The MPLS-OPS Archive

Cell Relay Retreat>MPLS-OPS Archive>month:2001-Feb> msg00107



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

Re: ISIS or OSPF?

  • From: Eric Osborne <eosborne@cisco.com>
  • Date: Mon, 19 Feb 2001 22:10:00 -0500
  • Cc: Alan Claughan <alan@is.co.za>, Eric Osborne <eosborne@cisco.com>, mpls-ops@mplsrc.com
  • Resent-Date: Mon, 19 Feb 2001 23:12:26 -0500
  • To: svrolyk@ieee.org
  • User-Agent: Mutt/1.2i
  • X-GPG-Fingerprint: 6412 0836 E440 B3EA 980C 4951 611E 1819 2E71 8562

On Mon, Feb 19, 2001 at 06:03:06PM -0800, Scot Vrolyk wrote:
> Please see below.......... Excerpted from Cisco MPLS Software Configuration
> Guide Rev 9.3.10................
>  IS-IS has somewhat better performance than OSPF in full mesh conditions
> because IS-IS has
> more sophisticated flooding capabilities (these capabilities, specifically
> the ability to pace flooding and
> block flooding on some interfaces, are also becoming available on OSPF).
>

Sure, and up until recently it could have been argued that OSPF scaled
better than IS-IS because it had better support for inter-{area|leve}
routing.  Or that IS-IS is better because it uses CLNS as its
transport protocol.  Or that OSPF is better because it uses IP as its
transport protocol.  Or that IS-IS is an ISO standard and OSPF is an
IETF standard, and clearly one of those groups is just morally
superior to the other.  Or maybe that OSPF engineers tend to have
better teeth than IS-IS folks, who have better hair.

So what?  

For a good comparison of is-is vs. ospf, see Dave Katz's slides from
the Albuquerque NANOG.  Or you can just skip to the punchline, which
is that both protocols are extremely similar.  I don't want to start
an is-is vs. ospf thread, but I *do* want to correct the occasional
misperception that cisco unilaterally recommends protocol A over
protocol B (or protocol I over protocol O, I suppose).  It's like
pepsi vs. coke - pick your favorite, savor the knowledge that the one
you happened to have picked is clearly the better one, and get on with
life.  :)

For the record, and I'm sure I'll be saying this again in a month or
two, cisco supports both is-is and ospf, both for regular IP routing
and for traffic engineering.  IS-IS was implemented first because
that's what our first TE customers were using.  There are some TE
features in IS-IS that aren't yet in OSPF, and vice versa.  But
believe it or not, we *do* occasionally listen to customers, and if
anyone out there is personally offended that we didn't do feature X
for protocol {O|I} yet, please let me know.  

Now I get off my soapbox and go back to my cave....:)




eric

 
> Scot R. Vrolyk
> Principal Network Architect
> 
> NetStream Communications, Inc.
> 1101 Creekside Ridge Dr. #150
> Roseville, CA	95678
> O: (916)677-4074
> F: (800)854-2764
> M: (214)906-7793
> 
> -----Original Message-----
> From: Eric Osborne [mailto:eosborne@cisco.com]
> Sent: Monday, February 19, 2001 1:14 PM
> To: Alan Claughan
> Cc: mpls-ops@mplsrc.com
> Subject: Re: ISIS or OSPF?
> 
> On Mon, Feb 19, 2001 at 08:56:48PM +0200, Alan Claughan wrote:
> > Cisco put the Traffic Engineering extensions into IS-IS first, so in the
> > early days of tag-switching IS-IS had to be used if TE was to be
> > configured. But these days both IS-IS and OSPF are supported. I would
> > advise sticking to the one you know and understand best!
> 
> Agreed.  From a cisco perspective, both IS-IS and OSPF are fully
> supported.  Use the one you're familiar with, or whichever one fits
> your needs.
> 
> 
> 
> 
> eric
> 
> -------
> The MPLS-OPS Mailing List
> Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
> Archive: http://www.mplsrc.com/mpls-ops_archive.shtml

-------
The MPLS-OPS Mailing List
Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
Archive: http://www.mplsrc.com/mpls-ops_archive.shtml