Great Circle Associates Firewalls
(July 1997)
 

Indexed By Date: [Previous] [Next] Indexed By Thread: [Previous] [Next]

Subject: Re: RIP vs. OSPF
From: pomeranz @ netcom . com (Hal Pomeranz)
Date: Wed, 2 Jul 1997 17:10:58 PDT
To: johns @ oxygen . house . gov (John Schnizlein), firewalls @ GreatCircle . COM, kgibbs @ best . com
In-reply-to: johns @ oxygen . house . gov (John Schnizlein) "Re: RIP vs. OSPF" (Jul 2, 6:13pm)

RIP has terrible convergence problems on very meshy networks-- routes
may never stabilize if you lose a critical device.  RIP doesn't support
variable-length subnets either, though I gather this is coming in RIP-2.

On the other hand, OSPF is a _pig_ on core routers in large networks.
There are also interoperability problems still lurking between Cisco
and Bay in my experience.  Is anybody still running mixed-vendor
network fabrics, though?

On Jul 2,  6:13pm, John Schnizlein wrote:
} The best solution for security purposes is to (hard) configure the default
} router into your host computers. Unfortunately, this is not the most robust
} configuration against network failure because it locks the host into a single
} path when multiple (valid) routers may be available.

See also Cisco's HSRP (Hot Standby Routing Protocol) which enables two
routers to back each other up by sharing an IP address (which you then
configure as default on your hosts).  Also IRDP, or whatever they're
calling these days.

Hal Pomeranz, Principal        Deer Run Associates        hal @
 deer-run .
 com
      Network Connectivity and Security, Systems Management, Training

Indexed By Date Previous: Labs that will do firewall perfomance testing.
From: "Dustin Goodwin"<dustin_goodwin @ smb . com>
Next: Re: Microsoft plans to offer a firewall
From: Visionprof @ aol . com
Indexed By Thread Previous: Re: RIP vs. OSPF
From: johns @ oxygen . house . gov (John Schnizlein)
Next: CFP: 1998 SNDSS (updated; last reminder!)
From: bishop @ cs . ucdavis . edu (Matt Bishop)

Google
 
Search Internet Search www.greatcircle.com