Recently I asked whether anyone could help with a routing problem we have on
several SPARCstation 2's. I received many replies, but unfortunately none of
them have helped me to solve the problem so far. The responses fell into a
number of categories :
I should get gated - 7 respondents (one person even sent me their config file -
much appreciated, but we haven't tried it yet)
Check our broadcast address - 5 respondents
Make sure ifconfig is run (correctly) BEFORE routed - 4 respondents
Make sure you're not running 'routed -q' (quiet option) - 4 respondents
Try running routed with the -s option - 1 respondent
Check that our CISCO gateway listens to, and forwards, RIP packets - 4 respondents
Check /sys/netinet/in_proto.c - 2 respondents
In response to these responses -
Yes, ifconfig has been correctly run with broadcast addresses set to all 1's,
on both interfaces, before routed is started. We have also tried killing and
restarting routed, and it has no effect. Even with the '-s' option.
All our campus CISCOs talk RIP, not IGRP. Apparently this can be a problem if
the Suns happily talk RIP but not IGRP, but the CISCOs only talk IGRP and not
RIP.
One of our machines has IPFORWARDING in /sys/netinet/in_proto.c set to 1, which
apparently is incorrect - we thought it might help though. Two of the other
machines still have it set to 0, which appears to be the correct value, and
they don't work either.
We have a number of machines all on the same subnet, which is then connected
through a CISCO router to the rest of the campus. The problem is evident on the
local subnet, so it isn't just a matter of the CISCOs not forwarding RIP.
In fact, when we run etherfind and look for broadcast rip packets, we see
ABSOLUTELY NOTHING from these hosts. The same command run on other subnets on
campus, looking for routing packets from other Suns with two interfaces running
SunOS 4.1.1 (not 4.1.2 as on our machines) shows that the other Suns are
broadcasting routing information.
I appreciate all your responses. If anyone has any idea why our Suns are not
broadcasting routes, given that they have correctly configured broadcast
addresses, I'd be interested to hear from you.
I have enclosed a list of respondents below. If anyone would like the
individual messages, just ask, and I'll send them to you.
"Adam W. Feigin" <feigin@iis.ethz.ch>
geertj@ica.philips.nl (Geert Jan de Groot)
alastair@eucad.co.uk
gdmr@dcs.edinburgh.ac.uk
birger@vest.sdata.no (Birger A. Wathne)
shollen@valhalla.cs.wright.edu (Sheila Hollenbaugh)
"Brian H. Powell" <brian@natinst.com>
montjoy@thor.ece.uc.EDU (Robert Montjoy)
stpeters@dawn.crd.ge.com (Dick St.Peters)
era@niwot.scd.ucar.EDU (Ed Arnold)
Mike Raffety <miker@sbcoc.com>
"John Paul O'Brien" <john@solar.nova.edu>
Tom Conroy <trc@NSD.3Com.COM>
Rich Brown <rabr@tc-savoy.gwl.com>
algold@server01.lncc.br (Alexandre Leib Grojsgold)
stern@sunne.East.Sun.COM (Hal Stern - NE Area Systems Engineer)
pbh@cfsmo.honeywell.com (Paul B. Henninger)
Mr T Crummey (DIJ) <tom@sees.bangor.ac.uk>
Arne Henrik Juul <arnej@lise.unit.no>
dan@engr.latech.edu (Daniel Len Schales)
agw@math.canterbury.ac.nz (Allen Witt)
hyder@niwot.scd.ucar.EDU (Paul Hyder)
Michael S. Maiten <msm@Energetic.COM>
Stefan Mochnacki <stefan@centaur.astro.utoronto.ca>
---------
Ivan Dean Here : x3261
Faculties Computer Unit There : (06) 249 3261
Australian National University Everywhere : +61 6 249 3261
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:47 CDT