krowitz@RICHTER.MIT.EDU (David Krowitz) (04/19/89)
I've nstalled SR10.1 on my DN560 using the /etc/hosts files
(as opposed to name servers). The DN560 is attached to our
ringnet which, in turn, is connected to our campus etheret
by a DN660 running SR9.7 and Domain TCP 3.1 (the AEGIS TCP/IP).
The ringnet is net 18.138 (using a subnet mask of 255.255.0.0) and
the campus ethernet segment for our building is 18.83. We use
the /etc/rc.local file distributed with sr10.1 pretty much as
it cam (starts up /etc/tcpd, runs /etc/iconfig with the command
/etc/ifconfig net0 `/etc/hostname` netmask defaultmask
runs /etc/routed with the -f and -h options [as it is set up in
the distributed file], and then starts inetd). The name
'defaultmask' is defined in /etc/hosts as 255.255.0.0, yet
when the machine is booted, we can not connect to any machine
(ie. ping, telet or ftp) not on the local ringnet *unless*
the rip_server (the AEGIS TCP/IP routing daemon) was running
on the gateway at the time the sr10 node was booted. The /etc/hosts
file lists both an address on the Apollo ringnet and the campus
ethernet for the gateway node, and the subnet mask should have
been defined correctly by the 'ifconfig' command, yet we
can not access the campus ethernet unless the routing daemon
was running on the sr9.7 machine whch is the gateway.
I've tried several different variations on the /etc/hosts file
(removing comments, changing the order of entries, etc), hardwired
the subnet mask into the 'ifconfig' command, etc. and the only
thing which seems to make a difference is the presense of the
routing daemon on the gateway node. I understand that the
'-f' option of /etc/routed clears the tcp daemon's routing
table pror to reloading it with fresh info, but shouldn't it
fall back on the /etc/hosts file when the info can't be
found in the internal tables? When I add a new host to /etc/hosts
I can access it immediately whether or not the machine is
on my local ringnet or on a remote network. Should I not be
running /etc/outed on a non-gateway node? The tcp manual for
sr10 and the defaults in the /etc/rc.local file suggest that
I should be running it.
-- David Krowitz
krowitz@richter.mit.edu (18.83.0.109)
krowitz%richter@eddie.mit.edu
krowitz%richter@athena.mit.edu
krowitz%richter.mit.edu@mitvma.bitnet
(in order of decreasing preference)weiner@novavax.UUCP (Bob Weiner) (04/21/89)
David Krowitz writes:
I've nstalled SR10.1 on my DN560 using the /etc/hosts files
(as opposed to name servers). The DN560 is attached to our
ringnet which, in turn, is connected to our campus etheret
by a DN660 running SR9.7 and Domain TCP 3.1 (the AEGIS TCP/IP).
Yet when the machine is booted, we can not connect to any machine
(ie. ping, telet or ftp) not on the local ringnet *unless* the
rip_server (the AEGIS TCP/IP routing daemon) was running on the
gateway at the time the sr10 node was booted. The /etc/hosts file
lists both an address on the Apollo ringnet and the campus ethernet
for the gateway node, and the subnet mask should have been defined
correctly by the 'ifconfig' command, yet we can not access the
campus ethernet unless the routing daemon was running on the sr9.7
machine which is the gateway.
We've had this same problem except that our Ethernet gateway node
along with all of our other nodes is running SR10.1 (BSD).
I finally found an Apollo tech rep in Texas who knew about this and
said that yes, none of the routing on the non-gateway nodes will work
after the gateway node is taken down and then rebooted. He said we
would need a script that checked every so often if the gateway had
gone down and then executed /etc/routed -f -h on all of the other
nodes again.
This is definitely a pain that Apollo should fix. We want dynamic
routing to be fully dynamic. Any tech help from the sharp Apollo R&D
minds that read this list would be greatly appreciated.
--
Bob Weiner, Motorola, Inc., USENET: ...!gatech!uflorida!novavax!weiner
(407) 738-2087