iSCSI Boot - LUN size

[Black_David@emc.com]

With WG co-chair hat off:

I realize that I'm supposed to be closing this issue,
but I think Mark has a point here.  To illustrate that
point, let me start with Luben's concise rebuttal:

> The argument regarding the error-prone user input and GUI problems
> has NOTHING to do with the issue. Those are HCI (human computer
> interaction) issues and should be raised as such elsewhere.

There's a hidden assumption buried in there, and that is that
all of the HCI issues caused by the introduction of iSCSI into
an environment have been addressed (i.e., all the appropriate
software components that have any need to be iSCSI-aware have
been appropriately modified).  While that's a fine assumption
for the long term, there are crucial components for which it
is manifestly not true in the short term, the DHCP servers.

When iSCSI goes into an environment, it has to work with the
existing DHCP server(s), not the possible future one(s) that
will be modified to deal with iSCSI HCI issues ... and those
modification requests are somewhere in a large stack of other
things that the DHCP server developers need to worry about.
Mark's point was that allowing a 16-bit LUN format (NOTE:
not requiring its use, 64 bits are still needed for full
functionality/generality) will simplify a transition issue
with *existing* DHCP servers.  I tend to agree.

With WG co-chair hat on:

The above issue is the only technical issue against the boot
draft.  It needs to be resolved here on the list.  Once it's
resolved, a new version of the boot draft containing the
resolution and responses to editorial comments will be forwarded
to the ADs/IESG - a second WG Last Call will not be necessary.

Thanks,
--David
---------------------------------------------------
David L. Black, Senior Technologist
EMC Corporation, 42 South St., Hopkinton, MA  01748
+1 (508) 249-6449            FAX: +1 (508) 497-8018
black_david@emc.com       Mobile: +1 (978) 394-7754
---------------------------------------------------