Hi All,
Wondering if you can help with something obvious, as I might not be seeing the wood for all the trees!
Running CM5.2.1, with resilient S8730's and G650 gateways at the core, with lots of G350/430/450's that have LSP capability at remote sites.
We have 10Mbps MPLS connections between all sites and have just installed low bandwidth backup DSL links for failover by HSRP to each site. I don't want VOIP calls running over the backup DSL as this is only for mission critical traffic, instead I'd like calls to route out via IGAR instead.
I've done plenty of investigation and know how to set up the remote gateways with D-CAC and RTR to check for link down on loopback interfaces, etc, and also how to configure the network regions to cope with the intervening region that we run..
BUT - If there is virtually no network downtime between MPLS failing and the backup DSL kicking in (HSRP does it job really well!) then how does the RTR establish that the link has actually gone down??
I will be setting loopback 1 interface as follows as our 10Mb MPLS connection only has 200k EF:
interface Loopback 1
dynamic-cac 200 255
keepalive-track 1
As soon as the backup DSL kicks in the RTR will see 200k available bandwidth, albeit ungoverned for QOS, so how will loopback 1 ever go down thus forcing calls over IGAR?
If RTR was to look at ping times that would make sense as they would be much lower over MPLS than backup DSL..
I'm rather confused - can anybody explain how it works?
thanks in advance
Oldskool
Wondering if you can help with something obvious, as I might not be seeing the wood for all the trees!
Running CM5.2.1, with resilient S8730's and G650 gateways at the core, with lots of G350/430/450's that have LSP capability at remote sites.
We have 10Mbps MPLS connections between all sites and have just installed low bandwidth backup DSL links for failover by HSRP to each site. I don't want VOIP calls running over the backup DSL as this is only for mission critical traffic, instead I'd like calls to route out via IGAR instead.
I've done plenty of investigation and know how to set up the remote gateways with D-CAC and RTR to check for link down on loopback interfaces, etc, and also how to configure the network regions to cope with the intervening region that we run..
BUT - If there is virtually no network downtime between MPLS failing and the backup DSL kicking in (HSRP does it job really well!) then how does the RTR establish that the link has actually gone down??
I will be setting loopback 1 interface as follows as our 10Mb MPLS connection only has 200k EF:
interface Loopback 1
dynamic-cac 200 255
keepalive-track 1
As soon as the backup DSL kicks in the RTR will see 200k available bandwidth, albeit ungoverned for QOS, so how will loopback 1 ever go down thus forcing calls over IGAR?
If RTR was to look at ping times that would make sense as they would be much lower over MPLS than backup DSL..
I'm rather confused - can anybody explain how it works?
thanks in advance
Oldskool
Comment