CISCO 3750s vs. A CISCO CORE SWITCH?


Hello All:

I have recently been informed that Cisco's core switches brag massive amounts of throughput on the internal switch matrix, but that the slots where the blades plug into are extremely limited (or bottlenecked) and that they cannot deliver nearly as much speed as the claim throughout the entire switch.

So I have been advised to someone who deals in hardware every single day, to purchase Cisco's new 3750 series switches as they will support VLANs and everything that we need to do at a much cheaper price point, and they also deliver MORE throughput than the standard Cisco Core Switch with a supervisor module etc, etc..

I am currently evaluating the possibility of installing TWO of these 3750s as opposed to ONE CORE switch, but I want more details on whether these statements are accurate, as well as what exact functionality I would be loosing when comparing to let's say a Cisco 4503 with the redundant power supplies, SUPERVISOR module installed, etc, etc?

Which switch is faster, which switch is a better purchase?  Is there any validity to these claims?  Currently I am looking at installing two 3750s and using the mini-GBICs to supply fiber between floors to some 5, 2950 (48pt and 24pt)  switches.   Is there anyway to make these 3750 switches act in a FAILOVER mode and make redundant connections in case one fails while no one is there, that it will automatically switch over to the other switch?    

Please supply me with all of your comments and thoughts, as I am trying to look into this decision from every single angle?

Thanks In Advance..
andgroupAsked:
Who is Participating?
 
lrmooreConnect With a Mentor Commented:
3750's make excellent core switches in smaller networks. They have a special cable that connects them together creating a 32GB backplane. You can get redundant power supply later, add up to 9 switches in the stack and the stack acts as one switch. Uplink both switches to the floors (dual-runs) and you have your automatic failover. Whichever one is the "stack master" will automatically failover to the other if the master fails.
I like these switches. They have an awsome web-based gui that the catOS based switches can't match. For 10/100/1000 integration, they are great.
0
 
MarkDozierCommented:
These are just general ideals based on how I reserch design questions like this
The first thing you need to do is determine what your needs/desires are.
Write out a list of requiremtn and questions
I would call Cisco sales rep's and ask them these questions. These folk are pretty good.
I would also read everything I find on the net, trade jorunals, etc..
0
 
andgroupAuthor Commented:
My initial thought process was to designate one switch as the primary and hookup the fiber runs to it, but I was also going to connect all the 7-8 servers to it with gigabit copper, and then if it went down, I'd physically go up there and pull all the cables out of the primary switch and move them down to the secondary switch.

If I want this solution to work with automatic failover, do I have to install a second NIC card in each server (which is already a pain because two of them are already clustered, and have a private heartbeat NIC in them already, so in two of the servers I guess I'd have to have 3 NICs installed in that server)?

What are you suggestions for making this failover TRULY automatic?

Thanks
John Woods

0
 
andgroupAuthor Commented:
Also keep in mind, that if I have multiple NIC cards in a server I may run into DNS/WINS issues
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.