From: Scott Morris (swm@emanon.com)
Date: Fri May 20 2005 - 21:53:59 GMT-3
It's actually a whole differernt decision-type process to give us the
reasoning behind all this!
Because of the way ISIS support for IPv6 was structured, the default
mechanisms employed a single SPF topology calculation. Which means that
your IPv4 routes and IPv6 routes needed to be fully deployed, or you may end
up black-holing yourself within ISIS.
The 63 is a maximum metric per link (default), the maximum metric for a
route is 10 bits (1023 max) by default.
Wide metrics increase these to a 24-bit per link and 32-bit max route metric
overall.
Wide metrics are not required for the "multitopology" implementataion of
ISIS with IPv6, but it is recommended. You'll follow the same defaults
(metric 10 per interface regardless of speed), so what you choose to do with
definitely depend on your network and what you want to accomplish! In the
lab, how many routers/routes/paths do you think you're going to have? I
would think that "narrow" metrics would be just fine in most instances
unless you're doing traffic-engineering (CCIE - SP), redistribution between
ISIS processes or Level 2 --> Level 1 or are required to do so.
Just my opinion.
Scott
PS. Check out:
http://www.cisco.com/en/US/products/sw/iosswrel/ps5187/products_configuratio
n_guide_chapter09186a00801d65f6.html
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
ccie2be
Sent: Friday, May 20, 2005 6:32 PM
To: 'Mark Lasarko'; ccielab@groupstudy.com
Subject: RE: isis metric-style transition
Hey Mark,
That's a different question altogether.
Off-hand, I donno for sure if ipv6 requires metric wide. But, I'm pretty
sure isis requires multitopology for ipv6.
Maybe in your NMC labs, you can find some ipv6 and isis examples. I've
heard some good things about the NMC labs when it comes to ipv6.
Sorry, I couldn't be more helpful.
Tim
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of Mark
Lasarko
Sent: Friday, May 20, 2005 5:49 PM
To: ccielab@groupstudy.com; ccie2be@nyc.rr.com
Subject: RE: isis metric-style transition
Greetings Tim,
Really?
I thought for sure IPv6 required this??
Something about the TLV's which advertise IPv6 info only being able to use
wide metrics...
I had not intended the question to be interpreted in the context of use for
IPv4 metrics > 63
Sorry if the question was not clear the way it was worded.
~M
>>> "ccie2be" <ccie2be@nyc.rr.com> 05/20/05 5:06 PM >>>
mark,
Unless you're told to use wide metrics either explicitly (unlikely) or
implicitly ( count the ways this could be done), I wouldn't start adding
misc commands to your config except for alias' or other "shopkeeping"
functions.
Unless you need to have isis metrics that go higher than 63, I think, don't
bother.
But, to each their own.
Tim
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of Mark
Lasarko
Sent: Friday, May 20, 2005 2:30 PM
To: ccielab@groupstudy.com
Subject: isis metric-style transition
Greetings GS,
Aside from the extra TLV's generated...
Is it a good idea to use this option in the lab?
Are there other reasons we should stick to "wide"?
TIA,
~M
BTW - I also noticed this recent memo relating to the subject:
M-ISIS: Multi Topology (MT) Routing in IS-IS
http://bgp.potaroo.net/ietf/ids/draft-ietf-isis-wg-multi-topology-10.txt
This archive was generated by hypermail 2.1.4 : Fri Jun 03 2005 - 10:11:59 GMT-3