RE: iSCSI: v15 R2T and DATA-OUT

[Paul Koning <ni1d@arrl.net>]

>>>>> "Eddy" == Eddy Quicksall <eddy_quicksall@ivivity.com> writes:

 Eddy> There should not be a requirement for the initiator to check
 Eddy> for consistency. This would just serve to impact
 Eddy> performance. It would be OK to check but the target should be
 Eddy> responsible to set the LUN correctly and the initiator should
 Eddy> just copy it as is.

What I would say is:

1. The target is responsible for sending the LUN correctly in R2T.
2. The initiator doesn't have to check it.

and that is all.

This leaves it unspecified what the initiator might do if the target
commits a protocol violation: it might send back the correct LUN, or
the wrong value from the R2T, or it might close the connection, or ...
I definitely would not want to see an expectation created that a
target can expect to see the LUN from the R2T echoed back to it if it
puts a value other than the correct one in there.

    paul