When using OpenSSL it by defaults uses /dev/urandom for random seeding.
 
When using OpenSSL just for signature verification & hash functions I don't need the random.
 
But:
1. When using OpenSSL for signing data, and for SSL sessions (both client side and server side), should I use the /dev/random instead?
2. If I should use /dev/random; Is it OK to initialize a file one time from /dev/random and then when I start my application I would load it from that file, and when I finish I will write my random status back to that file? Is this a truly strong good random?
3. How much bytes should I load from the random file?
4. Do I need to load it again once in a while, or is it enough to load it once, even if my system runs for a long time before restarting (weeks)?
 
Thanks,
 
Joe

Reply via email to