Question re Frame Relay FECN reflection

[r.waterman]

Hello

I have some queries re FECN reflection.

It's mentioned (in CCIE cert guide 3rd ed) that FR Switches & routers can send Q.922 test frames back to the sender.

i) What advantage is there in the destination FR DTE (router) sending back Q.922 frames if the FR Switch that originally noticed the congestion and is tracking the VC is going to send back a BECN anyway?

Is this just a matter of "just in case the FR switch that originally noticed congestion doesn't send a BECN back to the sender" type thing?, or perhaps due to the possibility of the route being taken by the PVC in one direction being different to the PVC established in the opposite direction?

[mheusing]

Hi,

You are right, one of the reasons could be different pathes for a PVC in different directions (though implementation usually avoids this). Another reason is not to rely on return traffic: what if you stream data in one direction and there is little to no return traffic to send the BECN? A test frame is more reliable to deliver the congestion information in a timely manner.

Regards, Martin