SAS Transport Layer Retries Teleconference Question

Kevin D Butt kdbutt at us.ibm.com
Wed Feb 5 11:40:37 PST 2003


* From the T10 Reflector (t10 at t10.org), posted by:
* Kevin D Butt <kdbutt at us.ibm.com>
*
This is a multipart message in MIME format.
--=_alternative 006BE3A607256CC4_=
Content-Type: text/plain; charset="us-ascii"


Tivoli Storage Manager (TSM) will fail the current operation upon
receipt of an aborted status and the job will fail.  TSM would not
survive this scenario.

Kevin D. Butt
Fibre Channel & SCSI Architect, IBM Tape Microcode, 
6TYA, 9000 S. Rita Rd., Tucson, AZ  85744
Tie-line 321; Office: 520-799-5280, Lab: 799-5751, Fax: 799-4138, Email:
kdbutt at us.ibm.com 



	"Jim Jones (Engineering)" <Jim.Jones at quantum.com> 
Sent by: owner-t10 at t10.org 


01/28/2003 05:25 PM 
        
        To:        T10 Reflector <t10 at t10.org> 
        cc:        "'John Lohmeyer'" <lohmeyer at t10.org>, "Dave Peterson
(E-mail)" <dap at cisco.com>, Paul Entzel <Paul.Entzel at quantum.com> 
        Subject:        SAS Transport Layer Retries Teleconference
Question 





* From the T10 Reflector (t10 at t10.org), posted by:
* "Jim Jones (Engineering)" <Jim.Jones at Quantum.com>
*
Paul Entzel and I worked with Dave Peterson on forming the question to
ask
backup application providers.  Here it is:

The T10 committee is currently working on a standard that does not
include
transport layer error recovery.  Errors detected at the transport layer
would result in commands returned with aborted command status.  It is
possible that this aborted status may be returned following the
successful
completion of a command.  Some concern has been expressed as to whether
backup applications can deal with this.

The T10 committee is requesting feedback from backup application
providers
indicating the extent of error recovery in current backup applications
(i.e., are the READ POSITION and LOCATE commands used for error
detection
and recovery purposes).

We would like backup application providers to respond to the following
question:

If you receive aborted status for a write, read, or space command that
may
or may not have completed successfully, will the backup application
recover
at the application layer?  

The necessary recovery may include detecting whether the command
completed
successfully (e.g., using READ POSITION), issuing a reposition command
(e.g., using LOCATE), and reissuing the failing command.

For example, if a write command returns aborted command status, the
backup
application needs to determine the current location, reposition to the
location prior to the write command and reissue the write command.

Using another example, if a space command returns aborted command
status,
the backup application needs to determine the current location, and
perform
the appropriate operation to complete the space command.

__________________________
Jim Jones
4001 Discovery Dr., Suite 2100
Boulder, CO 80303
Jim.Jones at quantum.com
720-406-5611


*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org




--=_alternative 006BE3A607256CC4_=
Content-Type: text/html; charset="us-ascii"


<br><font size=2 face="sans-serif">Tivoli Storage Manager (TSM) will fail the current operation upon receipt of an aborted status and the job will fail. &nbsp;TSM would not survive this scenario.<br>
<br>
Kevin D. Butt<br>
Fibre Channel &amp; SCSI Architect, IBM Tape Microcode, <br>
6TYA, 9000 S. Rita Rd., Tucson, AZ &nbsp;85744<br>
Tie-line 321; Office: 520-799-5280, Lab: 799-5751, Fax: 799-4138, Email: kdbutt at us.ibm.com</font>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td>
<td><font size=1 face="sans-serif"><b>"Jim Jones (Engineering)" <Jim.Jones at quantum.com&gt;</b></font>
<br><font size=1 face="sans-serif">Sent by: owner-t10 at t10.org</font>
<p><font size=1 face="sans-serif">01/28/2003 05:25 PM</font>
<br>
<td><font size=1 face="Arial">&nbsp; &nbsp; &nbsp; &nbsp; </font>
<br><font size=1 face="sans-serif">&nbsp; &nbsp; &nbsp; &nbsp; To: &nbsp; &nbsp; &nbsp; &nbsp;T10 Reflector <t10 at t10.org&gt;</font>
<br><font size=1 face="sans-serif">&nbsp; &nbsp; &nbsp; &nbsp; cc: &nbsp; &nbsp; &nbsp; &nbsp;"'John Lohmeyer'" <lohmeyer at t10.org&gt;, "Dave Peterson (E-mail)" <dap at cisco.com&gt;, Paul Entzel <Paul.Entzel at quantum.com&gt;</font>
<br><font size=1 face="sans-serif">&nbsp; &nbsp; &nbsp; &nbsp; Subject: &nbsp; &nbsp; &nbsp; &nbsp;SAS Transport Layer Retries Teleconference Question</font>
<br></table>
<br>
<br>
<br><font size=2 face="Courier New">* From the T10 Reflector (t10 at t10.org), posted by:<br>
* "Jim Jones (Engineering)" <Jim.Jones at Quantum.com&gt;<br>
*<br>
Paul Entzel and I worked with Dave Peterson on forming the question to ask<br>
backup application providers. &nbsp;Here it is:<br>
<br>
The T10 committee is currently working on a standard that does not include<br>
transport layer error recovery. &nbsp;Errors detected at the transport layer<br>
would result in commands returned with aborted command status. &nbsp;It is<br>
possible that this aborted status may be returned following the successful<br>
completion of a command. &nbsp;Some concern has been expressed as to whether<br>
backup applications can deal with this.<br>
<br>
The T10 committee is requesting feedback from backup application providers<br>
indicating the extent of error recovery in current backup applications<br>
(i.e., are the READ POSITION and LOCATE commands used for error detection<br>
and recovery purposes).<br>
<br>
We would like backup application providers to respond to the following<br>
question:<br>
<br>
If you receive aborted status for a write, read, or space command that may<br>
or may not have completed successfully, will the backup application recover<br>
at the application layer? &nbsp;<br>
<br>
The necessary recovery may include detecting whether the command completed<br>
successfully (e.g., using READ POSITION), issuing a reposition command<br>
(e.g., using LOCATE), and reissuing the failing command.<br>
<br>
For example, if a write command returns aborted command status, the backup<br>
application needs to determine the current location, reposition to the<br>
location prior to the write command and reissue the write command.<br>
<br>
Using another example, if a space command returns aborted command status,<br>
the backup application needs to determine the current location, and perform<br>
the appropriate operation to complete the space command.<br>
<br>
__________________________<br>
Jim Jones<br>
4001 Discovery Dr., Suite 2100<br>
Boulder, CO 80303<br>
Jim.Jones at quantum.com<br>
720-406-5611<br>
 <br>
<br>
*<br>
* For T10 Reflector information, send a message with<br>
* 'info t10' (no quotes) in the message body to majordomo at t10.org<br>
</font>
<br>
<br>
--=_alternative 006BE3A607256CC4_=--




More information about the T10 mailing list