After adding a 4th hard drive to my home server today, bringing the total storage space up to an unexpected 20 TB, I’ve been thinking a lot about backups and what my dwindling options are as this beast continues to grow even larger and awesomer than I would’ve ever expected only six short months ago…
I’m definitely well past the “it’s just TV and movies” phase and am now finding myself much more in the “I love this thing, and it would be a huge pain in the ass to replace!” phase instead! The trouble I’m gauging, though, is how to effectively manage a backup that big without spending a small fortune or driving myself absolutely insane!
Ironically, my original plan when I bought the two 4 TB drives to start this project was that one was supposed to be a backup of the other, however by the time I started getting my hands wet, not only had I concocted a plan to fill both drives that I already had, but I swiftly had another 6 TB on the way to give me “some wiggle room,” too.
Well, now that said wiggle room has flown the coop and I coincidentally just added another backup-less drive to my existing server, I’m starting to think that I need to reassess my options … first and foremost, because there’s literally not any space left in the case I have today for more drives to backup to anyways… 😛
Online backups are pretty much out, and note that I’m only talking about media server backups here – documents, photos, etc… are now triple backed up (something I need to write about one day) – so vital stuff is absolutely safe. It’s just my collection of Marvel movies and twenty-some-odd seasons of The Simpsons that I’m concerned about here today!
Anyways, online backups are out primarily for two reasons:
- Cost – A real backup service like Amazon S3 would be at least $200/month for 20 TB of data, even at their cheapest Amazon Glacier prices, and I’d be hesitant to push the luck of any of the “unlimited plans” that folks like Crashplan offer with that volume of data.
- Sensitivity – Let’s see, how do I put this gently??? I may have discovered while I was ripping my DVD collection that it was far more timely to just download copies of them off the Internet, and so even though I know that I have a huge crate of discs in my garage that justify such a huge library of technically illegal content, it’s not exactly something I’d entrust to a 3rd party who might be legally inclined to disagree… 😉
This pretty much limits me to local backups, which sucks because Florida gets hurricanes and whatnot occasionally, but there’s not much I can do.
So when looking at local backups, my latest plan up until this afternoon was going to be to essentially build a backup server that would be identical to my media server, except that its entire job would be to occasionally wake up, make copies of everything on the primary server, and then immediately go back to sleep.
This plan made sense until I started roughing out costs in my head for the next generation of my media server because realistically, my next expansion will need to put it into a proper rack mounted case, along with I’d also like to throw some beefier hardware at it like a dual-chip motherboard, multiple gigabit LAN card with bonded channels, and a RAID setup for the drives for better throughput and redundancy.
That alone is going to be expensive because I’ve never done RAID before and I’m realizing that I’ll need enough drives to build the entire array at once to have someplace to migrate my data over to, but aside from a healthy splurge, it’s still not terrible … until I realize that I have to double everything if I want to run a backup server alongside of it like I had originally been planning! 🙁
That all said, excessive flashy lights aside that a rack full of noisy hard drives will create, I had an interesting idea today that might change most of that for the better and that’s this … why do I need to have a live backup at all for data that’s almost NEVER changing???
Again, these are movies and TV shows, not working documents or even photos that I’m editing, so once I’ve got a ripped season of The Simpsons from Blu-Ray, that’s pretty much it until they re-release everything again on hypercube or holographic projection or whatever new-fangled media to get me to buy 25 seasons of cartoons all over again they come up with next!
So why not, I thought brilliantly while scrubbing myself clean in the shower, just take everything that’s static – pretty much every movie, and all of the older TV seasons – and just stick them on hard drives, and then put the hard drives in a waterproof case to throw in the closet or wherever?
It would save on moving parts because they’re literally only going to get written to a handful of times, it sort of works in that whole elements thing if they’re kept someplace safe, and … I think it kind of makes sense for a lower cost solution without all of the bells and whistles that frankly are kind of frivolous anyways…
I’d still need a system to keep track of what’s been offloaded to the drives for when it comes time to fill a new one, but I’d guess that software probably already exists for tape backups that could be used. I guess I’d have to test them every couple of years just to make sure that they’re still alive, but that could be part of adding new data to the collection which I’m sure would be a manual/annual effort or something, anyways.
I realized as I was completing some seasons of shows that I’d had a long time ago, but lost to a hard drive failure that really once they’re completed, they’re not going anywhere at that point, so maybe I’ve been overthinking this whole massive backup situation when instead I can just drop the one-time cost for another set of drives and a hefty, padded case to store them in and just be done with it!
Problem solved – now can I go back to playing in my data? 😉