FCP-2 Annex E

Robert Snively rsnively at Brocade.COM
Wed Oct 25 07:19:19 PDT 2000


* From the T10 Reflector (t10 at t10.org), posted by:
* Robert Snively <rsnively at Brocade.COM>
*

I have removed Annex E from FCP-2 for the following reasons:

A)  Part of it is wrong, making incorrect assumptions about 
	host adapters.

B)  The remainder of it is redundant with the examples in
	Annex D.

Please indicate if you agree with me or not.

Following is the text of the key comments that caused me to take
this approach.  Note that I have it fully updated if you indicate
strongly to me that it is a valuable portion of the document and
force me to put it back in.

A)  Incorrect:

4.632, Comment from George Penokie (Editorial)
	Page 103 - section e.2.1 - There is something very wrong with this 
	section. Although the sentences appear to be complete taken one at a 
	time when put together they do not make any sense. This section 
	needs to be rewritten to make it clear as to what is going on.

    Response:

	Accepted.

    Installation:

	I have rewritten this example, but I feel that the entire annex is 
	very suspect and should be deleted. It makes assumptions about 
	device behaviors that are well outside the norm and should not be 
	legitimized by such an annex. In particular, the assumption in 
	figure E.1 that host adapters cannot determine that an RO or SEQ_CNT 
	was missing seems very unlikely. That case also states incorrectly 
	an assumption that a "Complete" exchange can be retried. The case in 
	figure E.2 corresponds very closely to the normal behavior described 
	in the previous annex.

	For now, I am removing the annex from revision 5 and publishing it 
	as a separate document, pending approval for that action in the T10 
	meetings.


B)  Redundant:

4.658, Comment from George Penokie (Editorial)
	Page 106 - section e3.1 - There is something very wrong with this 
	section. Although the sentences appear to be complete taken one at a 
	time when put together they do not make any sense. This section 
	needs to be rewritten to make it clear as to what is going on.

    Response:

	Accepted.

    Installation:

	The example was rewritten. Note that this case is very similar to 
	figure D.19 in Annex D, but takes a different and equally optional 
	recovery process. I have removed this along with the rest of Annex 
	E. If anyone has a deep love for this, I would propose including it 
	as a second figure right after figure D.19 in Annex D.
*
* 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 mailing list