FC-TAPE Minutes, April 14-15, 1999, Palm Springs
STEWART_R_WYATT at HP-Boise-om2.om.hp.com
STEWART_R_WYATT at HP-Boise-om2.om.hp.com
Fri Apr 16 09:59:45 PDT 1999
* From the T10 Reflector (t10 at symbios.com), posted by:
* STEWART_R_WYATT at HP-Boise-om2.om.hp.com
Joint T11/T10 FC-TAPE AdHoc Meeting Minutes T11/99-231v0
April 14 and 15, 1999 - Palm Springs, California
Stewart Wyatt, Hewlett Packard, FC_TAPE Secretary,
stewart_wyatt at hp.com, 208-396-3594
1. Introductions: Group
2. Approval of this Agenda: T11/99-170v0 Group, approved as amended
3. Approval of 3/9 Minutes: T11/99-147v0 Stewart Wyatt, approved
4. Review Old Action Items: Stewart Wyatt
Action items carried over from before the March meeting:
#1. Dave Peterson/Bob Snively - Update the error recovery
documentation to resolve the time-out for ELS: Completed
#2. Group - Resolve the issue of signalling support for FC-Tape error
recovery, CRN support and data overlay in either/or inquiry data, mode
pages, PRLI or by some other means. Completed
#3. Bob Snively - Consider whether SRR is SCSI specific. Put in an
annex in FCP-2 for inclusion later in the FC-FS. Completed
#4. Dave Peterson - Report the omission of BLS time-out in FC-PH.
#5. Dave Peterson - Consider modifying the tape model to require that
the tape power up uninitialized, unloaded state. SSC issue. Completed
#6. Dave Baldwin - Investigate eliminating TPRLO. Rejected.
#7. Dave Peterson/group - Resolution of need for authentication
documentation. Completed - referred to FC-FS see minutes.
#8. Dave Peterson will attempt to document EOD behavior more precisely
in the SSC. Completed
Action items from the last meeting (March).
#1. Dave Peterson - Several reviewers noticed problems with the latest
draft of the profile. Check the master pages for revision and spelling
in the headers. In addition there are many cases of missing letters.
These usually occur at the end of the line but sometimes occur in the
middle of a line as well. Fixed
#2. Dale Lafollette - Check on status of new TapeAlert flags. The
editor now has a direct contact. See the minutes.
#3. Stewart Wyatt - Bob Snively questioned how TapeAlert flags are
cleared. He felt the flag should not be cleared until the underlying
cause had been corrected. He felt the text was unclear. Stewart will
forward his concerns to Stephen Gold for comment. Completed - Bob is
still unhappy and will post his issues to the reflector for review by
the TapeAlert working group. See the new action items.
#4. Dave Peterson - Clean up TapeAlert description using appropriate
standards text. Completed
#5 Erich Oetting - Consider specifying the compression algorithm like
supported density is specified. Forwarded to next meeting.
#6 Stewart Wyatt - Post Fibre Channel tape connector specification to
the reflector before the next meeting. Completed
#7 Bob Snively - Provide an updated copy of the FCP-2 by March 19th.
#8 Dave Peterson - Provide an updated copy of FC-TAPE on or before
April 1. Post issues to the reflector. Completed
#9 Group - Review letter ballot comments and post disagreements to the
reflector. No postings were made.
#10 Stewart Wyatt - Post action items to the reflector two weeks
before the next meeting. Completed
#11 Dave Peterson - See item 5 in the minutes for changes to the data
compression related mode pages. Completed
#12 Dave Peterson - See item 7 in the minutes for changes to the SRR
text description. Completed
5. FC-TAPE Document Review: T11/99-069v2 and T11/99-092v2 Dave
Dale LaFollette led the group through a page by page review of the
latest draft of the proposal. The changes were mostly editorial in
nature and are listed below.
Page 2 and 3: Reference numbers incorrect
Page 4 3.1.6 Login_BB_Credit include FLOGI and FDISC
Page 4: 22.214.171.124 Title bold and cross reference errors in 3.1.12 and
Page 6 Removed second sentence after 3.3 heading
Page 22 SRR is now an FC-4 Link Service. So it needs to be in a table
of its own.
Page 23: Note 3 PRLO misspelled
Page 25: Take out space in note 1
Page 26: Replace Clause 7 on Timers with reference to FCP-2. (FCP-2
will not reference AL_TIME and R_T_TOV)
Page 30 First two paragraphs will to be turned into a table. Reference
to SCC should be SAM-2.
Page 32 Table 22 FCP_CONF_REQ
Page 40 Mode Select Page 02 should be an I
Page 41 Mode Sense Page 02 should be an I
Also Page 18 is duplicated in both entries
Page 45 Table 27 header format fixed
Page Note 4 is not referenced in Table 28
Page 49 Note 5 upper case errors on "One"
Table 28 headers - more "Requirement" heading up one line
Index will be taken out
Stewart Wyatt, HP, questioned the implementation requirements for
FCP_CONF. This was reviewed and the group decided to leave the text as
6. FCP-2: T10 Working Drafts T10/99-168r0 Bob Snively
This agenda item was deferred until Thursday morning (15th). Bob had
created a document describing the changes made to FCP-2 revision 1.
The minutes from this agenda item reflect this document, T10/99-168r1.
Bob intends to update the document (to revision 2) with the comments
and decisions made in this meeting and present it to the T10 plenary
for approval next month. These minutes will not attempt to reproduce
Bob's document. The reader is advised to refer to Bob's document while
reading these minutes.
Bob started with his review with clause 5 "Items not planned for
Clause 5.1 Definition of mode page parameters for FCP, DTDC field - No
action taken. (This item is the proposal Jim Coomes presented in
agenda item 9.1.)
Clause 5.2 Restrict data overlay to error recovery. Decision: This
restriction should only be in the tape profile, not in FCP-2.
Clause 5.3 Continuously increasing sequence count. Decision: Okay -
adequately covered in other documents.
Clause 5.4 Data checking when doing overlay or out of order transfer.
Decision: Bob is adding a warning about the risk of incomplete data
transfers when data overlay is allowed.
Clause 5.5 FCP_RSP IU length. Some early implementations issued "good"
responses that were 12 bytes in length by not sending invalid fields.
These implementations are noncompliant. Action: Bob will look for a
place to make the statement that a response will be a minimum of 24
bytes in length.
Clause 5.6 Automatic persistent reservation registration update:
Decision already installed.
Clause 5.7 Reference for ABTS. Decision: Included in FCP-2 Annex for
inclusion in a later standard. Action: Needs to be reviewed with
respect to FC-PH
Clause 5.8 FCP_CONF and linked commands. Decision: FCP_CONF and linked
commands are mutually exclusive except for the last command or a check
condition that ends the link. Action: Bob will check to see that this
is explicitly prohibited in the text.
Clause 5.9 Obsolete Third Party Process Logout. Decision: Cannot be
obsoleted. Is documented in FC-FS
Clause 5.10 Rules for ELS generation before Login. Action: Bob will
talk to Bob Kembel and get a list of the specific BLS and ELS which
are allowed before login. (Login can be implicit.) List in annex for
moving to FC-FS
Clause 5.11 SRR transfer of initiative. Decision: No change, current
approach is right.
Clause 5.12 Retry Offset. Decision: If the starting point is ambiguous
recovery starts at the beginning. Action: Participants requested to
review - Bob believes its ok.
Clause 5.13 Review Enable. Decision: No change (This is the same issue
as clause 5.1.)
Clause 5.14 Correction of DDIS bit. Discussion: Could also include
FDISC and FAN in public environment. Bob noted that this violates
FC-AL2. Jim Coomes noted that it also violates PLDA, only shows up in
a limited environment. Decision: No change Action: Jim Coomes to make
a proposal to limit certain of the mode page bits to private loop
Clause 5.15 Consideration of additional ELS. A request for documenting
ADISC, PDISC, FDISC (and FAN) in the FCP-2. Decision: Documented in
other places - referred to in FCP-2. Bob noted that Jeff Stai believes
that the public loop discovery process documented in FLA has some
problems. Discussed where the verification process should be
documented. Decided it should go FC-FS. Action: Dave Peterson to write
a proposal carry the protocol for discovery to FC-FS.
Clause 5.16 CRN recovery. Action: Bob requested participants to
Next Bob reviewed clause 3 "Technical changes to be approved."
Clause 3.1 Timer Definitions RR_TOV, REC_TOV and ULP_TOV. Decision:
Clause 3.2 REC_TOV timer implementation. Relaxing the restriction of a
REC_TOV timer per exchange. Decision: Change the proposal to Matt
Wakeley's suggestion to require waiting at least REC_TOV and removing
the requirement of one timer per exchange.
Clause 3.3 Negotiation of FCP recovery capabilities. This issue is how
to determine that a device supports FC-Tape recovery. Decision:
Accepted a proposal made by Dave Baldwin to use the FCP_CONF bit in
the PRLI to indicate support for FCP_CONF, REC, and SRR capabilities.
The bit will renamed to TBD. Vote: 10 approve, 2 against, 5 abstain.
Clause 3.4 Obsolete mixed command/data and data/response obsolete. Bob
noted that these capabilities are incompatible with tape error
recovery. Voted 19 approve, 0 opposed, 0 abstain.
Clause 3.5 Obsolete First Burst Length parameter definition. Decision:
Changed proposal to say: Applies only if outbound (write) operation
transfer and write XFER-RDY disabled otherwise the field is ignored.
Action: Bob Needs to review proposal for improper references to loop
and tenancies. Vote: 11 approve, 0 opposed, 9 abstaining.
Clause 3.6 Interpretation of FCP_RSP. This proposal clarifies the
text. Vote: 14 approve, 0 opposed, 1 abstaining. Discussion: Kurt
request that when validity = 0, implies that the content is
undetermined as a general rule be explicitly stated. Action: Bob Check
to be sure that "ignored" = 0 length.
Clause 3.7 SRR Reason Code incorrect. Action: Neil Wanamaker to
provide corrected values.
Clause 3.8 SRR to be rewritten. This proposal is to make SRR an FC-4
link data frame. This is the first time that an FC-4 link data frame
has been defined. Vote: 10 yes. 0 opposed, 4 abstain.
Clause 3.9 Make class independent recovery optional for FCP-2. Wordage
changed to make the recovery process not tape exclusive. Decision: OK.
Next Bob reviewed Clause 1.0 Technical Changes. Only two items were
Clause 1.4 RR_TOV Timer defined in mode page. Decision: Changed
defined values for RR_TOV units to 1, 3, 5. 0 is defined to mean that
there is no timer specified in this mode page, all other values are
reserved. Voted: 13 approve, 0 opposed, 1 abstaining.
Clause 1.5 Long CDB definition. Discussion about the position on
FCP_DL unclosed. Action: Review proposal with respect to the position
7. SSC: T10 Working Drafts Dave Peterson
The next draft is waiting for TapeAlert text that is to come from a
TapeAlert working group meeting held the previous day. Dave expects
that he should have this input by next Tuesday and be able to issue a
new draft. There was some concern expressed that the SSC not be held
up if the information from the TapeAlert group is not received
8. T11 New Business Group
8.1 FC-Tape Possible actions, 2LB or 1PR, Dale LaFollette
The group discussed the next action to take with the FC-TAPE proposal.
The options are to request another letter ballot review from T11 or to
forward for public review. The discussion revealed that forwarding for
public review essentially publishes the document. Dale called a vote.
The results were 3 votes for a second letter ballot and 9 votes for
public review. Action: Dale Lafollette to forward FC-TAPE for further
processing for public review.
8.2 Connector Proposal, Stewart Wyatt, Hewlett Packard: Stewart
presented an argument to change the pinout in his proposal for a 68
pin SCA-2 connector from the existing 40 pin connector defined in
SFF-8067. There was no support for the change. Action: Stewart will
modify his proposal to match SFF-8067.
8.3 Private only. Joe Breher, Exabyte. Requested support for private
loop only operation - supported by Tak Seto, Quantum, and Jim Coomes,
Seagate. Customers have requested this as a means of limiting access
to the drive. Dale Lafollette said that he has customers that don't
want the tape to be a public device either. Dal Alan, ENDL, thought
that a "private only" mode page bit might be needed. Joe's proposal
was unanimously defeated. While a venndor can implement a private loop
device, this behavior will be non-compliant to this profile.
9. T10 New Business: Group
9.1 FCP-2 Jim Coomes, Seagate, Mode page control of FCP IUs in a Loop
Tenancy. An unspecified initiator requires that the FCP_RSP be sent in
a different loop tenancy than the last FCP_DATA read frame. Need to
control whether the FCP_RSP can be sent in the same loop tenancy as
the last FCP_DATA frame. Bob Snively suggested changing the proposal 1
IU per tenancy. The group voted 2 to 14 against incorporating this in
the FCP-2. A suggestion was made to document this in a vendor unique
manner and to make that information available for others who need it.
Note that this issue is also discussed in the FCP review, agenda item
9.2 SSC Carl Zeitler, Compaq. Information field value after space
command early termination. Copying from Carl's overhead, A change was
noted in SSC from SCSI-2. It relates to the last sentence of the
second paragraph (on filemarks) on page 44: "A CHECK CONDITION caused
by an early termination of any SPACE command shall not result in a
negative information field value." Carl illustrated a case where a
negative value can be calculated. Action: He asked the tape vendors to
clarify how this is implemented. Carl will post an explanation of the
issue to the reflectors.
10. Review New Action Items: Stewart Wyatt
General Action Items
#1. Erich Oetting - Consider specifying the compression algorithm like
the supported density is specified.
#2. Bob Snively - Post concerns about TapeAlert to reflector.
#3. Dale LaFollette - Ask T11.3 plenary for public review of FC-TAPE.
#4. Stewart Wyatt - New connector proposal matching 40 pin SFF-8067
#5. Carl Zeitler - post question about negative information field
values on early termination of SPACE command to reflector
#6. Tape vendors to reply to Carl's issue.
FC-TAPE action items
#1. Dave Peterson, make the changes identified in these minutes under
agenda item 5.
FCP-2 action items.
#1. Bob Snively will look for a place to make the statement that a
response will be a minimum of 24 bytes in length.
#2. Bob Snively will review the ABTS changes in the FCP-2 draft with
respect to FC-PH.
#3 Bob Snively will check the text to see that it states that FCP_CONF
and linked commands are mutually exclusive except for the last command
or a check condition that ends the link.
#4 Bob Snively and Bob Kembel and list the specific BLS and ELS which
are allowed before login including explicit Login in an annex for
later placement in FC-FS.
#5 Bob Snively requested that the error retry offset text (Clause 5.12
in T10/99-168r1) should be reviewed by the participants.
#6 Jim Coomes will make a proposal to limit certain of the mode page
bits to private loop applications.
#7 Dave Peterson to write a proposal to carry the protocol for
discovery to FC-FS.
#8 Bob Snively requested participants to review the CRN recovery
(T10/99-168r1 Clause 5.16).
#9 Bob Snively needs to review the proposal for first burst length
#10 Bob Snively check to be sure that "ignored" = 0 length in FCP_RSP
and other relevant places.
#11 Group review Long CDB definition proposal with respect to the
position of FCP-DL.
11. Adjournment: Group
Note: The minutes posted to the web site include an attendance list.
The list could not be pasted into this text email format.
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at symbios.com
More information about the T10