On Fri, Jun 13, 2008 at 13:26:43 -0400,
"Horst H. von Brand" <vonbrand(a)inf.utfsm.cl> wrote:
[AFAIR, the random pool is only 512 bytes in size, so it can't contain more
"entropy" than that]
I used to be able to recompile the kernel with a larger pool size, but I
haven't done that in a long time. If the requests are bursty, but enough
entropy is being generated over time to supply the amount needed without
delay, then upping the poolsize might be worth doing to reduce delays.