Cell Relay Archive

[Andy Bierman <abierman@cisco.com>]

For SVCs, and in most cases, the "probe-tap" (type-D) can monitor
the the signalling channel (VPI 0, VCI 5) to determine the NSAP
addresses associated with particular traffic on the VCCs in use
on the "tapped" link. 

For PVCs, a type D probe would have to obtain info from one
or both switch agents, or be configured with the PVC info.

> 
> Is the information from the ATM Supplemental MIB (section 5.3.4) dealing
> with giving this information to the NMS ?
> 
> Am I missing something really basic here ?

This section refers to support for UNI SVC Information Elements,
it does note that the Calling Party IE is not mandatory (which does
make implementation a little harder ;-), but it's not the text
referenced in the ATM-RMON document.

If implemented by the ATM switch agent, the type D probe can use
the atmVclAddrBindTable to determine the addresses associated
with a PVC, if any.

> 
> This is a basic query about the way the PVCs are setup :
> 
> When the NMS configures the endpoint addresses on each PVC to be
> monitored, the mapping information to the VPI-VCIs is fixed permanently.
> So for a given VPI-VCI the endpoint addresses (NSAP addresses) are always
> fixed. Is this what happens ?
> 

Yes. ("permanent" means until the PVC breaks, except for Soft PVCs)
PVCs are configured by a network administrator and are static in nature.
SVCs are dynamically configured by a signalling protocol (e.g., UNI 3.1).

> Yatin

Andy