[manet] MIB doctor "skim" of draft-ietf-manet-olsrv2-mib-11.txt

[Randy Presuhn <[e-mail]>]

Hi -

In reponse to my earlier comments, the editors of the OLSRv2
MIB module issued an updated I-D, which I understand has been
sent to the manet WG for a sanity check.  In order speed things
along, I did a quick skim of the document this evening.

These comments are in my opinion very minor, with the exception
of RMP-313, but I would appreciate comments from other MIB doctors
on the two "***" items below.  (As well as anything else if you
think I've missed the boat.)

RMP-301
top of page 4: "OLSRV2" should be "OLSRv2"

RMP-302
top of page 4: "intervals etc." -> "intervals, etc."

RMP-303
6.3: needs to mention IANAolsrv2LinkMetricType-MIB in
the list.

RMP-304
Olsrv2MetricValueCompressedFormTC: I'm surprised that
the display-hint is "x" rather than "d".  I hope the
WG discussed this - I'm ok with whatever, as long as
it's a conscious decision and not an oversight.

RMP-305
Olsrv2MetricValueCompressedFormTC: I'm also a little
surprised that the DESCRIPTION talks about set processing,
since it looks to me as though all uses of this TC are
for read-only objects.

RMP-306
Olsrv2MetricValueCompressedFormTC: has the WG discussed
what a management system should do if an implementation
reponds to a get/getnext/getbulk request for an object
of this type with a syntactically legal but mathematically
impossible value, such as 16776959?

RMP-307
olsrv2InterfaceTable: "This object is set either by
network management, or by other means, e.g., CLI."
Is the EXCLUSIVE-OR intentional?  If not, should
delete the "either".

RMP-308
Something I overlooked on previous passes:
olsrv2RxHoldTime, olsrv2PHoldTime, olsrv2FHoldTime,
olsrv2TpMaxJitter, olsrv2TtMaxJitter, olsrv2FMaxJitter:
are the maximum values for these *really* more than
49 days? 

*** RMP-309
olsrv2LinkMetricType: the DEFVAL should probably say
	{ unknown(0) } - I don't know whether all
	MIB compilers out there nowadays are smart
	enough to figure out { unknown } by itself.
	They *should* be able to, but this might be
	a case where conservatism is warranted.

RMP-310
olsrv2TibAdRemoteRouterSetMaxSeqNo and
olsrv2TibAttNetworksSetSeqNo: I seem to recall from
another document that the notion of "greatest" for this
bit of information isn't the same as in math.  If that
is so, it should be spelled out here, as has already been done
in the description for olsrv2TibAdRemoteRouterSetMaxSeqNo.

*** RMP-311
All items with SYNTAX TimeStamp - some might argue whether
UNITS clauses are appropriate for these.  I'm equivocal on
the point.

RMP-312
You might consider whether you'd like to put DEFVAL clauses
on the read-write items in olsrv2NotificationsControl.
DEFVALs are merely advisory - implementations are NOT obligated
to follow them (RFC 2578 clause 7.9).

RMP-313
Section 10: needs to add instructions for the registration
and spin-off of the MIB module with the metric textual convention.
(Those instructions currently are in the first paragraph
of Appendix A.  It's ok by me if thery're repeated there, but
I think it's crucial that IANA's instructions in section 10 are
complete.)

Randy