sapphyra

anarchist transbian robot girl

  • she/they

Hi! Just a cute trans girl on the internet. I have a plethora of random little hobbies and creative media I like to make stuff in, which I may or may not post when I have time. :3
This girl dreams of being a gamedev someday.
A little bit of a creature.

Going to miss Cohost so much when it's gone. :( Let's carry Eggbug in our hearts...


Discord
@missingfragment
You must log in to comment.

in reply to @OniLink's post:

lol i had a similar question--i use google drive, but 'course it's not the best way. dropbox is another good option. I've also tried making a (google) site with the videos, but it's going to take a while. it depends on what type of videos you have -- if they're just random youtube videos, cloud storage drives (rakuten drive, google drive, dropbox, 4shared, meganz, etc) are probably your best bet. if it's music (like what i have) try making a gsite or something. (if you need to embed video files into a google site, i would use catbox.moe. it's very easy to use and can make permanent links for files for FREE)

Unfortunately I'm not particularly trusting of cloud storage services for archiving. Especially Google given how prone they are to randomly shutting down various services. Then again, it could be a good potential method for storing an additional backup of the data... I'll see about working that into my plan, ty!

My current plan is to investigate reliable archive grade external drives, make a backup onto one of those to store in a completely different location, and then building a NAS with a backup RAID mode enabled to store the current drive in so I can still access it at any time. But I'm worried that I might be missing something when it comes to secure archiving.

The cheapest off-site backup I know of is Amazon Glacier. It's about a dollar per terabyte a month, for data you can wait 12-48 hours to access (the longer wait is cheaper, because retrieval is charged separately), and expect to keep in the service for 180 days or more. It's meant for business use, so it's not very user friendly, but it sure is cheap.

Compared to just buying a 4 TB hard drive and keeping it there, which is about $100 these days, it takes about 3 years to break even. Of course, hard drives are even more competitive with more data, 12TB hard drives are about $200 now, and then it only takes 17 months.

But I do also want to note that Amazon promises 11 9's of reliability on Glacier Deep Archive. In practice: your data is spread across at least three fully redundant data centers, with fully separate power and networking. As long as any two of those three data centers are intact, so is your data. (These data centers are within 60 miles of each other, so I suppose a very large bomb might take out all of them, but short of that, very little would knock out more than one.)

Within those zones, your data is also spread across multiple hard drives, which are replaced as they fail and restored from the copies. This is the risk that makes it only 11 9's, every now and then exactly the right set of hard drives fail and knock out a full set of data. This is much greater resilience, still, than you can expect from almost anything one person can do on their own. As Amazon once phrased it, if you store 10,000,000 objects you can expect a loss of one every 10,000 years. (I imagine they only removed that phrasing because it's simplified, what exactly counts as an object, a 128kb chunk?, etc, but the sense of scale is still about right)

Edit: regarding your concerns about cloud services getting shut down: Glacier is part of Amazon S3. That's the storage portion of AWS. AWS shutting down would be an incredible disaster for the entire internet. And it's profitable as is, unlike many cloud services. (Profitability is more common for business oriented services like this, but it's still not universal, Google's AWS equivalent isn't profitable, for example.)

As far as user friendliness, I've heard great things about the Arq backup software for handling keeping things in S3 for you automatically, I assume it has options for Glacier and such.

Edit 2: okay, cost of retrieval can be significant, on the order of ~$90 per terabyte. This is actually the same cost across all of S3 I think, but AWS users generally get the first 100GB free per month, and many users stay within that. I think you can transfer to AWS for free and grab that 100GB each month for free, paying only the AWS fees to host that 100GB long enough for you to download it. A huge hecking headache.

So this is heavily incentivized to be a secondary backup, I suppose, for restoring files that a less-reliable primary backup might have lost parts of.