[c-nsp] L2 or L3 Core using 3550-12G

Church, Chuck cchurch at netcogov.com
Wed Feb 16 10:04:38 EST 2005


Kevin,

	If I'm reading this right, your 'current L2 cores with L3
interfaces at the
distribution switches' refers to the old devices, right?  If so, I think
with some careful planning you can get this to work.  The 3550FX and 12G
will support uplink fast/backbone fast, in addition to Rapid STP.  I've
played with RSTP, and seen convergence times around 2 seconds (about the
time it takes to determine an interface is down).  That's pretty close
to OSPF (I don't believe the 3550s support sub-second OSPF hellos yet).
Keep in mind that the 12G has a limit of 16 (I believe) VLAN interfaces.
Not sure if this is an issue.  You might want to check out the 3750-12S
in place of the 3550-12G switches.  You'll have 2 additional fiber ports
to play with, eventual IPv6 support, a dedicated 32gig backplane between
the 2 cores, and possibly more VLAN interface capability (not sure about
that one though).
	But to answer your question, if you only want to route at the
core, you're going to need to rely on a L2 protocol for everything south
of the core.  STP is easy to load balance with, on a per-VLAN basis.
But the 3550FX will route as well.  So that's a possibility as well.
Whichever you do, read all the release notes for the switches you're
using.  There are many caveats regarding QOS placement on routed vs.
switchport ints, etc.


Chuck Church
Lead Design Engineer
CCIE #8776, MCNE, MCSE
Netco Government Services - Design & Implementation Team
1210 N. Parker Rd.
Greenville, SC 29609
Home office: 864-335-9473
Cell: 703-819-3495
cchurch at netcogov.com
PGP key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x4371A48D 


-----Original Message-----
From: cisco-nsp-bounces at puck.nether.net
[mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of
Kevin.X.White at corusgroup.com
Sent: Wednesday, February 16, 2005 3:40 AM
To: cisco-nsp at puck.nether.net
Subject: [c-nsp] L2 or L3 Core using 3550-12G


We are in the process of converting an existing FDDI site to a 3 layer
Cisco model using 2 off 3550-12Gs at the core and 2 off 3550FX in each
(of
10) switchblocks as distribution switches.
Our traffic levels are low for High priority traffic (<15Mbsec on FDDI)
except for low priority backup periods with around 150 users on each
geographicaly separated switchblock.

The question is can the current L2 cores with L3 interfaces at the
distribution switches be succesfully converted to do all the routing at
the
core allowing us to load balance across the cores and not have to wait
for
L2 convergance if a link fails or a switch goes down (yes I know about
backbone fast)

Will the 12G support 10 routed interfaces and continue to behave itself?

Kevin


_______________________________________________
cisco-nsp mailing list  cisco-nsp at puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/



More information about the cisco-nsp mailing list