Re: [Lime-oam-model] OAM Model analisys. First cut

[Gregory Mirsky <[e-mail]>]

Hi Deepak,
many thanks for your suggestions and questions. Please find my notes in-line and tagged GIM>>.

                Regards,
                                Greg

From: Deepak Kumar (dekumar) [mailto:[e-mail]]
Sent: Wednesday, February 04, 2015 6:23 PM
To: Gregory Mirsky; [e-mail]<mailto:[e-mail]>
Subject: Re: [Lime-oam-model] OAM Model analisys. First cut

Hi Greg,

Thanks for detailed Analysis. I have few queries.
In Metrics Is Lock is covered under Defect Indication?
GIM>> Frankly, I haven't thought of Lock. AFAIK, Lock is not to indicate a defect or failure but is the step to turn a service off of the given connection.
How's Test signal get covered?
GIM>> Do you refer to ETH-TST?
Should we talk about Throughput or as it's application ignore for now.
GIM>> Throughput measurement is not an operation but method that can be implemented using different operations.

Metric is covering "Source Information/MEP ID" and "MIP ID", but we should also extend to add other metrics related to Administrative Domain and MEP for completeness.
GIM>> Agree, will add.
For Eg:- Administrative Domain, OAM Technology Layer Type, Level (implicit or NA), MEP addressing and ECMP (Flow Entropy or Customer traffic information).
GIM>> In MPLS-TP Level is realized through SPME, which is a hack, IMO.
We should also add metrics for Server Layer, Client Layer and also concept similar to Server MEP (Which is lower layer of OAM sending AIS/LCK to upper layer to suppress Alarms).
GIM>> I think that Server and Client Layers are distinct only in
Also this concept is useful in OAM interworking between different technology also.

I believe providing information regarding server layer in data model will allow application using data model to efficiently go down one layer at a time and Isolate the Fault.

GIM>> I agree that will be beneficial. To add to your point, having Defect data model shared among reviewed OAM models will help as well.

Thanks,
Deepak

From: Gregory Mirsky <[e-mail]<mailto:[e-mail]>>
Date: Tuesday, February 3, 2015 5:22 PM
To: "[e-mail]<mailto:[e-mail]>" <[e-mail]<mailto:[e-mail]>>
Subject: [Lime-oam-model] OAM Model analisys. First cut

Dear All,
apologies for belated update. Attached please find first cut of the proposal for approach to OAM Model comparison and example of it being used.

Your questions, comments, suggestions always welcome and greatly appreciated.

                Regards,
                                Greg