape is not dead and there are several folks who would agree with me on this. Check out the Storage Expert’s LinkedIn group and search for tape is dead. You’ll find quite a few different views and all valid, well most of them anyway. The key thing here is do the right thing for your organization, look at all the facts (don’t listen to the FUD), line up the costs, and make the informed decision .vs. sticking with a technology just because you have always done it that way.
In summary, tape is not dead and will not go away for my guess another decade or so but it is slowly going away from what most people use this technology for. Think archive of government data… well there I go saying government and we all know the motto “if it ain’t broke fix it till it is” but this is a viable solution for tape. For now tape is still going to remain in the market but for how long one can only guess.