> On Sat, 24 Jan 1998, Jake Messinger wrote:
>
> > On Sat, 24 Jan 1998, Jason Hatch wrote:
> >
> > > In this example, all ethernet devices are on the same /24 network, shown
> > > here as "A". All ComOS are version 3.7.2 and all portmasters (finally)
> > > have 4 megs of ram, with exception of the IRX-114 which only has 1 meg of
> > > ram (it has either 2 or 300k free on it).
> >
> > Bump it to 4 if you are considering ospf or bgp.
>
> Must I really go to 4 if I'm not using BGP? I would think I would be able
> to squeeze OSPF in and maybe even SNMP:
>
> System memory 1048576 bytes - 731700 used, 316876 available
> 1152:0 640:0 80:5 176:0 32:0 128:2 16:96 256:1 48:1
> System nbufs 1100 - 52 used, 1048 available
>
> Besides, when it comes time to do BGP, I think I'll get a Cisco.
> Livingston's are great for quick and easy routers for downstream
> customers, but to think that all this time has passed (2 years I've had
> it) and I can't even assign multiple IP's to the ethernet interface. Just
> think, if I could do that, then I wouldn't be worrying about doing OSPF on
> my tiny POP, I'd just config the interfaces on all my PM's to have an IP
> on all my subnets and let proxy arp go to work for me.
>
> > > Now, when we dial into portmaster 1, with a static IP of B.253, something
> > > wierd happens. A traceroute from the Solaris box shows it going to the IRX
> > > (it's default route), over to portmaster 3, then to portmaster 1, then to
> > > its destination. Looking at the route tables of the IRX, we still see our
> > > RIP route of B.0 pointing to portmaster 3. Going into portmaster 2, we see
> > > the same route. But looking at portmaster 3, we see a RIP route pointing
> > > B.0 to portmaster 1. It looks as if portmaster 1 selectivly sent a RIP
> > > update to portmaster 3, but not the router and the other portmasters. This
> > > explains how portmaster 3 knew to send it to portmaster 1.
> >
> > Maybe cuz the ip address used belonged to the OTHER portmaster first?
> >
> > > So what happens when someone dials into portmaster 2 with a B.252 static
> > > IP with the B.1-30 assigned address pool on portmaster-3, and a user with
> > > B.253 on portmaster 1. Will it break then? :-)
> >
> > Not unless you have the static route defined in the portmaster. I assume
> > you are using radius and this will work as long as you are using a routing
> > protocol so that the static IP gets routed to properly... BUT, you have
> > PM # 3 in a separate subnet? Can a user with a static IP dial into 1, 2 or
> > 3? You prolly shoudnt be doing that because you could be pulling an IP
> > from one of the OTHER subnets.
>
> Actually, it did break. I am using RADIUS and RIP. Consider this,
> portmaster 3 has an IP address pool in subnet B. On boot, it sends a RIP
> broadcast to all the Portmasters and the IRX that it has B.0 (when it
> really just has B.1-30). A customer dials into portmaster 1 with a Static
> address of B.253. This works, as Portmaster 1 sends a RIP announcement to
> Portmaster 3 (but not the others) that it has B.0 now, so portmaster 3
> will send everything that is not dialed in (even idle IPs from its pool)
> to Portmaster 1. Now, I dial into Portmaster 2 with a Static IP of B.251
> and that user can do NOTHING but ping Portmaster 2 _until_ about 30
> seconds (time for next RIP announcement?) after the guy with B.253 on
> Portmaster 1 _logs_off_.
>
> Now, I know that RIP is bad to begin with, but what I don't understand is
> why did portmaster 1 send a RIP announcement to portmaster 3 only saying
> that it had B.0, which would cause IPs in B to be sent to Portmaster 3,
> then 1 (the router and other PM's think B is on 3), but why didn't
> portmaster 2 do the same, causing it to try 3, then 1 then 2?
>
> -Jason
>
Jason - I'm not trying to be rude or to piss you off, but rather to make a
recommendation: Go immediately to OSPF. Upgrade your RAM today - the
price is right; don't put it off. What you have (described above) is a
duct tape and hot glue solution; really - its a credit to your ingenuity
that it works at all!
When you bite the OSPF bullet (as I did) it'll take you an evening to get
it setup and _everything_ will be so much simpler to manage! I know OSPF
can be intimidating (I was intimidated), but its really not that difficult
to setup if you read the "this is how you do it" stuff and ignore all the
routing theory (for now).
Also Solaris 2.6 finally added support for VLSM - you might be very
surprised to learn that "/usr/sbin/in.routed -q" will work just fine if
you enable OSPF on your PortMasters and Cisco routers. No more trying to
get gated to run reliably on your Solaris 2.5.1 boxes.
Also I agree with your statement "BGP on Cisco". Livingston/Lucent is
really at the network edge - they do this very well. They don't try to
compete at the backbone routing level - that's a game only the big boys
can play. And if you look at a Cisco 3640 (or 3620) you'll see that this
is at the "sweet spot" in Ciscos price/performance curve right now. The
3640 especially - it may well be the last router you'll buy! Well....
maybe not!!
Just my $0.02.
Al Hopper Logical Approach Inc, Plano, TX. al@logical-approach.com
(972)-379-2133 or (972)-849-5765. Fax 972-379-2134
If you aren't confused, then you simply don't know what's going on!
-
To unsubscribe, email 'majordomo@livingston.com' with
'unsubscribe portmaster-users' in the body of the message.