Procurve Redundancy

[Phil_UK]

Hi

My current setup for redundancy is as follows:

I have 2 x 2810-24G and 3 x 2650-48 Procurve switches with spanning tree enabled.

The NICs on the ProLiant servers are teamed and one port from each card connected to one of the 2810s so each server is connected to both 2810s.
Each of the 1G uplink ports on the 2650s are connected to one of the 2810 so each 2650 is connected to both 2810s. PCs and printers are connected to the 2650s.
The 2810s are connected to each over with two trunked ports.

I have no issues other than the fact I'm running short of availble ports on the 2810s

Following the takeover of another company I have aquired 9 x 3400cl switches and would like to utilise them.

What would be the best configuration for redundancy? Should I replicate the above configuration or are there better features to use on these switches.

Any advise would be welcome.

Thanks

[Pieter 't Hart]

basically there seems no need to change the topology, if you only need to add more ports at the "access-layer".

but i think the 3400 has more capabilities than the 2810.
if so it could be an idea to replace your 2810's as "core-layer" by 3400's.

teaming across switches will be fault-tolerant only (not trunks) in both situations. so there's no improvement.

[Pieter 't Hart]

apart from that you may use "switch meshing"
to have all redundant links active instead of some blocked by spanning-tree.

see attachment for single image
or http://ftp.hp.com/pub/networking/software/6400-5300-4200-3400-AdvTrafficMgmt-Oct2006-59906051-Chap07.pdf