[T11.3] RE: FC-LS/FCP-4: REC ELS reason code inconsistency

David Peterson dpeterso at brocade.com
Sun Mar 18 20:36:00 PDT 2007


Yes, FCP-2 error detection since its inception specified that the reason
code to return is "Logical error", and the belief is that the
inconsistency was born via the transfer to FC-FS.
Given the inconsistency, we need to specify the reason code and reason
code explanation values that are expected to achieve the proper result.
...Dave
(no disclaimer)
-----Original Message-----
From: Neil Wanamaker [mailto:Neil_Wanamaker at pmc-sierra.com] 
Sent: Sunday, March 18, 2007 7:31 PM
To: Robert Snively; Pappal, Sam; David Peterson; t10 at t10.org;
t11_3 at t11.org
Subject: RE: [T11.3] RE: FC-LS/FCP-4: REC ELS reason code inconsistency
It could be argued that FCP-2 recovery existed long before FC-LS, and
the inconsistency was created when the the FCP-2 link services got
copied into FC-FS. 
Neil Wanamaker
PMC-Sierra
100-2700 Production Way
Burnaby, BC V5A 4X1
604.415.6053 x 2435
Neil_Wanamaker at pmc-sierra.com
-----Original Message-----
From: t11_3-bounces at listserve.com [mailto:t11_3-bounces at listserve.com]
On Behalf Of Robert Snively
Sent: Friday, March 16, 2007 1:31 PM
To: Pappal, Sam; Dave Peterson; t10 at t10.org; t11_3 at t11.org
Subject: [T11.3] RE: FC-LS/FCP-4: REC ELS reason code inconsistency
Sam,
FCP-4 makes use of the tools provided by the lower layer for recovery
capabilities and for verification of continuing forward progress, as
well as various clearing functions.
These are actions mapped from the SCSI Architecture to the FCP-4.  So
yes, it should be using the values presented by FC-LS.	Unfortunately,
through a case of parallel evolution and early implementations, this
discrepancy escaped notice.  Dave's concern is whether correcting the
discrepancy in the standards will cause any implementers undue grief.
Implementations will tell the story of which standard should be
corrected and what correction should be made, since it really doesn't
matter very much architecturally which value is selected.
Bob
-----Original Message-----
From: owner-t10 at t10.org [mailto:owner-t10 at t10.org] On Behalf Of Pappal,
Sam
Sent: Thursday, March 15, 2007 7:33 AM
To: Dave Peterson; t10 at t10.org; t11_3 at t11.org
Subject: RE: FC-LS/FCP-4: REC ELS reason code inconsistency
* From the T10 Reflector (t10 at t10.org), posted by:
* "Pappal, Sam" <Sam.Pappal at gdc4s.com>
*
Maybe I'm missing something here but isn't FCP-4 at a layer above the
FC-LS fibre channel layer and somewhat independent?  If it's a FCP-4
layer detected unknown exchange then the reason code returned
corresponds with the FCP-4 layer (logical error), and if it's a FC-LS
layer detected exchange unknown then the FC-LS layer code is returned
(Unable to perform command request). 
Sam 
-----Original Message-----
From: owner-t10 at t10.org [mailto:owner-t10 at t10.org] On Behalf Of David
Peterson
Sent: Wednesday, March 14, 2007 12:15 AM
To: t10 at t10.org; t11_3 at t11.org
Subject: FC-LS/FCP-4: REC ELS reason code inconsistency
* From the T10 Reflector (t10 at t10.org), posted by:
* "David Peterson" <dpeterso at brocade.com>
*
FCP-2/3/4 error detection implementers,
FC-LS and FCP-4 are inconsistent regarding the proper reason code to
return if the Originator S_ID,=20 OX_ID, or RX_ID fields are
inconsistent (i.e., the Exchange is unknown).
FC-LS specifies the reason code shall be 09h "Unable to perform command
request". Note this reason code has been specified since FC-FS.
FCP-4 specifies the reason code shall be 03h "Logical error". Note this
reason code has been specified since FCP-2.
Please respond indicating the reason code you return and/or expect if
the Originator S_ID,=20 OX_ID, or RX_ID fields are inconsistent, along
with any associated specific reasoning.
And please respond by end of business March 30th so I can address this
issue at the April T11 meeting week.
Thanks...Dave
(no = disclaimer)
*
* 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
_______________________________________________
T11_3 mailing list
http://mailman.listserve.com/listmanager/listinfo/t11_3
_______________________________________________
T11_3 mailing list
http://mailman.listserve.com/listmanager/listinfo/t11_3



More information about the T10 mailing list