routing protocols sco unix
From: "Brian K. White" <email@example.com> Subject: Re: Problem with routes disappearing. Date: 10 Aug 2005 16:38:27 -0400 Message-ID: <003601c59deb$6a782020$6b00000a@venti> References: <firstname.lastname@example.org> ----- Original Message ----- From: "Alex" <email@example.com> Newsgroups: comp.unix.sco.misc Sent: Wednesday, August 10, 2005 3:17 PM Subject: Problem with routes disappearing. >I am actually having two completely different sets of servers at > different locations have a problem that is very similar but not quite > identical. All of the servers in question are IBM SC5200's with Sco > Openserver 5.0.6 What is happening is that I have to add routes for > the IPSEC VPN tunnels we have to these clients. The routes seem to just > disappear. On one of the sets of servers the route is actually > completely gone from the routing tables and has to be readded. I tried > to use a cron job to readd the routes periodically but that doesnt seem > to work either. On the other set of servers the route is still there > but stops working and cant be seen with netstat -rn. I have to delete > it and readd it to get it to start working again. Here's an example of > the second problem. The route in question is route add -host > 192.168.1.14 10.1.39.1 > > This first example is taken after the route stopped working, as you can > see it is not visible in this netstat. > > server1 # netstat -rn > Routing tables > Destination Gateway Flags Refs Use Interface > default 192.168.1.254 UGS 0 0 net0 > 10.1.39 10.1.39.2 UC 1 0 net1 > 10.1.39.2 127.0.0.1 UGHS 0 0 lo0 > 127.0.0.1 127.0.0.1 UH 7 124 lo0 > 192.168.1 192.168.1.201 UC 1 0 net0 > 192.168.1.201 127.0.0.1 UGHS 0 0 lo0 > > server1 # route add -host 192.168.1.14 10.1.39.1 > add host 192.168.1.14: gateway 10.1.39.1: File exists > > server1 # route delete -host 192.168.1.14 10.1.39.1 > delete host 192.168.1.14: gateway 10.1.39.1 > > server1 # route add -host 192.168.1.14 10.1.39.1 > add host 192.168.1.14: gateway 10.1.39.1 > > server1 # netstat -rn > Routing tables > Destination Gateway Flags Refs Use Interface > default 192.168.1.254 UGS 0 0 net0 > 10.1.39 10.1.39.2 UC 1 0 net1 > 10.1.39.2 127.0.0.1 UGHS 0 0 lo0 > 127.0.0.1 127.0.0.1 UH 7 124 lo0 > 192.168.1 192.168.1.201 UC 1 0 net0 > 192.168.1.14 10.1.39.1 UGHS 0 0 net1 > 192.168.1.201 127.0.0.1 UGHS 0 0 lo0 > > Any ideas as to whats happening to this route??? cd /etc mv routed routed.disabled mv gated.conf gated.conf.disabled # ignore error if not exist shutdown -g0 -i6 -y # this = reboot so warn first! Explanation: Probably a routing daemon is doing it's job, which is to dynamically re-write the routing table based on instructions it receives from other routing daemons on other machines on your network. SCO Open Server includes at least 3 different routing daemons out of the box, namely routed and gated and irdd. 5.0.6 and below ship with routed enabled by default. 5.0.7 ships with none enabled by default. /etc/tcp starts and stops various network services and performs other network start/stop actions like ifconfig and "route add" and starting either routed or gated at boot and any time you manually run /etc/tcp start or stop. /etc/tcp (on 5.0.6 and lower) will start gated if /etc/gated.conf exists, which it doesn't unless you created one. If gated.conf doesn't exist and routed does exist (which is does unless you move it) then it will start routed. So on a stock 5.0.6 (or lower) box, routed is enabled by default. There is no "nicer" way I know of to disable routed other than simply renaming /etc/routed. /etc/tcp does look first to see if routed exists it doesn't just try to run it so I guess that was always the official way to disable it. irdd is installed and commands to start it exist in /etc/tcp but are completely commented out so nothing ever starts that automatically. This is on 5.0.6 and lower. On 5.0.7 (possibly also 5.0.6 after some update) you can edit ROUTERDAEMON* variables in /etc/default/tcp to turn any routing daemons on or off and they are all off by default. OR, ... Maybe you need the routing daemon because your network is large and needs to be automatic/dynamic. In which case, find and fix the machine that's broadcasting the bad RIP packets that the SCO box is picking up. Brian K. White -- firstname.lastname@example.org -- http://www.aljex.com/bkw/ +++++[>+++[>+++++>+++++++<<-]<-]>>+.>.+++++.+++++++.-.[>+<---]>++. filePro BBx Linux SCO FreeBSD #callahans Satriani Filk!
Got something to add? Send me email.
(OLDER) <- More Stuff -> (NEWER) (NEWEST)
Printer Friendly Version
Increase ad revenue 50-250% with Ezoic