07-07-2008 12:35 PM - edited 03-06-2019 12:02 AM
We are in the process of redesigning the core.
There are several ideas.
The one I like the best is:
2 6509's
720's in each
and bridge the 2 together as one switch, not sure of the exact terminology.
Basically want it to look like a switch stack. Admin from one connection to both devices.
The question is, if we lose one sup will layer 2 still work on the chassis that lost the blade?
What controls the layer 2 function on the switches?
Is the MSFC configured on each side or is it configured on one side and pushed to the other side, kind of like the 3750 stack?
What I am looking for is a redundant core, available all the time. The other idea is 2 seperate set-up's running HSRP, which I am fine with, but do not feel it is the best set-up for us.
Also, at this point money is not really an issue, I know sounds almost too good to be true!
07-07-2008 12:44 PM
You're looking for VSS. Here's a link that should answer all your questions.
Hope that helps.
07-07-2008 12:47 PM
yes, that is exactly what I was looking for!
Thanks!
Could not remember what it was called.
07-07-2008 12:55 PM
Have you done this?
One quick question before I put myself to sleep with this tonight?
Does this pass all layer 2 info if the sup is down on one of the switches or being upgraded?
Basically, one function we are looking for is either dual sups or the VSS. I happen to like the VSS option but not 100% up on it.
Does any of this make sense?
07-07-2008 01:04 PM
I have NOT done this, we went with two switches and layer 3 links. If one Sup is down the other takes over for all layer 2 and layer 3 services.
07-07-2008 01:21 PM
At this moment only one sup per chassis is supported. If it goes down only half part of virtual switch will be working.
Support for two SUP per chassis will be in IOS 12.2.33 SXI
07-07-2008 12:59 PM
what suprevisors do you have?
you can ran VSS only with DFC3C or DFC3CXL
so you must have SUP720-10G
07-07-2008 01:04 PM
we can get whatever we need.
The idea is what is the best option for what we want to accomplish.
If we need new Sup's then we will get them.
This core must be solid with High availablity and not really be able to take any down time unless we are hit by a bomb. Our data center is actually a storm bunker that can take just about anything, except a bomb dropped from a plane...no lie.
07-07-2008 01:16 PM
You might want to check out some of the SRNDs too.
http://www.cisco.com/en/US/netsol/ns656/networking_solutions_design_guidances_list.html
Like I stated before, we went with two switches and multiple L3 links. For us, this provides high availability with better scalability, and higher throughput capabilities on our network.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide