Re: Old song - timeouts

[mea@nic.funet.fi]

...
>>> BTW. I noticed that changing the timeout values in smtpserver.h doesn't
>>> help a single bit. That precisely (and a quick look at the smtpserver
>>> code) gave me the idea that there must be something wrong with the alarm()
>>> usage. Don't you think that it would be better to use some form of virtual
>>> timing (e.g. by virtualizing timers with setitimer?)
>> 
>> Nope, nothing in the timers has changed to cause this type of problems.
>
> But how would you explain that changing the timeout values in smtpserver.h
> doesn't affect the smtpserver behavior at all?

Perhaps you could watch system a bit, and attach 'strace' to active
smtpservers -- or perhaps even to use
	strace -fp smtp-daemon-pid

The 'strace' on Linux has several other options, 'man' tells more.
(Like ways to store traces into separate files per pid.)

/Matti Aarnio