we finally added up what our tape infrastructure was actually costing per year, and i wish we hadn't.
this is one of those situations where the number wasn't surprising in hindsight but it still felt like a punch when we put it on paper.
tape was the "inexpensive" choice that we've had around for quite some time now, and individually, the costs seem relatively insignificant. hardware maintenance agreements. the replacement of the LTO drives on a periodic basis. licensing costs for the backup management software layer. offsite vaulting costs that increased silently every year. the manpower involved, not as part of a specific job, but rather split between two positions.
add it up across a year and it wasn't what i'd call small. add it up across five years and compare it to what equivalent cold cloud storage actually costs now and the math stopped making sense.
the part that got me was the migration cost sitting at the end of the road regardless. at some point the current LTO generation goes end-of-life. you migrate to the next one, which means new hardware, new software, staff time, and a verify-everything cycle. that cost doesn't go away by keeping the tape infrastructure, it just defers.
we operated with tape ark for the migration piece when we eventually made the call, partly because doing it in-house was going to cost more in staff time than just outsourcing the whole thing. the economics of that decision were cleaner than i expected.
not saying cloud is the answer for everyone. but if your org has never actually calculated total tape TCO including the deferred migration cost, the number might shift how you're thinking about it.
has anyone else done this calculation recently? specifically asking about the numbers for different org sizes.