Physically Immutable Optical Archive Libraries | Hacker News Reader

Physically Immutable Optical Archive Libraries

savartus.com

41 pointsthunderbong17 comments

So it's like a tape library, but for 1Tb optical disks?

I think a tape library still wins. LTO-9 allows 18Tb on one tape (without compression) and LTO-10 is 30/40Tb

And tape is going to be fine, as long as it's kept in reasonable conditions. No need to do clean rooms to store it or anything like that.

The tricky bit, especially for the small operater is keeping a working drive as long as you want to keep the storage.

Yeah, the back-compat story for LTOs is not great. They are going to try and improve it starting with LTO-10.

It was partially on purpose for the earlier LTO generations. The capacity/speed ratio was not too high, so you could just keep upgrading your library in the background every decade. It outweighed the complexity of designing backwards-compatible hardware.

But it's becoming less practical now. LTO-4 is 1.6Tb with 160Mb read speed, so each tape takes just about 3 hours to read. LTO-10 is 40Tb with 400Mb read speed, resulting in more than 24 hours of time to read one tape.

Drive availability is an interesting question in itself. Laser drivers are pretty complicated beasts themselves that require highly precise mechanics. Tape drives are much simpler, because the areal information density is about 3 orders of magnitude smaller than on optical media.

The mechanical stuff that loads and unloads the tape and keeps it in tension is crazy, but it's all just regular "macroscopic" parts that can be custom-manufactured if needed.

> Tape drives are much simpler

Maybe back in the 90's :)

LTO-9 increased track density, increased the total number of tracks by reducing their width and increased linear density.

The "how" certainly does not sound "much simpler" to me ?[1]

    The new, narrower track width was achieved through a combination of improvements in tracking
    performance, new, narrower tunneling magnetoresistive (TMR) readers within the head assembly
    (now <1000nm in width) and optimized writers. The narrower tracks also necessitated
    improvements to the media magnetic layer. Magnetic particles are Barium Ferrite (BaFe) as in LTO-8, but with
    improved characteristics to support the higher linear densities.

    Changes were also made to the tape substrate, with the thickness reduced by 7% from 5.6µm to
    5.2 µm. This in turn enables more tape to fit on the reel – tape length per cartridge has
    increased from 960m to 1035m. The new substrate also features optimized stability
    characteristics to mitigate tape dimensional variations that occur due to fluctuations in
    environmental conditions.

In addition the LTO Program's own technical paper describes how LTO-9 manages dimensional changes in the new substrate with a one-time-per-tape calibration algorithm to control 32-channel recording via closed-loop servo algorithms, plus a longer-block Reed-Solomon C2 code — together getting LTO-9's uncorrectable bit error rate to 10⁻²⁰, a 10× improvement over LTO-8[2]

[1] https://at.ingrammicro.eu/api/cfs/Icecat/PDF.ashx?l=en&s=957... [2] https://www.lto.org/wp-content/uploads/2022/08/LTO-UBER-Tech...

Sure, but that's an order of magnitude better than optical drives. Blu-Ray optical media use feature sizes on the order of 100nm (0.1um) and the read head needs to be able to move in two dimensions to keep it in focus.

I repaired a couple of LTO drives, and they are pretty approachable once you disassemble the Ruby Goldberg tape management devices.

A lot of LTO complexity also matters for _writing_, not reading. If you're an archivist trying to restore a tape, you don't need to worry about optimal tensioning, calibration, etc. You can also read the tape at a fraction of the normal speed so that your control loops can be much looser.

With that modern LLMs are capable of I'n seriously considering investing in at least M-DISC storage or similar to store stuff I don't want to lose :). It's unlikely that both iCloud and Synology get hacked at the same time, but the probability is higher than I could imagine before.

I burned some optical installation media for OpenBSD and Linux for the first time in a long time to have some RO boot media if another Internet Worm comes around.

If you are concerned with durability mostly in terms of the next few decades, I think S3 deep archive would be difficult to beat. Your data is offline by default. It would be logistically infeasible for a rogue agent to physically mount every tape for purging before someone would notice and stop the tape library.

The risk with any cloud is that while your data might be physically intact, you can still lose access to it for a variety of reasons (e.g. being born in a wrong country)

How does this compare vs Microsoft’s Project Silica? Available for purchase being the immediate differentiator, but curious how they fare technically.

https://unlocked.microsoft.com/sealed-in-glass/

Silica uses femtosecond lasers for writing and supposedly mostly ordinary silicate glass as the medium. This would probably mean more costly drives and severely cheaper media (Folio claims $3/TB now, $1/TB eventually. I wager Silica would be at least 10x cheaper). If Silica keeps its promises, we might be looking at archival prices cratering. But it's been in development for over a decade, I wouldn't hold my breath.

Really interesting would be R/W speeds.

I'm still waiting for the Microsoft’s Vacuum cleaner.

These don't seem really physically immutable, just protected by the firmware of the drives/robot against being written twice. With enough effort couldn't the same laser write _something_ over them again? Maybe optical discs need a hole that gets punched out or something, like the old floppy disk write protect. (And then make that control a physical laser power switch)

The video says they use 200GB Blu-ray media. Is this something new?

No. In fact, Sony had their own version of this that had three generations of media before getting shitcanned. I think it wound up going to something like 300GB discs - which is like 2.5x what consumers can do with BDXLs.

We where just talking about tape and other long term back up for Langrange Point Cloud. Fun to see the lasers.

I’ve always wanted to have a media format as described in Infinite Jest. This is in that neighborhood…

Reply on news.ycombinator.com