Hi,
 
  Popularity of LISTSERV lists for classroom use is increasing in leaps and
bounds here on our campus.  Unfortunately, this popularity has helped cause
problems for our campus POP server.  We (as I'm sure others are) are
experiencing exponential growth in electronic mail service which burdens
the load on our mail server so severly sometimes, that mail can be delayed
for hours.  Ignoring the fact that the POP server is a problem and that most
of the time LISTSERV can keep up with the high demand placed on it (we had
a 300,000+ delivery day just the other day), we are attempting to come up
with ways to meet the demand for LISTSERV lists while at the same time
lessening the load on the POP server during peak hours.  The POP server
is a separate machine in this case, some high-end HP workstation, soon
to be a DEC alpha machine.
 
One way I have come up with for doing this is to make larger lists into digests
(by default) and have the digest delivered at off-peak hours.  The problem
with this approach, of course, is that nothing approaching real-time
interaction can be accomplished.  What would help somewhat is if I had
the ability to send a digest more than once a day.  Is this possible?
The "size" parameter on the Digest keyword won't work because it is not
possible to predict the size of the digest on any given day.  What would
be better is if I had the ability to send the digest twice a day at
predetermined times regardless of their size.  I'm sure the answer is
"no", but I was wondering if anyone else has encountered this and has
devised a solution?
 
Let me say we know the problem is with our implementation of POP and a
home-grown server, but that is not likely to change in the near future.
With that said does anyone have any ideas for me to increase our service
to the campus community while avoiding a system crashing load on the mail
server during peak hours.
 
Thanks in advance for any and all advice you can offer.
 
Oh yeah!  The LISTSERV I use for most lists today is:
 
> release
LISTSERV-TCP/IP for unix version 1.8b, managed by:
  [log in to unmask]
  [log in to unmask]
  [log in to unmask]
  [log in to unmask]
  [log in to unmask]
  [log in to unmask]
  [log in to unmask]
  [log in to unmask]
 
Master nodes file version:   95/11/06 16:14:26 (VERS9511)
NJE peers file version:      95/11/15 17:22:46
Internet peers file version: 95/11/15 17:26:24
Service file version:        95/11/08 13:51:12 (VERS9511)
Running under:               AIX 3.2
 
Summary of resource utilization
-------------------------------
 CPU time:        0.050 sec                Device I/O:        0
 Overhead CPU:    0.010 sec                Paging I/O:       13
 CPU model:         000109454600
 
 
--Greg Kroll
  Virginia Tech Computing Center