From: Troy Levin (troylevin@hotmail.com)
Date: Thu Mar 02 2006 - 21:54:10 GMT-3
Using a Fully Routed to the Access Layer Design has its advantages:
High availability
- Deterministic convergence times for any link or node failure in an
equal-cost path Layer 3 design of less than 200 msec
- No potential for Layer 2 Spanning Tree loops
Scalability and flexibility
- Dynamic traffic load balancing with optimal path selection
- Structured routing permits for use of modular design and ease of growth
Simplified management and troubleshooting
- Simplified routing design eases operational support
- Removal of the need to troubleshoot L2/L3 interactions in the core
However, you need to consider a few things before making that decision which
may or may not effect you.
1. Use of IP Addressing - as you push L3 to the access layer you use up
more address space. The use of /31's make this less of a concern
2. No ability to Span VLANs across switches. Each switch in the access
layer is a unique broadcast domain and VLAN assignment. If you need to SPAN
then this design is not possible.
My 2 cents....
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
ccie_2006@comcast.net
Sent: Thursday, March 02, 2006 4:50 PM
To: Leigh Harrison; FORUM
Subject: Re: To route or not to route.....
I meant to say I would prefer running IGPs to the closet instead of
Spanning-tree.
We just implemented the exact same solution where I work. We have a huge
voice network with Meeting place, video conferencing and IPTV. As matter of
fact we have absolutely no POTS lines. We have a total of 5000 users with a
possible 5000 more joining the campus. When we originally purposed it, a
bunch of old school CCIEs started laughing at us. But after testing and
comparing how long spanning tree takes to reconverge compared to routing
(OSPF or EIGRP). And not to mentioned, how easy it is to troubleshoot unlike
spanning-tree. Where it can takes almost a day to find where the loop was
started (not really a day exaggerating here no flames please :) but a long
time. Everything is running fine now I would prefer running IGPs to the
closet instead of Spanning-tree. I think Cisco is going to push this more
and more as time goes on.
From: postmaster@onnurimail.com
To: ccie_2006@comcast.net
Subject: [Err] Re: To route or not to route.....
Date: Thu, 2 Mar 2006 21:45:19 +0000
Content-Type: text/plain; charset="us-ascii"
X-Converted-To-Plain-Text: from Multipart/mixed by GroupStudy
X-Converted-To-Plain-Text: Alternative section used was text/plain
Transmit Report:
joykang@onnurimail.com?!0T 8^@O 9_<[@; 29x =C55G_Av88 =GFPGO?4=@4O4Y.
(=GFP @L@/ : 550 joykang@onnurimail.com... User unknown(127.0.0.1))
<B|0m> =GFP @L@/?! 4kGQ <38m
User unknown :8^@O@; <v=EGR ;g?k@Z0! A8@gGOAv >J@=
Socket connect fail:<v=E 8^@O <-9v?M ?,0a =GFP
DATA write fail :<v=E 8^@O <-9v7N 8^<<Av <[=E =GFP
DATA reponse fail :<v=E 8^@O <-9v7N:NEM 8^<<Av <v=E =GFP
Reporting-MTA: dns; onnurimail.com
Final-Recipient: rfc822;joykang@onnurimail.com
Diagnostic-Code: smtp; 550 error - joykang@onnurimail.com... User
unknown(127.0.0.1)
Action: failed
Status: 5.0.0
X-RECEIVED-IP: 207.44.210.9
Received: (from sympa@localhost) by lists.groupstudy.com (8.11.6/8.11.6)
id k22Li6B19343; Thu, 2 Mar 2006 16:44:06 -0500
Received: from groupstudy.com (www.groupstudy.com [209.51.144.7]) by
lists.groupstudy.com (8.11.6/8.11.6) with ESMTP id k22Lhwm19315 for
<ccielab@lists.groupstudy.com>; Thu, 2 Mar 2006 16:43:58 -0500
Received: from groupstudy.com (groupstudy.com [127.0.0.1]) by
groupstudy.com (8.12.11/8.12.10) with ESMTP id k22Li4wS001887
GroupStudy Mailer; Thu, 2 Mar 2006 16:44:04 -0500
Received: (from listserver@localhost) by groupstudy.com
(8.12.11/8.12.11/Submit) id k22Li4K3001885 for GroupStudy Mailer; Thu,
2 Mar 2006 16:44:04 -0500
Received: from sccrmhc14.comcast.net (sccrmhc14.comcast.net
[204.127.200.84]) by groupstudy.com (8.12.11/8.12.10) with ESMTP id
k22LhxGZ001854 GroupStudy Mailer; Thu, 2 Mar 2006 16:43:59 -0500
Received: from smailcenter61.comcast.net ([204.127.205.161]) by
comcast.net (sccrmhc14) with SMTP id <2006030221434601400e3141e>; Thu,
2 Mar 2006 21:43:51 +0000
Received: from [71.126.157.45] by smailcenter61.comcast.net; Thu, 02 Mar
2006 21:43:46 +0000
From: ccie_2006@comcast.net
To: Leigh Harrison <ccileigh@gmail.com>, FORUM <ccielab@groupstudy.com>
Subject: Re: To route or not to route.....
Date: Thu, 02 Mar 2006 21:43:46 +0000
Message-Id:
<030220062143.6684.440767120000BEE600001A1C2205884484C0CFCFCDA10A070C0C@comc
ast.net>
X-Mailer: AT&T Message Center Version 1 (Aug 4 2005)
X-Authenticated-Sender: Y2NpZV8yMDA2QGNvbWNhc3QubmV0
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
X-Converted-To-Plain-Text: from multipart/alternative by GroupStudy
X-Converted-To-Plain-Text: Alternative section used was text/plain
X-ASK-Info: Our key was found in the mail (2006/03/02 16:44:04)
Sender: nobody@groupstudy.com
Reply-To: ccie_2006@comcast.net
X-Loop: ccielab@groupstudy.com
X-Sequence: 50198
Errors-to: ccielab-owner@groupstudy.com
Precedence: bulk
X-no-archive: yes
List-Id: <ccielab.groupstudy.com>
List-Help: <mailto:sympa@groupstudy.com?subject=help>
List-Subscribe: <mailto:sympa@groupstudy.com?subject=subscribe%20ccielab>
List-Unsubscribe:
<mailto:sympa@groupstudy.com?subject=unsubscribe%20ccielab>
List-Post: <mailto:ccielab@groupstudy.com>
List-Owner: <mailto:ccielab-request@groupstudy.com>
We just implemented the exact same solution where I work. We have a huge
voice network with Meeting place, video conferencing and IPTV. As matter of
fact we have absolutely no POTS lines. We have a total of 5000 users with a
possible 5000 more joining the campus. When we originally purposed it, a
bunch of old school CCIEs started laughing at us. But after testing and
comparing how long spanning tree takes to reconverge compared to routing
(OSPF or EIGRP). And not to mentioned, how easy it is to troubleshoot unlike
spanning-tree. Where it can takes almost a day to find where the loop was
started (not really a day exaggerating here no flames please :) but a long
time. We decided to go with it.
Everything is running fine, now I would prefer running spanning to the
closet instead of Spanning-tree. I think Cisco is going to push this more
and more as time goes on.
-------------- Original message --------------
From: Leigh Harrison <ccileigh@gmail.com>
> All,
>
> I'm currently working on a design for a customer. Straight forward
> design with Access and a Core. 3750's in the access layer and a 6513
> in the core (yes there is only 1, but the customer already has it, it
> has dual sup cards and dual power supplies...) the 3750's are in
> stacks and there is dual gig links back to the core.
>
> I was at a Cisco seminar recently where Cisco said that the best
> practice is to route, rather than use spanning tree and switch,
> essentially turn off spanning tree. I'm quite happy to run either way,
> but I do have a question:-
>
> We are running VoIP on the network and there is call recording
> software going in. This needs to have the ports of the gatekeepers
> span'd to it so that it can do the recording. If I'm routing my
> network, what are the options for accomplishing this if my gatekeepers
> are not connected to the same switch?
>
> I presume that someone out there has run into a similar issue, so any
> insight would be greatly appreciated.
>
> Best Regards
> LH
> #15331
>
> ______________________________________________________________________
> _ Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Sat Apr 01 2006 - 10:07:37 GMT-3