ip sla monitor reaction-configuration & reaction-trigger

[mercedesbroening]

Hi,

My goal is to cause a route change based on a MOS threshold violation for the following:

ip sla monitor 11

type jitter dest-ipaddr 10.200.200.101 dest-port 16486 codec g729a

tag VPN-SF-Boston-VoIP-UDP-Jitter-SLA

frequency 30

I am unclear how to do this. I can do a reaction-configuration to catch the MOS violation but how do I trigger the route change to a cleaner line?

Thanks

[Pavel Bykov]

If I understand correctly, you need to configure action type "trigger" or "trigger and traps" (e.g. using "ip sla reaction-configuration"), and then configure ip sla reaction-trigger that refers to your operation 11 and triggers another operation where you would define cleaner line preference (metric, next hop, or something else).

I do not have much experience with this, so take it with a grain of salt

[mercedesbroening]

Ahh yes, that is the rub. A trigger will only trigger another ip sla object.. I am unclear if there is a way I can get the triggered object to cause a route change..

[elwood]

Did you ever get this to work? I am trying to do the same thing now, as the routing protocol won't trigger use of a backup network if MOS falls below our configured threshold. If you found away, I'd appreciate a replied post showing how it was done.

thank you

[haroon.shaikh]

Recently I had a case which might be helpful to you:

One of our branches had 2 routers: One managed by us and another managed by service provider. This SP router had DSL link and was connected to our router via Ethernet. If the DSL goes down the Ethernet link still stays up so our router never knew of the lost link.

What we did was configured an ip sla monitor which pings the DSL ip and if it fails 5 times, it causes a TCL to run which changes the route back and forth.

You might be able to find some TCL scripts on cisco's SLA site.

[mercedesbroening]

I never figured out how to make it work with a mos value, but I did figure out how to use it with a latency violation. And since mos value is calculated from latency (and a few other things, like jitter) I was ok with that. But then I decided for my purposes it was simpler to do the following:

You have probably seen this before but this is what I did:

ip sla monitor 20

type echo protocol ipIcmpEcho 10.200.200.2

frequency 30

!

ip sla monitor 21

type echo protocol ipIcmpEcho 10.200.200.102

frequency 30

!

!

ip sla monitor group schedule 2 20-29 schedule-period 30 start-time now life forever

!

!

!

!

track 20 rtr 20

!

track 21 rtr 21

!

!

route-map IP-SLA permit 10

match ip address "put your access list here"

set ip next-hop verify-availability 10.200.200.2 1 track 20

set ip next-hop verify-availability 10.200.200.102 2 track 21

!

And then apply this to the correct interface.

you can tweak how many packets are sent and the timeout rate to determine if the interface is up or down, so it is pretty sensitive. you can also configure how quickly the routes flip back to avoid flapping

[elwood]

Thank you,

I am still working on this too , and was going down the path of latency/rtt response. I appreciate the config post and feedback.

-