Greg Hudson <ghud...@mit.edu> writes: > /dev/random starvation explains the clock skew errors
This is a troubleshooting point that we should emphasize more: A Kerberos message might be within the clock skew tolerance when sent, but for whatever reason, the receiver might delay processing it until it is no longer within the clock skew tolerance. This can result in clock skew errors that don't appear to correspond to any observable clock skew between systems. ________________________________________________ Kerberos mailing list Kerberos@mit.edu https://mailman.mit.edu/mailman/listinfo/kerberos
