Letter ballot comments on THRESHOLD EXPONENT field (fixed)
George.Penokie at lsi.com
Thu Aug 22 13:01:12 PDT 2013
Formatted message: <a href="http://www.t10.org/cgi-bin/ac.pl?t=r&f=r1308222_f.htm">HTML-formatted message</a>
Ignore the first note.
OK, so Fred thinks this wording would work.
A threshold exponent field indicates the threshold set size. A threshold
exponent field set to zero indicates that the logical unit does not support
logical block provisioning thresholds.
If logical block provisioning thresholds are supported, then the threshold
exponent shall be a non-zero value selected such that:
Bye for now,
3033 41 St NW
Rochester , MN 55901
george.penokie at lsi.com
From: Knight, Frederick [mailto:Frederick.Knight at netapp.com]
Sent: Wednesday, August 21, 2013 11:43 AM
To: Kevin D Butt; Roger Hathorn; Penokie, George; T10 Reflector
Subject: RE: Letter ballot comments on THRESHOLD EXPONENT field
Yup, Kevin is correct. We shouldn't really invert this because of the LOG
page use of this field.
From: owner-t10 at t10.org<mailto:owner-t10 at t10.org> [mailto:owner-t10 at t10.org]
On Behalf Of Kevin D Butt
Sent: Tuesday, August 20, 2013 1:21 PM
To: Roger Hathorn; George.Penokie at lsi.com; T10
Subject: Fw: Letter ballot comments on THRESHOLD EXPONENT field
IBM does not accept this revision to SBC-3. See below for the reasoning.
Kevin D. Butt
SCSI Architect, Tape Firmware, T10 Standards
Data Protection & Retention
MS 6TYA, 9000 S. Rita Rd., Tucson, AZ 85744
Fax: 520-799-2723 (T/L:321)
Email address: kdbutt at us.ibm.com
----- Forwarded by Kevin D Butt/Tucson/IBM on 08/20/2013 10:19 AM -----
From: Roger Hathorn/Tucson/IBM
To: Kevin D Butt/Tucson/IBM at IBMUS,
Date: 08/19/2013 06:20 PM
Subject: Re: Fw: Letter ballot comments on THRESHOLD EXPONENT field
I completely disagree with this statement:
"A threshold exponent field set to a non-zero value indicates the logical
unit supports logical block provisioning thresholds (see 220.127.116.11)."
That is a change to the current definition and adds a normative requirement
that if I support resource counts in the LBP log page, then I shall also
support logical block provisioning thresholds, which also states that I have
to support the logical block provisioning mode page.
I am OK with saying "A threshold exponent field set to zero indicates that
the logical unit does not support logical block provisioning thresholds.",
because that is already in the draft. I am not OK with the inverse.
This is the main reason for my comment that was rejected:
The threshold exponent is also used for expressing available and used
resource counts in the LBP log page. One could support providing of the log
page resource counts without supporting LBP thresholds.
The response was:
"No change - The one usage of threshold exponent in the LBP log page
references back to this VPD page and now has modified wording to make the
i.e, clearer (see SBC-3 revision 35f or later). The threshold exponent
referenced in the LBP log page is also the same threshold exponent defined in
the VPD page. If the contents of the THRESHOLD EXPONENT field are non-zero
then the device supports these thresholds. It is not just implied it is
stated as such in the text."
The content of the LPB log page does not contain thresholds, it contains
resource counts. They are useable without supporting thresholds. I don't see
where it is stated anywhere in the letter ballot version of the text.
Feel free to forward this response.
Roger G. Hathorn
STSM, Storage Systems Development
IBM Systems and Technology Group
Tel: 520-799-5950 (T/L: 321-5950)
More information about the T10