Valhalla
Revision as of 22:43, 18 September 2014 by Mithrandir (talk | contribs)
This wiki page is a collection of ideas for Project Valhalla.
<SketchCow> Basically, we have this situation where we have stuff that is being threatened, and it's huge, and then it's either not so threatened or it's in a weird quantum state. So, this really stretches the bounds of what IA does. It's a huge amount of data, it's not likely to be overly touched if the originals are up, and IA will spend/lose a lot of money pulling it into their infrastructure. So maybe we can discuss actual, not pie-in-the-sky possibilities of what we can do to have some sort of not-IA pile of storage.
Join the discussion in #huntinggrounds.
Options
- Hard drives
- These would have to be live. HDDs decay quickly, and if they're not spinning, you can't detect failures.
- Possible software for this kind of thing; syncthing, Tahoe-LAFS, ...?
- Commercial/archival-grade tapes
- Consumer tape systems (VHS, Betamax, cassette tapes, ...)
- Vinyl
- PaperBack
- Optar
- Blu-ray: lasts a LOT longer than CD/DVD but should not be assumed to last more than a decade
- M-DISC"
- Unproven technology, but potentially interesting.
- Flash media
- Wears out quickly, not-so-good long term storage
- Soliciting donations for old flash media from people, or sponsorship from flash companies?
- Glass/metal etching
Non-options
- Ink-based Consumer Optical Media (CDs, DVD, etc.)
- Differences between Blu-Ray and DVD? DVDs do not last very long.
- BitTorrent Sync
- Proprietary (currently), so not a good idea to use as an archival format/platform
- Amazon Glacier
- Amazon Glacier seems like a a great idea, until you realize they mean 1 cent per gigabyte per month. This is $120 per terabyte per year. The transfer out of 100TB would also run over $10,000 the month its pulled from the system.
- Floppies
- "Because 1.4 trillion floppies exists less than 700 billion floppies. HYPOTHETICALLY, if you set twenty stacks side by side, figure a quarter centimeter per floppy thickness, excluded the size of the drive needed to read the floppies you would still need a structure 175,000 ft. high to house them. Let's also assume that the failure rate for floppies is about 5% (everyone knows that varies by brand, usage, time of manufacture, materials used, etc, but lets say 5% per year). 70 million of those 1.4 trillion floppies are unusuable. Figuring 1.4 MB per floppy disk, you are losing approximately 100MB of porn each year. Assuming it takes 5 seconds to replace a bad floppy, you would have to spend 97,222 hrs/yr to replace them. Considering there are only 8,760 hrs per year, you would require a staff of 12 people replacing floppies around the clock or 24 people on 12 hr shifts. Figuring $7/hr you would spend $367,920 on labor alone. Figuring a nickel per bad floppy, you would need $3,500,000 annually in floppy disks, bringing your 1TB floppy raid operating costs (excluding electricity, etc) to $3,867, 920 and a whole landfill of corrupted porn. Thank you for destroying the planet and bankrupting a small country with your floppy based porn RAID." (source)
From IRC
<Drevkevac> we are looking to store 100TB+ of media offline for 25+ years <Drevkevac> if anyone wants to drop in, I will pastebin the chat log <rat> DVDR and BR-R are not high volume. When you have massive amounts of data, raid arrays have too many points of failure. <rat> Drevkevac: I work in a tv studio. We have 30+ years worth of tapes. And all of them are still good. <rat> find a hard drive from 30 years ago and see how well it hooks up ;) <brousch_> 1500 Taiyo Yuden Gold CD-Rs http://www.mediasupply.com/taiyo-yuden-gold-cd-rs.html
"<Drevkevac> still, if its true, you could do, perhaps, raidz3s in groups of 15 disks or so? <SketchCow> Please add paperbak to the wiki page. <SketchCow> Fuck Optical Media. not an option;. <Drevkevac> that would give you ~300GB per disk group, with 3 disks
Costs
These are just estimates. Calculation: $/TB = Total Cost / Total Capacity
Purpose | Cost (USD) Per TB |
---|---|
Tape Media | $36.4 |
Hard drives | $43 |