[MIB-DOCTORS] 'MIB Doctor' review of draft-ietf-pce-pcep-mib-06

[Juergen Schoenwaelder <[e-mail]>]

Hi,

please find below my review of draft-ietf-pce-pcep-mib-06. Overall,
the document is in a pretty good shape. It is nice to review MIB
modules that are well put together.

The MIB module compiles fine with smiling 0.4.8 and produces no
warnings. Below is a list of comments in document order of appearance.
Some may require an action other maybe not. I have marked the ones
that I find more important with (*).

a) I guess you should spell out PCE In the title.

b) The abstract says the I-D defines an "experimental portion" of the
   MIB while the document is marked as standards-track. This does not
   seem to be consistent.

c) In section 2, ATD should STD.

d) The LAST-UPDATED and REVISION clauses as well as the Copyright year
   in the DESCRIPTION clause of the MODULE-IDENTITY all date back to
   2013 even though the I-D was last updated in 2014. You may want to
   move to a date in 2014.

e) I suggest to replace pcePcepMIBObjects with pcePcepObjects. This
   aligns more with the recommendation given in appendix D of RFC
   4181.

f) I think the description of pcePcepEntityTable should use plural,
   that is, "... about PCEP Entities". Or will this table always have
   only a single row?

g) If you want to describe the enums of an INTEGER, please do so in
   the DESCRIPTION clause and not in comments (since comments may not
   be preserved by tools processing MIB modules).

h) pcePcepEntityAddr seems to serve two purposes, depending on whether
   the peer is a pce and a pcc. The pcePcepPeerRole object makes me
   believe an entity can also be both at the same time. If so, then
   how does one interpret pcePcepEntityAddr? (*)

i) pcePcepEntityConnectMaxRetry has a UNITS clause saying it is in
   seconds. This might be a cut and paste error?

j) pcePcepPeerAddr has a length restriction which essentially also
   restricts pcePcepPeerAddrType. pcePcepPeerAddrType kind of has a
   restriction inline but not explicit via subtyping. It might be more
   inline with the INET-ADDRESS-MIB to not restrict this. If you want
   to restrict things, you may want to do that in a compliance
   statement. (*)

   Despite future proofing considerations, the restrictions make it
   impossible to have PCE entities with non-global that is non-unique
   addresses (because you may not be able to disambiguate them).

k) pcePcepPeerDiscontinuityTime refers to RFC 1907. I think the
   current version it should refer to is RFC 3418.  In any case, if
   you think explicit text concerning sysUpTime is needed, I think
   this warrants a normative reference added to RFC 3418 and a
   statement in section 5.5. (I am not sure this is really needed
   since you kind of import knowledge about sysUpTime via the
   TimeStamp TC you are using.

j) The *AvgRspTime, *LWMRspTime, *HWMRspTime objects have a resolution
   of seconds. This means that times < 500ms will be represented as 0
   seconds and thus indistinguishable from no message at all. I do not
   know whether this is an issue or not. Is it possible to have
   sub-second response times? (*)

k) pcePcepSessState typo: entiries -> entries

l) pcePcepSessPeerDeadTimer typo: for for -> for


-- 
Juergen Schoenwaelder           Jacobs University Bremen gGmbH
Phone: [phone]         Campus Ring 1, 28759 Bremen, Germany
Fax:   [phone]         <http://www.jacobs-university.de/>