Re: SCSI URL scheme

["John Hufferd/San Jose/IBM" <hufferd@us.ibm.com>]

Raghavendra Rao,
Today, and perhaps into the future, it is up to the Storage Controller to
permit access, from the approprate Host (initiators), to the approprate LUs
(disks).  They only return to the Report LUNs requestor those LUNs that the
initiators are authorized to see.  I think that is a good thing.

In the example you focus on, says that it is a problem for 100s of  systems
issuing a Report LUNs, and Inquiry commands.  Since most of the remote
systems will probably only request one disk, it will probably be only the
logical LU0 which they are permitted to see.  I do not see a problem there.
In fact even if they have a 100 LUs that they are permitted to get at, I do
not see a problem with the Report LUNs sending back the LU#s that each of
those systems are authorized to use, and those systems issuing the Inquiry
commands to those authorized LUs.  (The interaction is trivial.)

This is the way things work today, and at the moment I do not think it is
broken.   It is the ultimate target that is responsible for enforcing the
authorization process that the installation establishes.  Today, with the
lack of good security in the FC network, we have a problem which we can
address with iSCSI.  But after we get good Authentication, it seems to me
that the Storage Controller enforcement of what initiators can address
which LUs is good and approprate.


.
.
.
John L. Hufferd



Raghavendra Rao <jpr@divyaroot.India.Sun.COM>@ece.cmu.edu on 10/04/2000
09:45:27 AM

Please respond to Raghavendra Rao <jpr@divyaroot.India.Sun.COM>

Sent by:  owner-ips@ece.cmu.edu


To:   ips@ece.cmu.edu
cc:
Subject:  Re: SCSI URL scheme




>
> First off, this is suppose to be done (normally) once, but perhaps at
> various (logical) Plug & Play  points.  If we are talking about the Plug
&
> Play points in time, then there could be some additional conversation.
But
> independent of that, which so far I think we have been, the Host
determines
> ALL his addressable LUs when he comes up, and today that means "REPORT
> LUNS", followed by approprate INQUIRYs.  This is not a big deal at bring
> up.
>

It depends on the storage controller and the configuration. If iSCSI indeed
becomes very popular, it is not hard to visualize a lot of compute nodes
(of the order of 100s just to start with) accessing/sharing a big storage
device that could have a lot of LUs (of the order of thousands at the
least).

It is possible to configure the storage device to selectively report LUs
in the REPORT LUN response so as to limit the number of INQUIRYs fired from
the initiators - But I see this only as a work around - not a solution
to the problem.

I think a better way to translate a LU-WWN to LUN value is required to
avoid
overwhelming number of INQUIRYs fired from all the compute nodes trying
to access the storage device; This is not only wasting the network
bandwidth,
but is also slowing down the host's reconfiguration/booting.

-JP