Re: [EFM] RE: OAM Proposals - a ping by any other name
- To: Roy Bynum <rabynum@xxxxxxxxxxxxxx>
- Subject: Re: [EFM] RE: OAM Proposals - a ping by any other name
- From: Geoff Thompson <gthompso@xxxxxxxxxxxxxxxxxx>
- Date: Mon, 22 Apr 2002 10:13:25 -0700
- Cc: "Romascanu, Dan (Dan)" <dromasca@avaya.com>, <bob.barrett@fiberintheloop.com>, "Taborek, Rich" <rich.taborek@intel.com>, "Martin Nuss" <nuss@internetphotonics.com>, <Kevin.Daines@worldwidepackets.com>, <hsuzuki@cisco.com>, <MSquire@hatterasnetworks.com>, "Richard Brand"<rbrand@nortelnetworks.com>, stds-802-3-efm@ieee.org
- In-Reply-To: <5.1.0.14.2.20020422090726.01b88b80@xxxxxxxxxxxxxxxxxx>
- References: <AAB4B3D3CF0F454F98272CBE187FDE2F01396851@xxxxxxxxxxxxxxxxx al.avaya.com>
- Sender: owner-stds-802-3-efm@majordomo.ieee.org
Roy-
You point is well taken about every MAC chip (or equivalent) in every piece
of test equipment requiring replacement.
However, I don't think that was Dan's point. I don't think he was talking
about test equipment. He seemed to be talking about imbedded SNMP
management. Same problem.
Geoff
At 09:11 AM 4/22/02 -0500, Roy Bynum wrote:
>Dan,
>
>I agree with you. Preamble as it is may require a new chip set for all
>the test equipment. A baseline including the "null" frame of a preamble
>without an Ethernet MAC frame will require the replacement of every piece
>of Ethernet test equipment that every enterprise and service provider
>currently has. I am not sure that the industry can support that right now.
>
>Thank you,
>Roy Bynum
>
>
>At 01:10 PM 4/22/2002 +0300, Romascanu, Dan (Dan) wrote:
>> >
>> > A system will send an OAMinP preamble approximately once
>> > every 125u seconds,
>> > either by substitution on the next available frame, or by
>> > sending a null
>> > frame. The null frame to be of current min size PDU, or
>> > shorter, if EFM
>> > defines such a new PDU for null frame.
>> >
>>I might have missed some previous discussions, but why should EFM define
>>a PDU null frame of a size shorter than the current min size PDU. This
>>would break all existing performance monitoring tools, and require a
>>re-shuffle of the size distribution counters in the existing standard
>>Ethernet and RMON MIBs.
>>
>>Regards,
>>
>>Dan