Topic: Is Dropbox a safe place keep my TrueCrypted BTC wallet backup? (Read 5332 times)

I will use both (Dropbox and Google Drive)

I think it should be safe, i personally use google drive as i use my google account for a lot of stuff.

But I I will never find out if someone have a copy of my wallet.dat.

Because I know it's there now?
No, you have some strength with that system. Even if someone got the wallet you have a long time to move the coins out of it.

you can put your truecrypted wallet to any filehoster you want. mh, why not torrent it?

it's safe.

My wallet.dat is encrypted via the Bitcoin-QT and then put inside an an encrypted rar archive. Both passwords are 190 bits according to KeePass (30 characters long mixture).

Why is that not secure if I put it on Dropbox?

Do we know what a truecrypt file's metadata contains?

I would assume it doesn't have any, for the obvious security reasons. Yes, you could immediately know the file doesn't match its extension, but if that's the case you could throw in a bunch of red herrings as well! Take a random jpg, encrypt it with 40 different passwords and put them all in the same place. They'd have a 1 in 41 chance of choosing the right file to start with.

For the JPG's you could use embarrassing party photos, which provides motivation for the encryption, but no value to the hacker--so they would assume all the files are the same and there is no reason to keep cracking them.

The most important thing is probably keeping a low profile AND hiding your stuff. No one's going to look very hard if they don't suspect something is there.

It depends on how thorough the hacker is.

This a Bad Idea. The file's internal metadata (which is what anyone searching for a needle in a stack of needles will be looking at) will be a dead giveaway. Renaming a file with a different extension will not conceal it at all, and may actually make it even more conspicuous with tools that automatically flag files whose filename extension doesn't match the internal metadata. Security through obscurity doesn't work at all.

Well, it's just bruteforcing passphrases. For every character you add to the passphrase length, you change the difficulty of bruteforcing quite a lot. Let me demonstrate


OK, lets imagine. [source: http://calc.opensecurityresearch.com/ ]

Bruteforcing SHA256 at 138 terahash/s when key length is.. (oh, I just checked: according to blockchain.info the hashrate is today 174 terahash/s)
8: less than a second (lowercase alphanumeric) | 1 second (mixed alphanumeric)
9: less than a second | 1 minute 39 seconds
10: 27 seconds | 1 hour 43 minutes
11: 16 minutes | 4 days
12: 10 hours | 275 days
13: 15 days | 47 years
20: 3 billion years | 164 trillion years
256 lowercase alphanumeric characters (SHA256 hash): 

So keep your passphrases long. I suppose to maximize key strength one could hash the passphrase before using any standard truecrypt algorithm. It's rather likely that an attacker would attempt to bruteforce against a known hashing algorithm (or a sequence of them), so that's one more hindrance.

Thanks for the link, I was looking if hashcat had been used for anything besides password hashes, but hadn't found this.

PBKDF2-HMAC-SHA512 / AES: 95 kHash/s
PBKDF2-HMAC-RipeMD160 boot-mode / AES: 451 kHash/s

OK, RipeMD took 40 seconds, SHA512 is 5 times slower, so it will take 10 minutes. Let's guess it was a simple password, it would take much longer with a better SHA512 password. But then remember it was just two HD6990. Let's see what we can do with a Cluster of 25 Titans: 25 x 163.0 kH/s = 4075 kHash/s. Or with a cluster of 25 HD7970: 25 x 233.0 kH/s = 5825 kHash/s.

You also have to take into consideration that you don't have to find the exact passphrase for pasword cracking, you just have to find a combination of symbols that gives you the exact same hash. d7a8fbb307d7809469ca9abcb0082e4f8d5651e46d3cdb762d02d0bf37c9e592 is the SHA256 hash of "The quick brown fox jumps over the lazy dog". So "a!kL07gS1" might give you the same hash (it is just an example, it doesn't have the same hash) and you could decrypt the file with that as well. So what can you do for more security is either using a hashfunktion that generates a longer hash (07e547d9586f6a73f73fbac0435ed76951218fb7d0c8d788a309d785436bbb642e93a252a954f23 912547d1e8a3b5ed6e1bfd7097821233fa0538f3db854fee6 is the SHA512 hash of "The quick brown fox jumps over the lazy dog") or using an slow algorythm like blowfish. But with more hashing power the time needed to solve even those puzzles gets shorter. Just imagine what you could do with the 138 Thash/s SHA256 hashpower that the bitcoin network currently has LOL. Sure not much people here would have the money to build a cluster of 200 Titan, HD7970 or Xeon Phi just to crack some passwords, but it's no problem for the secret services (Do I hear Prism or GCHQ? ) or some criminal organisations to do this.

Then it's not so shitty after all, is it? I'm a noob in encryption, anyone got any counter arguments?

Yeah.. it's gonna take 72 years to crack a 8 letter alphanumeric password.

Guess you can hack everything if you want to and have good equipment sooner or later.

I guess it was a short password to prove it is working, as it was done in 40 seconds with two 6990.
This is only the beginning, the programmer has speed up the whirlpool hashing now by 58% and yet to work on cascaded modes.
But the times where Truecrypt was secure will be surely over soon, when you now can hack every windows password with a 25 GPU cluster (10x HD 7970, 4x HD 5970 (dual GPU), 3x HD 6990 (dual GPU), 1x HD 5870):
http://arstechnica.com/security/2012/12/25-gpu-cluster-cracks-every-standard-windows-password-in-6-hours/

...in a safe, sunken in the deepest ocean.

The safest will be to keep your backup wallet in an offline USB drive.

Seems like a shitty encryption if it's that easy.

Truecrypt? LOL ... it can be cracked in no time with some Radeon GPUs:
http://hashcat.net/forum/thread-2301.html