Scheduler is losing track of successful deliveries

[Andy Poling <andy@jhunix.hcf.jhu.edu>]

I'm seeing this problem consistently.  After a transport agent has
successfully delivered a message (shown by a "r+" in the ctl file) scheduler
sometimes seems to miss it and keeps re-trying it.  This is a pain because
all the transport agent does is ignore it, saying nothing to scheduler which
promptly tries it again.  I'm losing a significant number of cycles to this
looping mis-behavior.  It doesn't even seem to help to send additional "ok"
statuses to the scheduler when re-trying - it is already too confused.

If a message is destined for multiple addresses, it gets even more screwy:
scheduler will become internally convinced that it has already delivered to
addresses to which it has not, and that it has not delivered to address
which it has.

If I kill scheduler and start a new one, the new one takes care of
everything, but all expiry information is lost.  This means that the mail
queue just grows continually, bloated with undeliverable mail that never
gets expired.  And this bloated queue seems to cause scheduler to lose its
sanity even more quickly.

The affected messages show in a mailq listing as having nothing for their
status.

I've got a steadily ramping mail queue here.  Any suggestions?  Help!

-Andy

Andy Poling                              Internet: andy@jhunix.hcf.jhu.edu
UNIX Systems Programmer                  Bitnet: ANDY@JHUNIX
Homewood Academic Computing              Voice: (410)516-8096    
Johns Hopkins University                 UUCP: uunet!mimsy!aplcen!jhunix!andy