[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
frameworkpib: Prid and InstanceId question
- To: RAP workinggroup <rap@ops.ietf.org>
- Subject: frameworkpib: Prid and InstanceId question
- From: Freek Dijkstra <freek@macfreek.nl>
- Date: Sat, 3 Aug 2002 23:17:02 +0200 (MET DST)
I just read draft-ietf-rap-frameworkpib-09.txt and got confused. Can
someone please de-confuse (profuse? :-) ) me?
Most data structures defined by a PIB start with some id attribute, which
is of type InstanceId.
It is my understanding that this InstanceId is something different than
the PRID. For one thing, the definition of InstanceId, ReferenceID and
PRID are in RFC 3159. ReferenceID and PRID are different.
However, in draft-ietf-rap-frameworkpib-09.txt most (if not all) classes
start with an attrbute of type InstanceId, and are called
something like 'frwkBaseFilterPrid' (ending with the word Prid).
This is very confusing: is it a Prid or a InstanceId, or are InstanceId's
and Prid's the same?
If they are not the same (as I hope), I strongly suggest to change the
naming in your draft, becaused it really confused me! (Make it
'frwkBaseFilterId' and similar).
Background, why I believed (perhaps incorrectly, but hopefully correct)
that InstanceId's and Prid's are different: It is my understanding that a
PRID is encapsulated in a COPS message in the following way:
COPS common header (client type, action)
COPS message-header (decision, command: install/delete)
COPS-PR header (s-num = 1)
PRID
COPS-PR header (s-num = 3)
PRI
COPS-PR header (s-num = 1)
PRID
COPS-PR header (s-num = 3)
PRI
COPS message-header (named client info?)
COPS-PR header (s-num = 1)
PRID
COPS-PR header (s-num = 3)
PRI
Is this correct? It's quite some time that I tried to figure this
out, by reading RFC 3084 (COPS-PR).
This made me assume that the PRID is an identifier which is created and
maintained by the SPPI / COPS API, while the InstanceId is just an Id
maintained by the application which maintains the rest of the data.
So please tell me that they are really different, and it was just the
naming that confused me.
Thanks and regards,
Freek