On 03/05/2013 19:29, Ulrich Herberg wrote:
Alexey,
thank you very much for your review. Please see my reply below:
On Fri, May 3, 2013 at 6:39 AM, Alexey Melnikov
<[email protected]> wrote:
I am the assigned Gen-ART reviewer for this draft. For background on
Gen-ART,
please see the FAQ at
<http://wiki.tools.ietf.org/area/gen/trac/wiki/GenArtfaq>
Document: draft-ietf-manet-olsrv2-mib-07
Reviewer: Alexey Melnikov
Review Date: 03 May 2013
IETF LC End Date: 23 April 2013
IETF Telechat Date: 16 May 2013
Summary: This draft is ready for publication as Proposed Standard (with
nits)
- Major Issues: None
- Minor Issues:
In multiple places:
This object is persistent and when written
the entity SHOULD save the change to
non-volatile storage."
Why is this a SHOULD? I.e., is there a good reason not to save
such values (other than crashing while doing that, which I think is
not something you can prevent anyway.) It really looks like MUST
is actually appropriate here.
This is in accordance with the NHDP-MIB Module (RFC6779). The
rationale is that MANET routers may be very constrained in terms of
memory, much more than classical Internet routers. So while it makes a
lot of sense to recommend storing the information persistently, it may
just not be feasible (running out of memory) or undesirable (too
little memory / no memory).
I think providing some examples for why this is a SHOULD (as you
explained above) would be valuable.
_______________________________________________
Gen-art mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/gen-art
