Home > IT Answers > Networking > #com v's Cisco
Help

Question

  Asked: Nov 20 2006   4:06 PM GMT
  Asked by: bouncybrit

#com v's Cisco


Networking, Hardware, Routers, Switches, Hubs, Cabling, 3Com, Avaya, Cisco, Performance management

Hi,
I am in the process of scaling up a Datacenter to handle what over the next 3 to 5 years is expected to be a doubling of the current workload. I am pushing my current stackable switches to the limit at peak times and the performance impact is very visible even to end users.

I am managing a network with approximately 1000 nodes.
multiple SQL servers for medical records and imaging (lots of data) SQL servers for financial data, admissions billing all the standard stuff. we are also doing a lot of Voice over IP and Voice over WiFi. we have 25 different VLANS, and obviously we have to deal with regulatory requirements for logging, security, availability etc.
I have been working on updating the Core Switching and for the various buildings and remote sites I have had no problems with the midrange 3Com series of switches the 40x0's. Of course they have just been end of lifed and I dont think that they will be really capable of handling everything in the core that we are going to throw at them over the next couple of years, so....

I have narrowed my choices down to a Cisco 6509 or a 3Com 8800. both switches claim to have the modules and support for the various protocols I am going to be throwing at them. (gigabit ethernet, 10 Gigabit ethernet down the road, OSPF, QoS, VLAN Access controls etc) I am not looking outside of 3com or Cisco as I really dont want to go down the road of learning a brand new interface and command language. Does anyone have experience positive or negative with either switch that they are willing to share with me.

thanks
Jason

Subscribe to Alerts! Get questions and answers delivered to your Inbox.


E-mail me updates on this question



   SUBSCRIBE

hidden modal window
Help

Answer Wiki (Improve, edit or add to this answer)

IMPROVE THIS ANSWER
View History
 RATE THIS ANSWER
0
Click to Vote:
  •   0
  •  0
Last Answered: Nov 20 2006  6:31 PM GMT  by astronomer




Jason:
I don't have experience with the models you mentioned but my advice would be to stick with 3com if they have worked well for you.
There was a time when cisco consistently delivered the goods while their competitors stumbled. That time has passed. I believe cisco has reached the point where their marketing is significantly more innovative than their engineering. We use HP switches and cisco routers here. The lifetime warranty support from HP has, in general, been significantly better than cisco.
As long as the switch fabric is fast enough and the features are adequate for the forseeable future, I don't see a good reason for picking one brand over the other. Stick with what you are comfortable with.
rt
  • AddThis Social Bookmark Button

Browse more Questions and Answers on Networking, DataCenter and VoIP.

Looking for relevant Networking Whitepapers? Visit the SearchNetworking.com Research Library.


RSS - Discussions Help

Discuss This Answer


You must be logged-in to discuss a question. Log-in/Register

mortree  |   Nov 23 2006  2:09AM GMT

Yes look at fabric/backplane speeds of core switches first. Many switches let you add modules way beyond what the fabric can handle.

The 3Com looks good because total 720Gbpps of dual fabric doesn’t reach max packets until the packet size drops to an average of about 200 bytes per packet. So it is all usable speed. Too bad it is split over dual fabric…though I guess that does give you redundancy.

The Cisco needs that 720 Supervisor Engine to match it…maybe. I didn’t see packet handling for Cisco or how many fabrics that speed was split among. Let me guess this is more expensive?

I am wondering if you really want to continue to take growing VoIP onto your Core though. I have been hearing more and more that the volume of growing VoIP traffic is too hard to predict and control. I keep hearing that companies are looking at splitting VoIP off from data in network cores around datacenters. I guess the philosophy is that “yes we can save lots of money on wiring from the distribution level downwards. But really it is worth a few more coins in the core to split the wiring or use optical mixers on common fiber to keep core data flowing smoothly.”

 

mortree  |   Nov 23 2006  2:09AM GMT

Yes look at fabric/backplane speeds of core switches first. Many switches let you add modules way beyond what the fabric can handle.

The 3Com looks good because total 720Gbpps of dual fabric doesn’t reach max packets until the packet size drops to an average of about 200 bytes per packet. So it is all usable speed. Too bad it is split over dual fabric…though I guess that does give you redundancy.

The Cisco needs that 720 Supervisor Engine to match it…maybe. I didn’t see packet handling for Cisco or how many fabrics that speed was split among. Let me guess this is more expensive?

I am wondering if you really want to continue to take growing VoIP onto your Core though. I have been hearing more and more that the volume of growing VoIP traffic is too hard to predict and control. I keep hearing that companies are looking at splitting VoIP off from data in network cores around datacenters. I guess the philosophy is that “yes we can save lots of money on wiring from the distribution level downwards. But really it is worth a few more coins in the core to split the wiring or use optical mixers on common fiber to keep core data flowing smoothly.”