From: "Elliott, Robert (Server Storage)" <Elliott@hp.com>
To: "t10@t10.org" <t10@t10.org>
Date: Wed, 30 Apr 2008 23:38:35 +0000
Subject: RE: Transports per physical phy
X-Message-Number: 8752
Formatted message: HTML-formatted message

There should be one TT and one MT per SAS logical phy.	An 8-wide HBA with
multiplexing can have connections open to 16 STP or SMP targets at the same
time.  Although the diagram otherwise recognizes the logical phy concept,
that statement is still broken.
If a connection is already open to STP target N and a connection request
comes in from STP target N, something is wrong - multiple connections for one
I_T nexus in STP is not allowed.  If a connection request comes in from an
SMP target, something is wrong - they are not supposed to send connection
requests.
---
Rob Elliott, HP Industry Standard Server Storage
________________________________
From: Bill.Martin@Emulex.Com [mailto:Bill.Martin@Emulex.Com]
Sent: Wednesday, April 30, 2008 6:12 PM
To: Elliott, Robert (Server Storage); t10@t10.org
Subject: Transports per physical phy
Rob:
In figure 29 of SAS-2, there is a statement at the bottom of the figure of
"One or more SAS phys (with matching number of TTs and MTs) per SAS port"
This implies that there is only one STP or SMP transport layer per physical
phy.  Is that the intent of this diagram, and what is the response if there
is one STP or SMP connection on one logical phy when a second connection
request comes in on the other logical phy of this physical phy?  This is
analogous to multiple connection requests on multiple physical phys (which
might be part of a single wide port).  In the later case the request would be
rejected by the STP or SMP layer, not at the port layer.
I am not sure if the diagram needs to be corrected or if there is some other
answer already available in the specification.
Thanks,
Bill Martin
Emulex
Office of Technology
Industry Standards
916 772-3658
916 765-6875 (Cell)
bill.martin@emulex.com