Wol wrote: > On 23/09/2023 14:35, Dale wrote: >> Another question. Are people trying to work on better encryption >> given current encryption can be cracked? I read some things changed >> after Snowden. I'm just not sure what and if more changes are needed >> even today. > >> If you wanted the most secure and hard to crack encryption, what >> would you use? How does one tell cryptsetup to use it? I have >> several encryption options here but no idea what is the best or even >> just good. > > If you want encryption that can't be cracked, go for RSA. It's > uncrackable. > > Now you might be wondering why I say that, given that is a simple, > well-known attack, but it's true. You can trick me into encoding as > much plain text as you like, where you can intercept the cipher text, > and you will not be able to crack the cipher itself. What you need to > do is get hold of ONE of my key-pairs. The public one of course is > usually freely available, and if you get hold of the private one it's > game over. > > You can then mathematically solve "the puzzle of the keys" from my > public pair and recover the private key. This is why RSA keys keep > getting bigger - it takes more and more brute force to solve. > > I don't know enough about ECC - do you crack it or solve it? > > Both these ciphers however have a massive weakness - make a mistake > setting them up and the solution becomes easy. RSA relies on > multiplying two huge primes together. Dunno what ECC relies on. If one > of your RSA primes is not, in fact, prime then factoring the huge > product becomes easy, and recovering all the keys built from it is > simple. > > ECC specifies various parameters, and the official standard ECC > parameters were discovered to contain a flaw. Was that an intentional > back door? It's thought it was an accident. > > But I think cryptographers have abandoned crackable ciphers now - if > it's crackable then it's easily crackable. And all other ciphers > simply rely on the asymmetric effort taken to create a key or solve a > key. > > Cheers, > Wol > >
When I run cryptsetup to encrypt my drives, I have no idea what it is using. I assumed the defaults would be the most secure. This is the luksDump info, some may be changed or snipped, not sure if it is something I should make public. ;-) root@fireball / # cryptsetup luksDump /dev/sdo1 LUKS header information Version: 2 Epoch: 3 Metadata area: 16384 [bytes] Keyslots area: 16744448 [bytes] UUID: 967257e5-ccc8-48ab-8f46-c6b05dc3bf37 Label: (no label) Subsystem: (no subsystem) Flags: (no flags) Data segments: 0: crypt offset: 16777216 [bytes] length: (whole device) cipher: aes-xts-plain64 sector: 4096 [bytes] <<<< SNIP >>>> Digests: 0: pbkdf2 Hash: sha256 Iterations: 83062 Salt: 20 d5 f5 3b 51 43 31 29 8a b0 31 dc ad 56 0c 15 50 18 aa f8 df a0 4e 9e 8e e1 b2 bb f1 04 67 01 Digest: 96 18 90 9e 89 7a 16 71 72 d0 97 ec 84 e1 b5 38 fc cb ea 97 93 29 19 4c 83 a6 fb 4e e9 ba 79 7b root@fireball / # I'm not to clear on this but it looks like it is using 'aes-xts-plain64' to me. If so, is that good enough? Is there better? While I'm mostly worried about someone maybe stealing my rig, I also don't want someone with some skills getting in there either. Some crooks may know someone. ;-) Dale :-) :-)
