18 Nov 2010 12:05
RFC 5066 Comment on PME initialization and aggregation
Baud, Yann <yann.baud <at> schmid-telecom.ch>
2010-11-18 11:05:09 GMT
2010-11-18 11:05:09 GMT
I wish to comment on the
penultimate paragraph of RFC 5066 section 3.1.3:
*Quoted*
Note
that the PCS port does not have to be operationally 'down'
for
the
connection to succeed. In fact, a dynamic PME addition
(and
removal) MAY be implemented with an available PME being initialized
first (by setting its ifAdminStatus to 'up') and then added to an
operationally 'up' PCS port, by modifying a respective ifStackTable
(and respective ifInvStackTable) entry.
removal) MAY be implemented with an available PME being initialized
first (by setting its ifAdminStatus to 'up') and then added to an
operationally 'up' PCS port, by modifying a respective ifStackTable
(and respective ifInvStackTable) entry.
In fact, PME addition to a
PCS is impossible once the PME has been initialized, because
the discovery and
aggregation phases, which are done using ITU-T G.994.1 (G.Hs), must
occur before PME training and
activation in the PME
initialization process. Once the PME
is initialized, it is impossible
to configure the
aggregation on the peer pme (IEEE 802.3 2008 61.4.7).
I would much welcome a
clarification on this paragraph.
Best
Regards,
Yann
_______________________________________________ Hubmib mailing list Hubmib <at> ietf.org https://www.ietf.org/mailman/listinfo/hubmib
RSS Feed