[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Review Request for draft-orly-atommib-rfc3895bis-00.txt



Hi,
 
Since I have to fix the enum names, I will also fix the other two issues marked in red. That will eliminate the very last comment too.
I will do so after May 2nd.
 
Orly


From: C. M. Heard [mailto:heard@pobox.com]
Sent: Monday, April 10, 2006 23:51
To: MIBS Mailing List
Cc: Romascanu, Dan (Dan); Orly Nicklass
Subject: Re: Review Request for draft-orly-atommib-rfc3895bis-00.txt

On Mon, 10 Apr 2006, Romascanu, Dan (Dan) wrote:
> The author of draft-orly-atommib-rfc3895bis-00.txt required that this
> document be processed as an independent submission sponsored by the AD
> to become a standards track document.
>
> I believe that there is a need for a couple of expert reviews in order
> to ensure that the document is at the proper quality level for
> standards track. C.M. Heard volunteered to perform a MIB Doctor
> Review, and Kam Lam volunteered to help with the transport Protocols review.
>
> The reviewers and all other parties interested in this work are
> invited to send their comments to ietfmibs@ops.ietf.org until May 2nd,
> 2006 the latest.

Dan,

Here is the promised MIB Doctor review of Orly's recently-submitted document draft-orly-atommib-rfc3895bis-00.txt.  Since RFC 3895 itself went through a MIB Doctor review, I believe that it is sufficient to review just the things that have changed since RFC 3895.  And since I did a reasonably thorough job of that on the pre-submission version of Orly's document, I think it suffices this time just to go over the things that I flagged before plus new issues introduced in the latest round of edits.

Before I do that, however, I'd like to make some general comments about the changes relative to RFC 3895.  As noted in Section 2.1 of the draft, the important ones are:

(1) Values were added to dsx1LineType to support J1 types.
(2) The object dsx1LineImpedance was added.
(3) All DM (degraded minutes) related objects were deprecated
    following their removal from ITU and ANSI performance
    monitoring standards.
(4) Relevant text and reference section were updated.
(5) Changes in Compliance Statements to include new values.

I audited the smidiff report (a copy of which is attached), and I agree that the above list is a reasonably accurate summary of what was changed.  The new object dsx1LineImpedance is writeable and is mentioned in the Security Considerations section, as required.  One point that should be noted about the new compliance statements is that in addition to removing the requirement to implement DM-related objects, they add a requirement to implement dsx1LineImpedance.
Thus, existing implementations will need to be updated in order to claim compliance to the 3895bis spec.  That should probably be noted when the document is last-called.  Note that the same approach was taken for RFC 3895 itself, that is, the objects dsx1LineMode and dsx1LineBuildOut that were added in that rev of the spec were made mandatory-to-implement.

Now for the re-check off items flagged in the pre-submission review.

On Saturday, March 11, 2006 C. M. Heard wrote:
> 1.) Sections 1.1 through 1.3 of RFC 3895 have been removed.  These
> cover the revisions back to RFC 1406 and the list of companion
> documents.  I would recommend that these sections be reinstated.
> They should probably be Sections 1.2 through 1.4 in your draft.

Sections 1.1 and 1.2 of RFC 3895 have been reinstated as Sections
2.2 and 2.3 of the 3895bis draft.  Section 1.3 of RFC 3895 has not been reinstated.  I have no objections to dropping that section, but its absence does result in some uncited informative references (more on this below).done

> 2.) I notice that the references to G.704 Table 4a/4b have been
> changed -- in most places -- to G.704 Table 5a/5b, even though the
> underlying reference is the same.  If this was intended, then no
> action is necessary;  I am just checking to make sure that no error
> was introduced.  (If the change is NOT backed out, then presumably the
> reference to Table 4a in the ds1MibE1PriCompliance OBJECT clause for
> dsx1Fdl needs to be corrected).

This was left unchanged, so I assume that no change was required.  I notice that the reference to Table 4a in the ds1MibE1PriCompliance is for something different so I withdraw the parenthetical comment.

> 3.) The date on the REVISION clause for RFC 3985 is "200409190000Z"
> in your draft but is "200409090000Z" in RFC 3895 itself.  The date
> that you need to use is "200409090000Z", i.e., the one in RFC 3895.
> Doing so would obviate the following smilint messages:
>
> /usr/local/share/mibs/ietf/DS1-MIB:48 [3] {revision-removed} revision
> `2004-09-09 00:00' removed
> DS1-MIB.x:66 [5] {revision-added} warning: revision `2004-09-19 00:00'
> added

This has been fixed and is no longer an issue.

> 4.) A typo has crept into the DESCRIPTION clause of dsx1LineCoding:
> s/specifed/specified/ in the last paragraph of the clause.

This has been fixed and is no longer an issue.

> 5.) The DESCRIPTION clause for the new object is a bit thin, and the
> named-number labels could be improved.  Suggested change:
>
> OLD:
>           dsx1LineImpedance   OBJECT-TYPE
>                SYNTAX      INTEGER {
>                               notApplicable (1),
>                               unbalance75ohms (2),
>                               i100ohms (3),
>                               balance120ohms (4)
>                            }
>                MAX-ACCESS  read-write
>                STATUS      current
>                DESCRIPTION
>                       "Impedance setting for E1 lines."
>                ::= { dsx1ConfigEntry 23 }
> NEW:
>           dsx1LineImpedance   OBJECT-TYPE
>                SYNTAX      INTEGER {
>                               notApplicable (1),
>                               unbalanced75ohms (2),
>                               balanced100ohms (3),
>                               balanced120ohms (4)
>                            }
>                MAX-ACCESS  read-write
>                STATUS      current
>                DESCRIPTION
>                       "Nominal line impedance.   For T1 and J1 lines
>                        the value is typically balanced100ohms(3).
>                        For E1 lines the value is typically
>                        unbalanced75ohms(2) and balanced120ohms(4).
>                        When this object does not apply, or when
>                        the appropriate value is not know, the
>                        value should be set to notApplicable (1)."
>                ::= { dsx1ConfigEntry 23 }

The suggested DESCRIPTION text was adopted but the enum labels were not changed to match it:done

          dsx1LineImpedance   OBJECT-TYPE
               SYNTAX      INTEGER {
                              notApplicable (1),
                              unbalance75ohms (2),
                              balance100ohms (3),
                              balance120ohms (4)
                           }
               MAX-ACCESS  read-write
               STATUS      current
               DESCRIPTION
                      "Nominal line impedance.  For T1 and J1 lines the
                      value is typically balanced100ohms(3).  For E1
                      lines the value is typically unbalanced75ohms(2)
                      and balanced120ohms(4).  When this object does not
                      apply, or when the appropriate value is not known,
                      the value should be set to notApplicable (1)."
               ::= { dsx1ConfigEntry 23 }

Either the enum labels should be changed to match the DESCRIPTION text or the DESCRIPTION text should be changed to match the labels.  I would prefer changing the labels (s/balance/balanced/), but can live with either option.done

> 6.) I suggest the following formatting change at the end of the MIB
> module:
>
> OLD:
>              ::= { ds1Groups 13 } END
> NEW:
>              ::= { ds1Groups 13 }
>           END

This has been fixed and is no longer an issue.

> 7.) In Appendix A, which explains the historical usage of dsx1IfIndex
> and dsx1LineIndex in RFC 1406, the reference to RFC 1213 was changed
> to RFC 2863.  This is not appropriate in this context, since the text
> is being quoted from RFC 1406, and the version of the Interfaces Group
> applicable to RFC 1406 was RFC 1213 (RFC 2863 did not show up until
> much later).  The following rollback is therefore
> recommended:
>
> Text in draft-ietf-orly-rfc3895bis-00-5-3-06.txt:
>    The dsx1IfIndex column of the DS1 Configuration table relates each
>    DS1 interface to its corresponding interface (ifIndex) in the
>    Internet-standard MIB [RFC2863].
>
> Please re-instate to text in RFC 3895:
>    The dsx1IfIndex column of the DS1 Configuration table relates each
>    DS1 interface to its corresponding interface (ifIndex) in the
>    Internet-standard MIB (MIB-II STD 17, RFC 1213) [RFC1213].
>
> and re-instate RFC 1213 as an informative reference.

This has been fixed and is no longer an issue.

> 8.) There are two typos in the Security Considerations section:
>
> - at the end of the list of writable objects, please change
> "dsx1LineImpedant" to "dsx1LineImpedance";
>
> - in the paragraph on readable objects, please change "DS1/E1/DS2/E2"
> to "DS1/J1/E1/DS2/E2".

These have been fixed and are no longer issues.

> 9.) The Informative References section needs to be fixed.  Based on
> the discussion above, this is what I would recommend:
>
> 6.2.  Informative References
>
>    [RFC3410]       Case, J., Mundy, R., Partain, D., and B. Stewart,
>                    "Introduction and Applicability Statements for
>                    Internet-Standard Management Framework", RFC 3410,
>                    December 2002.
>
>    [RFC3895]       Nicklass, O., Ed., "Definitions of Managed Objects
>                    for the DS1, E1, DS2, and E2 Interface Types", RFC
>                    3895, September 2004.
>
>    [RFC2495]       Fowler, D., Ed., "Definitions of Managed Objects for
>                    the DS1, E1, DS2 and E1 Interface Types", RFC 2495,
>                    January 1999.
>
>    [RFC1406]       Baker, F. and J. Watt, Eds., "Definitions of Managed
>                    Objects for the DS1 and E1 Interface Types", RFC
>                    1406, January 1993.
>
>    [RFC2494]       Fowler, D., "Definitions of Managed Objects for the
>                    DS0 and DS0 Bundle Interface Type", RFC 2494, January
>                    1999.
>
>    [RFC3896]       Nicklass, O., Ed., "Definitions of Managed Objects
>                    for the DS3/E3 Interface Types", RFC 3896, September
>                    2004.
>    [RFC3592]       Tesink, K., "Definitions of Managed Objects for the
>                    Synchronous Optical Network/Synchronous Digital
>                    Hierarchy (SONET/SDH) Interface Type", RFC 3592,
>                    September 2003.
>
>    [AT&T-UM-305]   AT&T Information Systems, AT&T ESF DS1 Channel
>                    Service Unit User's Manual, 999-100-305, February
>                    1988.
>
>    [RFC1213]       McCloghrie, K. and M. Rose, "Management Information
>                    Base for Network Management of TCP/IP-based
>                    internets: MIB-II", STD 17, RFC 1213, March 1991.

Done;  however, [RFC3592], [RFC2494], and [ANSI-T1.102] are not cited in the text (see above), and the RFC Editor will probably remove them.

All the issues I have raised here are minor, and I see no reason to hold up IETF Last Call on their account.

Regards,

Mike