Timeline for Using PBKDF2 in combination with AES-KW defeats usage of BCrypt password hashes?
Current License: CC BY-SA 4.0
8 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 14 at 18:32 | history | became hot network question | |||
| Nov 14 at 12:01 | comment | added | Gh0stFish | But Bcrypt is not very resistant to FPGAs (see this article from 2019) or to ASICs, which is what you're comparing to when you look the hash rates on at mining rigs. | |
| Nov 14 at 11:43 | comment | added | oleg_zh | @Gh0stFish: yes, it may be unfair, however the issue remains the same, namely SHA-x is paralellizable on GPU, while bcrypt is not. I think we will go with Argon2, as suggested below | |
| Nov 14 at 11:40 | vote | accept | oleg_zh | ||
| Nov 14 at 11:01 | answer | added | Ja1024 | timeline score: 5 | |
| Nov 14 at 10:58 | comment | added | Gh0stFish | And it also seems rather unfair to compare the cracking rate of unspecified ASICs (which probably don't even support PBKDF2) to a rather arbitrary "0.1s per hash" for bcrypt. | |
| Nov 14 at 10:54 | comment | added | Gh0stFish | OWASP doesn't generally recommend using PBKDF2 - the main recommendation in the Password Storage Cheat Sheet recommends using Argon2id. PBKDF2 is only recommended "If FIPS-140 compliance is required". And even in that case, they're recommending a much higher work factor than you use in your example. | |
| Nov 14 at 10:32 | history | asked | oleg_zh | CC BY-SA 4.0 |