[nos-bbs] JNOS SMTP session hangs

[Michael Fox - N6MEF]

Hi Maiko (and others),

Recently, we have experienced two separate SMTP session hangs on two
separate JNOS machines.  The symptom is that messages on BBS-1 which are
destined for users on BBS-2 are stuck in queue on BBS-1 and do not go out -
ever - until we discover the problem and manually intervene (bad!).  The BBS
seems to be working fine for local message delivery.

Logging in to BBS-1 and pinging BBS-2 we see that the serial port is working
just fine and that RF and IP connectivity between the two is just fine.
Looking at the SMTP queue on BBS-1 with "smtp list", we find that one of the
messages has an "L" next to it, indicating a lock file.  Sure enough, the
lock file exists (in .../spool/mqueue or .../spool/mail - I forget which).
Stopping JNOS, deleting the lock file, and then restarting JNOS causes the
mail to start flowing again.

We have the following smtp parameters set in autoexec.nos:
smtp dtimeout 24
smtp t4 300
smtp timer 600
smtp tdisc 300
smtp usemx off
smtp maxclients 3

So, a few questions:
1)  What would cause an SMTP session to either hang or terminate improperly,
leaving the lock file there for all eternity?  

2)  Is there some type of watchdog timer in JNOS that can be set such that
it would be able to tell that the session failed and then clean up by
deleting stale lock files, etc. and running the queue again?

3)  Has anyone else seen this problem and, if so, what did you do to address
it?

Thanks,
Michael
N6MEF