[c-nsp] Top of Rack Switching

[Wojtek Zlobicki]

>
> The ToRS model would consist of a pair of 4948's in every server rack
> providing 5-8 VLANs to the servers in the rack (Productionx2, Backup,
> Management, etc).  The 4948's would have a 2x1Gb EthC between them and
> they would each have a 2x1Gb EthC up to a 6509 for aggregation;
> basically creating a U-shaped topology from a spanning-tree
> perspective.  Due to other design requirements, my L3 boundary can't
> come any closer.


Do you really have the port density in each rack to justify this?  We're
talking 96 ports in one rack, that's a lot of 1U boxes with dual Ethernet.
My biggest concert other than port density and wasted ports, would be
security.  In most organizations I have worked for, there is a separation of
people that have access to sever boxen and network gear.  Do you have any
issues with your staff and customers having physical access to your switch?


> What I would like to hear about are:
> - Issues with managing the complexity of the environment (managing 122
> switches in an IDF is not common)


Managing 40 or 120 is not significantly different as long as you standardize
your processes from the beginning.  Remember to keep a spare chassis around
if at all possible.  I am personally partial to the 6500's.  Depending on
traffic patterns among racks, I would rather have data traversing a switch
fabric than inter switch trunks.  With the amount of gear you're purchasing,
I assume that you're not expecting to pay list.  Second, with this type of
capital expenditure, I'd be surprised if Cisco and or your vendor would not
be willing to lend a hand with recommendations and design.  The 6500 will
also allow you to extend other services through the platforms modules.
Being able to install an FWSM of CSM module can mean being able to provide
more services out of one chassis.