Thanks for all the responses, this list is great. For [1], there were general suggestions that those were indeed the correct files to edit, and Jem included /etc/dumpadm.conf [optionally]. However, David Melnik and Alan Pae also suggested using sysunconfig as the best way to go, as it should take care of all of the above automatically. I'll also look into that, as I suspect it might do other things like change the /var/crash/nodename directory, etc. For [2], Jem Richards and Jeff Horowitz really nailed it (and [1] too. :) ) by guessing my problem, and several others were on the right track given the lack of information in my original email. Yes, the "remote" snoop was on the same network as dmfe1, and thus the traffic was going out dmfe1 like it was supposed to do. My apologies, I'm still learning this, and assumed the default gateway would always be used. :) I realize now that it will go out dmfe1 if it's the same network So the network traffic is probably fine and I'll verify this tommorow. Since we intend to use this extra interface for jumpstart management anyway, i'll probably just unplumb it post jump & boot, which should eliminate any confusion and extra host authentication (e.g. tcp wrappers) for the extra interface. Thanks also to: Glenn Harrison, Zaigui Wang, Rob De Langhe, mlh, and Darren Dunham! ivo Ian Ivo Veach wrote: >Greetings - > >We've got an unusual situation where we're testing jumping a netra x1 from >it's second interface [the jumpstart network], and then configuring it as >if it were the host based from it's primary interface. I'm wondering the >best way to go about the name change? > >The netra is jumped as "foojump," [dmfe1] but we eventually want the >machine to be called "foo." [dmfe0] The jumpstart process has correctly >set up the hosts, hostname*, netmasks, and defaultrouter to use the >hostname "foo" on reboot - I've confirmed these are accurate. Two >questions: > >[1] What is the best way to change the system name? We can use hostname, >which doesn't seem to help much. We can change /etc/nodename and >/etc/net/tic*/hosts so that they all refer to foo. This does seem to >change the name after reboot, but... > >[2] Even after 1, the system is in a state of confusion. It thinks it is >called "foo" at a glance. But snooping traffic from another machine >indicates traffic is coming from "foojump" [dmfe1] and NOT "foo." [dmfe0] >What's worse, is that routing is very weird. Given: > >foo# ifconfig -a >lo0: flags=1000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4> mtu 8232 index 1 > inet 127.0.0.1 netmask ff000000 >dmfe0: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 2 > inet 10.10.1.9 netmask ffffff00 broadcast 10.10.1.255 > ether 0:3:ba:6:1a:e4 >dmfe1: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 3 > inet 10.10.2.20 netmask ffffff00 broadcast 10.10.2.255 > ether 0:3:ba:6:1a:e4 >foo# netstat -nr > >Routing Table: IPv4 > Destination Gateway Flags Ref Use Interface >-------------------- -------------------- ----- ----- ------ --------- >10.10.1.0 10.10.1.9 U 1 2 dmfe0 >10.10.2.0 10.10.2.20 U 1 3 dmfe1 >224.0.0.0 10.10.1.9 U 1 0 dmfe0 >default 10.10.1.254 UG 1 11 >127.0.0.1 127.0.0.1 UH 3 12 lo0 > >foo# more /etc/hosts |grep foo >10.10.2.20 foojump oldloghost >10.10.1.9 foo > >Would you expect a remote snoop to see traffic from 10.10.2? Because that >is what we are seeing still - traffic from foojump, associated with >10.10.2, even though the default gateway is 10.10.1.254 (and >/etc/notrouter exists). Any suggestions? > > >cheers and thanks, > >Ian Veach, imail@nevada.edu >_______________________________________________ >sunmanagers mailing list >sunmanagers@sunmanagers.org >http://www.sunmanagers.org/mailman/listinfo/sunmanagers _______________________________________________ sunmanagers mailing list sunmanagers@sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagersReceived on Thu Mar 13 02:58:33 2003
This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:43:05 EST