Autore: Neal Becker
Data:
To: exim-users
Oggetto: race condition?
I noticed this happened twice:
--- mainlog ---
1996-04-09 16:42:44 pid 242: SIGHUP received: re-exec daemon
---------------
--- paniclog ---
1996-04-09 16:42:44 socket bind() to port 25 failed: Address already in use
---------------
This was on hpux9.05. Is it possible that a delay is needed? Did the
old process forget to close the socket before re-exec?
Note: I have sighup'd exim on other machines without this problem, so
either it is a function of the machine or the load or is just
intermittent.