redundancy using WAN ethernet connections

CCDECCDE9 · ‎07-27-2009

Hi

I have 100Mbps ethernet link between two sites terminating at CAT4006(Site A) and CAT3750(Site B).The interfaces that connect the sites are in the same sunet 10.10.10.0/24 i.e VLAN 20(on Cat4006) and VLAN1 (on 3750) and static route is used to reach other subnets or vlans in these switches.My question is what needs to be done from config point of view if I get another same type of link(ethernet)with less bandwidth (20Mbps) between these sites.

Thanks

Giuseppe Larosa · ‎07-27-2009

Hello,

you can:

a) use floating static routes to have this second link used only in case of failure of primary faster link

b) you can use PBR to divert some specific traffic flows over this secondary link during normal operations conditions.

OR you can use EIGRP with its capability of load balancing over unequal cost paths.

Hope to help

Giuseppe

CCDECCDE9 · ‎07-27-2009

Do I need floating static route ?I thought switches use the second ethernet link automatically when primary fails using the existing as is routes through RSTP built into the switches..Please comment

Giuseppe Larosa · ‎07-27-2009

Hello,

I see if these are L2 links yes RSTP will manage L2 redundancy and you will be careful in assigning an higher STP cost to slower link on both sides.

Clearly my first post comments are for L3 links I had missed this point.

Hope to help

Giuseppe

CCDECCDE9 · ‎07-27-2009

do I need to configure anything on the switches ?or just put the new link in the same vlan as primary

Giuseppe Larosa · ‎07-27-2009

Hello CCDE,

>> do I need to configure anything on the switches ?or just put the new link in the same vlan as primary

if I were you I would increase the STP cost on secondary link

int gx/y

spanning-tree cost 500

on both sides = switches

this to be sure that primary link is used when available.

Hope to help

Giuseppe