Armory long term cold storage concerns

[rax]

Hello everyone!

It's been quite some time since I learned about "cold storage" for the first time, but I've always been too lazy to try and figure out how it actually works (besides I don't own that many coins). That changed last week when I stumbled upon armoryguide.com (great resource!) and spent the last weekend playing with the program.

I'm pretty happy with the result. I recycled an unused old laptop which now acts as the vault, with its offline Armory wallet. My everyday PC now runs Armory too, and has both a regular wallet and a watch-only copy of the offline one. I stored away paper backups for both of them. I moved some coins back and forth between them, testing both regular and offline transactions. I even nuked the laptop and later restored the offline wallet to see if I could recover the coins it still held. So far soooo good.

All in all I'm almost ready to move the main bulk of my web wallet into cold storage and forget about it, but there is something that still bugs me. Given the current trend of the blockchain size growth, I'm pretty confident that in a few years from now I won't be able to afford running a full bitcoin node in my computer. That means I won't be able to run an online instance of Armory. And that means I won't be able to manually create and broadcast offline transactions anymore, so the stored coins would be stuck forever!

Are my assumptions correct? If so, are there any web or lightweight clients today that support the creation of raw, unsigned transactions compatible with Armory?

[picobit]

If that happens (and I agree that is a worry!) they you have two options.

1.  Hope that Alan has finished the lightweight version of Armory not needing a full node.

2.  Export your private keys one by one from the offline wallet, import them in another client, and move the money like that.

The latter will always be possible, so there is a plan B if everything else fails.

[rax]

Thank you, picobit! I certainly did not think about option 2. As long as I hold its private key I can restore an offline wallet elsewhere, a web client for instance... Even though by doing this I would lose the benefits of cold storage. Let's hope for 1, then

[7queue]

How would you not be able to afford running a full node & armory?

What am I not seeing that you are?

8 )

[rax]

How would you not be able to afford running a full node & armory?

What am I not seeing that you are?

8 )

If you visit the link to the blockchain size chart that I provided above you'll see that 3 years ago the blockchain was just about ~450MB. Today it's fast approaching 30GB and I'm afraid that at some point in the future I'd need to buy dedicated hard drives just to store my local copy of it. And personally I'm not willing to go that far.

[josephbisch]

Plus Armory uses more space for its database, so you need more space to run Bitcoin Core + Armory than for just running Bitcoin Core.

[7queue]

How would you not be able to afford running a full node & armory?

What am I not seeing that you are?

8 )

If you visit the link with the blockchain size chart that I provided above you'll see that 3 years ago the blockchain was just about ~450MB. Today it's fast approaching 30GB and I'm afraid that at some point in the future I'd need to buy dedicated hard drives just to store my local copy of it. And personally I'm not willing to go that far.

Oh, I thought it might have been something else I was missing.

I just put two separate drives in a system, one for each of bitcoind and armory db's, did this at the get-go in anticipation of growth as transactions increase.

I tried a full node with in-bound connections at my residence but the annoyance of attacks prompted me to move that out right-quick!

I think coinbase has a vault option, have to check on that and see what they offer for long term storage.

8 )

[rax]

Yes but one of the main points of cold storage is not having to trust a third party with your coins, isn't it?  

Plus, be aware that this is a 66x size increase in just over 3 years, and we are reportedly still in the early stages. If Bitcoin is ultimately successful only big companies will be able to afford the kind of facilities needed to store the whole blockchain. We're talking entire HDD farms here, Google style.

[7queue]

Yes but one of the main points of cold storage is not having to trust a third party with your coins, isn't it?  

Plus, be aware that this is a 66x size increase in just over 3 years, and we are reportedly still in the early stages. If Bitcoin is ultimately successful only big companies will be able to afford the kind of facilities needed to store the whole blockchain. We're talking entire HDD farms here, Google style.

True, if it goes over 4TB then I'm out, hopefully by then there will be some options available for us to utilize.

8 )

[doug_armory]

Yes but one of the main points of cold storage is not having to trust a third party with your coins, isn't it?  

Plus, be aware that this is a 66x size increase in just over 3 years, and we are reportedly still in the early stages. If Bitcoin is ultimately successful only big companies will be able to afford the kind of facilities needed to store the whole blockchain. We're talking entire HDD farms here, Google style.

I understand where you're coming from. I also feel that fear over the blockchain size is, to a certain extent, exaggerted. In the past year, the blockchain has expanded to ~25 GB, with ~14 GB coming in the previous year. In theory, if blocks were maxed out and the 1-block-every-10-min. target was hit, that would be ~52 GB. Let's further assume that blocks are actually mined every 5 min. due to continued explosive growth in the mining sector. That would be ~104 GB. On top of that, we'll assume Armory will always need a DB roughly the size of the blockchain. So, we're at ~50 GB right now, with potential for ~200 GB/yr growth if Bitcoin growth really did take us to the moon.

People here have said they're out when the blockchain hits 4 TB. Good thing that such an event probably won't happen for at least 20 years. Or, with Armory, at least 10 years. 6 TB hard drives are already out. By 2025, I'd imagine we'll be somewhere in the double digits, with 4 TB drives being the bare minimum available in shops. In other words, storing the basic chain won't require a farm. A fast connection and a lot of time (or somebody mailing you a drive) if you're starting from scratch, yes, but not a farm. Perhaps a farm would be needed if somebody was trying to track all the sidechains and such should sidechains ever get integrated into Bitcoin.

Yes, the blockchain is getting pretty large, and there is a need for lighter solutions. Yes, it's still possible to manage the blockchain from home and will continue to be so for a long time, although it will admittedly be less and less convenient over time. We'll be keeping an eye on the blockchain size and will react as is appropriate.

[7queue]

Yes but one of the main points of cold storage is not having to trust a third party with your coins, isn't it?  

Plus, be aware that this is a 66x size increase in just over 3 years, and we are reportedly still in the early stages. If Bitcoin is ultimately successful only big companies will be able to afford the kind of facilities needed to store the whole blockchain. We're talking entire HDD farms here, Google style.

I understand where you're coming from. I also feel that fear about the blockchain size is, to a certain extent, exaggerted. In the past year, the blockchain has expanded to ~25 GB, with ~14 GB coming in the previous year. In theory, if blocks were maxed out and the 1-block-every-10-min. target was hit, that would be ~52 GB. Let's further assume that blocks are actually mined every 5 min. due to continued explosive growth in the mining sector. That would be ~104 GB. On top of that, we'll assume Armory will always need a DB roughly the size of the blockchain. So, we're at ~50 GB right now, with potential for ~200 GB/yr growth if Bitcoin growth really did take us to the moon.

People here have said they're out when the blockchain hits 4 TB. Good thing that such an event probably won't happen for at least 20 years. Or, with Armory, at least 10 years. 6 TB hard drives are already out. By 2025, I'd imagine we'll be somewhere in the double digits, with 4 TB drives being the bare minimum available in shops. In other words, storing the basic chain won't require a farm. A fast connection and a lot of time (or somebody mailing you a drive) if you're starting from scratch, yes, but not a farm. Perhaps a farm would be needed if somebody was trying to track all the sidechains and such should sidechains ever get integrated into Bitcoin.

Yes, the blockchain is getting pretty large, and there is a need for lighter solutions. Yes, it's still possible to manage the blockchain from home and will continue to be so for a long time, although it will admittedly be less and less convenient over time. We'll be keeping an eye on the blockchain size and will react as is appropriate.

Western Digital readies first 10TB hard drive, ships new 8TB drive

http://arstechnica.com/information-technology/2014/09/western-digital-readies-first-10tb-hard-drive-ships-new-8tb-drive/

Dang, well there ya go...

[rax]

All right all right I'll buy in, folks. I didn't want to stir a scalability debate. And I don't stand a chance against an Armory developer anyway Personally I'll bring down my node when the clock ticks 100GB or so, and I know that this is gonna happen sooner rather than later, more so if I'm looking at a 30 or 40 year timeframe.

That's why I wanted to know if as today there is already a lightweight client that supports Armory offline transactions, that's all.

[njaard]

A little birdie told me that you can use https://counterwallet.io/ to create transactions and view your balance of Armory offline wallets, without ever revealing your private keys.

[rax]

A little birdie told me that you can use https://counterwallet.io/ to create transactions and view your balance of Armory offline wallets, without ever revealing your private keys.

Whoa! I belive this is exactly what I was looking for. Thank you, njaard!

[picobit]

A little birdie told me that you can use https://counterwallet.io/ to create transactions and view your balance of Armory offline wallets, without ever revealing your private keys.

Whoa! I belive this is exactly what I was looking for. Thank you, njaard!

But do read the small print: It only works with addresses that have already been used to SEND bitcoins at least once, because they need the public key, and they get that from the blockchain.  Of course if there is a market for it, they could easily make an interface reading it directly from the watch-only wallet, making their site equivalent to the online Armory computer (except for privacy issues, of course).

[Dabs]

If you have the private keys, you can get the public key, without spending.

[Raize]

For $20 you can own a 64gb flash drive:
http://www.amazon.com/PNY-Turbo-64GB-Flash-Drive/dp/B00FDUHDAC/

This is over double the current blockchain size.

If you're doing long-term storage, my guess is you've got more than 1 BTC that you are storing. $20 shouldn't be much to ask for.

While I understand the concerns, I find it likely that $20 will continue to be sufficient for at least holding the blockchain for 5-10 years. In that time, if it gets out of hand, it'll mean adoption has increased, not decreased, and the price along with it.

That said, the better arguments are ones already made, notably these two:

1.  Hope that Alan has finished the lightweight version of Armory not needing a full node.

2.  Export your private keys one by one from the offline wallet, import them in another client, and move the money like that.

But let's address these...

I know that this is gonna happen sooner rather than later, more so if I'm looking at a 30 or 40 year timeframe.

Plus, be aware that this is a 66x size increase in just over 3 years, and we are reportedly still in the early stages.

66 times in 3 years for 40 years:
40/3 = 13.33
66*13.33 = 880
25gb * 880 = 22 TB

40 years ago top of the line storage:
http://royal.pingdom.com/2012/09/04/hard-drive-market-1974/

5MB could be rented, but we should find a purchase price.

This link is better:
https://en.wikipedia.org/wiki/History_of_IBM_magnetic_disk_drives#IBM_3330

Purchase price:
http://www-03.ibm.com/ibm/history/exhibits/storage/storage_3330.html

$87,000 for 5MB (and this is a whole system)...
$87000 / 5 = $17400/MB

Today we can pay $20 for 64GB, so how much per MB is that?
$20 / 64000 = $.0003125/MB

So technology has increased to a difference per megabyte per dollar of:
17400 / .0003125 = 55.68 million

Assume that technology growth sucks hardcore and we only see a 1 million technological difference over the next 40 years:
64GB * 1 million = 64 PB

That means in 40 years we can expect a 64PB drive to cost $20. That is still more than capable of carrying a 22TB blockchain. Even accounting for inflation, you're still talking at least a 1PB drive for $20 in 2054.

I mean, we can fiddle with these numbers all we'd like, but we have absolutely no idea what's in store even 5-10 years from now, let alone 30-40. Even assuming we have technological growth at 1/55th the rate we have the last 40 years, we're still well within reason of storing the blockchain on a $20 drive. And if you asked any scientist in the 1970s (because, keep in mind, we couldn't ask the average person who had little idea what computers even were) if it was likely that in 40 years we'd have 64gb drives available for 55 million times less per megabyte than we did in the 1970s, they'd probably laugh and say it might be better, but not THAT much better.

I agree with scalability concerns, but mostly on the network bandwidth front, not so much on the back-end storage aspect.

[Dabs]

My guess is that your average laptop will have 10% hard drive (or equivalent) space available, good enough for the blockchain, in the year that you buy that laptop (or it was manufactured the year before.)

This means, that in 2020, which is just 6 years from now, you can probably still use a laptop you bought today, which might have anywhere from 300 GB to 2 TB, of which maybe the blockchain is anywhere between 100 GB to 200 GB.

By 2030, which is another 10 years, you should have bought a laptop (or whatever device is available then) within the last 1 to 3 years, and it will have maybe 10% to 20% of it's space available, which you can use to store the blockchain, whatever it's capacity is then.

Your 1 to 10 bitcoins that you bought in 2014, would still be 1 to 10 bitcoins in 2030, unless you spent it all.