I use OpenSSL with aes-256-cbc to encrypt a 30 GB tar file containing a recent backup of my hard drive. Now suppose somebody wants to run a brute-force attack on this file, trying out all possible password combinations. Does OpenSSL have to scan through the entire 30 GB of data in order to say whether the password matches or whether it doesn't match or can OpenSSL determine really quickly whether the specified password is the correct one?

In other words: Does it take longer to run a brute-force attack using OpenSSL on a 30 GB file than on a 100 KB file?


it depends, but - generally - no. AES is a block cipher, so you're breaknig a thing by 256 bit blocks in your case. And 100KB is more than enough to guess a type of "what's inside", by MIME, for example, after that a task is a way more simplier

  • Thanks, but I don't understand the second half of your answer. How can you guess what's inside by MIME without knowing the password? I thought it's only possible to know what's inside if you actually know the password?! – Andreas May 20 '16 at 15:57
  • 1
    It's a nit, but I think AES256 has a 128 bit block size en.m.wikipedia.org/wiki/… – Neil Smithline May 20 '16 at 16:09
  • 2
    @Andreas when decrypting you guess the password and you know you have the right one because the data looks good and makes sense. For example, the data looks like a disk image. Alexey is suggesting that ~100k is enough to know if it makes sense or not – Neil Smithline May 20 '16 at 16:11
  • tar has strongly-nonrandom metadata; 1kB is more than enough to verify. If you actually have tar compressed with gzip or bzip2 or lzma, which are very common for large backups, probably even less. – dave_thompson_085 May 20 '16 at 23:21
  • @NeilSmithline of course, AES256 block size is 128, I was talking about a portion of the beginning of the file sufficient for MIME detection. – Alexey Vesnin May 21 '16 at 13:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.