Re:Hash Functions all the way down
by qwertyuiop924 on Wednesday August 03, 2016 @05:41PM
(#12221771)
Wait, don't you want minimum acceptable throughput on security-related hashing, so that it becomes less likely that somebody can just brute-force it?
[ Parent
| on HN ]
Not everything hashish ends up being salted passwords. Hashing is great for just hashtables and partioning.
Re:Hash Functions all the way down
by qwertyuiop924 on Wednesday August 03, 2016 @06:49PM
(#12222114)
Ah, that makes sense. Thanks.
Re:Hash Functions all the way down
by revelation on Wednesday August 03, 2016 @06:01PM
(#12221904)
This is evaluating hash functions for use in something like a hash table where the concern is first and foremost for speed at goodish distribution, not collision resistance.
The security (DoS rather) issue with hash tables is more that 1) attackers often control the input, and H(a) == H(a) obviously and 2) the hash function output is reduced to a few bits to decide which bucket to put something into. In those cases making your hash function slower just makes you more susceptible.
Yeah I forgot to mention that even if most of the time you don't care about "security concerns" with regular hashtables, there are cases where a weakness in your hashtable can be used to bring some of your services down. That has happened!
However, many of the times that is not a concern, but as always, case by case basis.
Re:Hash Functions all the way down
by JoshTriplett on Wednesday August 03, 2016 @06:12PM
(#12221972)
For hashing, you want the maximum possible throughput. If you want to hash a password, you can always iterate the hash function to add difficulty, or better yet, use a mechanism specifically designed for hashing passwords, such as scrypt, bcrypt, or PKBDF2.