[c-nsp] Top of Rack Switching

[Pete Templin]

Scott Altman wrote:

> For quantities we are talking about 60 pair of 4948's per IDF, so net
> is 120 4948's, 2 6509's.  I chose to use the SFP version of the 4948's
> vs. the 10Gb based on the availability of upstream 10Gb ports in the
> 6509 (4 10Gb, vs 48SFP) and my bandwidth requirements don't warrant
> greater than 4Gb out of each server rack.

I think you mean (or should consider referencing) 2Gb out of each server 
rack.  Without MPLS TE or something similar, you'd have a hard time 
getting _functional_ 3Gbps out of the switch pair in the event of a link 
failure.  You also need to plan for a switch failure bumping you down to 
2Gbps available throughput.

I know of an enterprise person at a company near me who's got 6509s with 
numerous Cat4500s exchanging routes via EIGRP.  I think he said he's got 
100 Cat4500s, so I don't think your numbers are of any great concern.

See Cisco's recent Campus Design fundamentals for key pointers on this 
sort of design.  Particularly, STP root placement and HSRP/VRRP/GLBP 
alignment are key.

pt