iSNS: Last Call technical comments

[Joshua Tseng <jtseng@NishanSystems.com>]

Hi Folks,

A few of the coauthors went through a final round of review
of iSNS revision -14 for WG last call, and came up with the following
two technical comments:

1)  The spec needs to say that Discovery Domain membership for Storage
Nodes MUST be persistent, even after deregistration of a member
Storage Node.  When a deregistered device is re-registered into the
iSNS database, it retains the DD membership state that it had
immediately prior to its deregistration.

The text to address this has been added to the last paragraph of
section 3.2.2.

2)  In iSNS revision -14, attempted deregistration of a non-existent
Discovery Domain (section 6.6.5.10) or Discovery Domain Set (section
6.6.5.12) returned an error code.  We do not believe this is standard
behavior for query-response protocols, and recommend changing this
so that attempted deregistration of non-existent entries does not
return an error code.

The text to address this has been added to sections 6.6.5.10 and
6.6.5.12 to explicitly state that attempted deregistrations of
non-existent entries SHALL not be considered an error.

The remaining comments are all editorial.  I am submitting iSNS
revision -15 to IETF today, but a .pdf version can be found here:

www.nishansystems.com/ietf/draft-ietf-ips-isns-15.pdf

All deltas from the -14 revision, including the editorial updates
and the two technical changes, are highlighted in the .pdf file.

Regards,
Josh Tseng