| Two possible answers:
1. In a dual-attached concentrator, to put one on the primary and one on
the secondary ring, so you can hear management etc. requests on either.
This assumes that there are management nodes which are connected only
to the secondary ring and not to the primary. We don't think this makes
much sense: the primary ring is primary -- so you're either connected
to that ring, or perhaps to both rings -- but not to the secondary only.
2. "Roving MAC" case: to run some proprietary protocol or algorithm at
station insertion. A commonly mentioned example is "graceful insertion"
which claims to allow station insertion without loss of the token.
Apart from being non-standard, such a procedure has questionable merit
in practice. It also adds greatly to the complexity (and therefore
the cost and failure rate) of the concentrator.
paul
|
|
Re. -1
Would you please describe in more detail how the 'grateful insertion'
works ? I know this word from the 'system level description' but
I never know how it operate. Are we the only vendor providing this
feature ? If so, do you think we need to emphasis this feature
to the customer, just from 'selling' point of view ?
I have another question, but not related to the above question.
According to the manual, our bridge adapts RFC1042 (RFC1103 ?) for
packet translation. Then for the following configuration,
VAX or any
other host
on Ethernet
|
___|_________DECbridge 500 ---VAX on FDDI Ring
Suppose systems on both Ethernet and FDDI ring talk via IP and
DECbridge just get a packet from the VAX on the FDDI ring. What
I want to ask is that: How can the DECbridge 500 knows that the
VAX (or other system) on the Ethernet side is speaking 802.3+802.2
or Ethernet format so that it can translate it into correct format
? Does it convert the packet to Ethernet format by default or it
just sends two packets (one for Ethernet and one for 802.3+802.2)
onto the Ethernet network (IBM-8209 works like this ?) ?
Thanks in advance.
Kin
|