[bmwg] IGP/Data-plane convergence proposal

BMWG,

Now that we have made significant progress toward completing the
OSPF Convergence I-Ds, it has been proposed that the BMWG support
a new work item relative to 3 individually-submitted I-Ds, titled:

"Benchmarking Applicability for IGP Route Convergence"
http://www.ietf.org/internet-drafts/draft-poretsky-igp-convergence-app-00.txt
"Terminology for Benchmarking IGP Convergence"
http://www.ietf.org/internet-drafts/draft-poretsky-igp-convergence-term-00.txt
"Benchmarking Methodology for IGP Convergence"
http://www.ietf.org/internet-drafts/draft-poretsky-igp-convergence-meth-00.txt

Goal:

If undertaken by the WG, the objective is to produce applicability,
terminology, and methodology drafts that specify the benchmarking
of IGP convergence in routers and systems under test.  Route convergence
is assessed by active measurements of packet transfer, called data plane
measurements to distinguish them from measurements of the control plane.

Scope:

The focus of this effort is to benchmark the route convergence of
Routers running a link state IGP such as OSPF or ISIS.  Other routing
protocols are outside the scope of this effort.

Motivation:

Service Providers seek minimal IGP Convergence time, and that time
is a key metric of system/network design and router design.
Network convergence time may be improved through deployment of fast
converging routers.

Network users consider packet loss a key service quality metric.
Intervals of high packet loss are externally observable events having
direct impact on their application performance.
IGP Route Convergence time is a Direct Measure of Quality (DMOQ)
when benchmarking the data plane's ability to support continuous flows.

For these reasons it is important to develop a standard
methodology and terminology for benchmarking router IGP
convergence that measures performance in the data plane.
The applicability document describes all of the factors
that influence a convergence measurement, and how a purely
black box test can be designed to account for them.

Schedule:

If you wish to offer your thoughts as to whether the BMWG should
or should not undertake the effort as a WG deliverable, please
email your comments to this list or acmorton@att.com no later
than 21 May 03.

_______________________________________________
bmwg mailing list
bmwg@ietf.org
https://www1.ietf.org/mailman/listinfo/bmwg