SAM2:SCSI client server model
Sanjeev Bhagat (TRIPACE/Zoetermeer)
iscsi_t10 at sanjeevbhagat.com
Wed Oct 3 06:54:32 PDT 2001
* From the T10 Reflector (t10 at t10.org), posted by:
* "Sanjeev Bhagat \(TRIPACE/Zoetermeer\)" <iscsi_t10 at sanjeevbhagat.com>
Thanks a lot for this response. It certainly doesnt leave any room of doubt
for me. I am very clear now.
This silly doubt came while i was going through the thread on kicker and
ISID issue you are discussing on IPS list
----- Original Message -----
From: "Jim Hafner" <hafner at almaden.ibm.com>
To: "Sanjeev Bhagat (TRIPACE/Zoetermeer)" <iscsi_t10 at sanjeevbhagat.com>;
<t10 at t10.org>
Sent: Wednesday, October 03, 2001 12:24 AM
Subject: Re: SAM2:SCSI client server model
> I'll try to take a stab at this.
> The short answer to your question is NO. There is no relationship between
> the iSCSI connection/session (and FC login) and the application client.
> The I_T nexus is between a SCSI Initiator Port and a SCSI Target Port.
> That can live as a relationship as long as the transport layer wants to
> keep it (or even pretend that it is there).
> The application client sits above the SCSI Initiator Port. It is created
> when a SCSI task needs handling, is encapsulated by CDB or task management
> function and handed to the SCSI Initiator Port for delivery. When status
> is returned, the application client goes away. It's been called a "thread
> of execution" of a single SCSI task. While it may come and go, the SCSI
> Initiator Port and SCSI Target Port may stay and any relationship (like
> login) between them may stay in place as well.
> In short, the application client is above the SCSI Initiator Port and the
> session is below the port (in the implementation of the port's interface
> the transport).
> Does that help?
> Jim Hafner
> "Sanjeev Bhagat \(TRIPACE/Zoetermeer\)" <iscsi_t10 at sanjeevbhagat.com>
> @t10.org on 10/02/2001 01:37:31 pm
> Sent by: owner-t10 at t10.org
> To: "T10 Reflector" <t10 at t10.org>
> Subject: SAM2:SCSI client server model
> * From the T10 Reflector (t10 at t10.org), posted by:
> * "Sanjeev Bhagat \(TRIPACE/Zoetermeer\)" <iscsi_t10 at sanjeevbhagat.com>
> Refer to SAM2r20, section 4.3, page 29
> In the architecture model , an application client is created to issua a
> single SCSI command or task management function. An application client
> ceases to exist once the command or or task management function ends.
> also section 4.7.3 page 27
> This model assumes that there is one application client for each pending
> command or task management function.
> Does the above two lines imply that everytime an application client wants
> send a command to the target it will have to start a new connection and
> session and end it as the command ends??
> * For T10 Reflector information, send a message with
> * 'info t10' (no quotes) in the message body to majordomo at t10.org
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org
More information about the T10