RE: [EFM] Usage of management channel
Ariel
You just convinced me that my proposal for side band (which is similar to
two others) can be done without affecting the payload bandwidth. Given that
SNMP 'works' on 56k and 64k links in-band it would work on 100k (or less)
dedicated side band with no problems. It's a datagram protocol, and it
matters not how or where the datagrams are propagated, so long as they get
to their destination. Personally I would run SNMP agents at the head end and
use a thinner mechanism for remote OAM, but that would be a vendor specific
implementation choice. Either would work in 1GE, band width might be tighter
within the copper spec., but then if it is the ITU PHY there is already a
spec. for an OAM channel that can be leveraged.
Bob
> -----Original Message-----
> From: owner-stds-802-3-efm@majordomo.ieee.org
> [mailto:owner-stds-802-3-efm@majordomo.ieee.org]On Behalf Of Ariel
> Maislos
> Sent: 06 October 2001 21:26
> To: stds-802-3-efm@ieee.org
> Subject: [EFM] Usage of management channel
>
>
>
> Just trying to clarify something that I do not understand.
> What is the use of the OAM channel we are all talking about?
>
> Is it for link management?
> - If so then this is very low rate signaling, especially assuming
> the links
> is as stable as reported.
> - When the bandwidth loss is below the 100ppm tolerance (100ppm for 1G is
> 100K), then it is immeasurable by Ethernet standards.
>
> Is it for station management?
> - If so then the likely protocol would be SNMP.
> - And in that case, how do we route SNMP packets into the OAM
> channel? Don't
> they belong in-band?
>
> Ariel
> Passave
>