First off, thanks to:
clark rudder <rudder@infonet1.infonet.com>
gdonl@gv.ssi1.com (Don Lewis)
Stefan Mochnacki <stefan@centaur.astro.utoronto.ca>
heasley@merl.com
ulla@dmi.min.dk (Ulla Fischer)
Claus Assmann <ca@informatik.uni-kiel.de>
"C.R. Ritson" <C.R.Ritson@newcastle.ac.uk>
glenn@uniq.com.au (Glenn Satchell - Uniq Professional Services)
Eckhard.Rueggeberg@ts.go.dlr.de (Eckhard Rueggeberg)
J.H.N.Chin@reading.ac.uk
gdonl@gv.ssi1.com (Don Lewis)
Neil Sobrado <sobrad22@zodiac.cs.newpaltz.edu>
pamela@Legato.COM (Pamela Pledger)
Dan Stromberg - OAC-DCS <strombrg@bingy.acs.uci.edu>
Neil Rickert <rickert@cs.niu.edu>
Jacques.Beigbeder@ens.fr (Jacques Beigbeder)
eww@hobbes.crc.com (Eric Wampner ORL)
"Fuat C. Baran" <fuat@tintin.cc.columbia.edu>
Bill Morrow <morrow@cns.ucalgary.ca>
The general consensus is that -q[1-5]m is too small and simply wastes
resources. Most people suggested -q30m or -q1h.
What most people did not mention (and I did not know about) is that the
reason that mail delivery had been slow at -q5m was that these systems
were set up to do mail processing at queue time *only*. I don't know who
had set it up this way, but the machines here had the line `Odqueue' in
their /etc/sendmail.cf's, which had to be changed to `Odbackground' so
that they would initially attempt to deliver mail at send-time, and only
use the queue if the initial attempt failed.
Thanks to Don `Truck' Lewis (gdonl@gv.ssi1.com) for this insight.
Please note that `Odqueue' and `Odbackground' are equivalent to `Odq' and
`Odb' respectively.
Thanks again to all who responded.
LT
,-----------------------------------------------------.
| Yale Economics Dep't | Lenny Turetsky |
| System Administrator | lturetsk@econ.yale.edu |
|-------------------------+---------------------------|
| My employers paid for some of my time and energy. |
| My opinions were never for sale. |
`-----------------------------------------------------'
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:09:13 CDT