Minutes: IETF STORM WG meeting 3/25/10 9:00 AM - 10:00 AM Co-Chairs: - Tom Talpey (Microsoft) – attending via Webex - David Black (EMC) David presented the "Note Well". Craig Carlson agreed to take meeting notes. Agenda bashing – Agenda had been bashed the previous evening to add Unicode/newprep item. No additional changes requested David presented charter Milestones x All major drafts are posted, except for iSCSI MIB update. Waiting for WG decisions on what to take out of iSCSI, and for iFCP MIB update (same author) to clear MIB Doctor review x Two iFCP-related drafts are through WG Last Call, publication will be requested soon. x Near term WG focus: iSER draft, MPA draft and iSCSI feature update draft. iSCSI Update Draft Status – Fred Knight (NetApp) x SCSI version descriptors for current revisions of SAM-5 and SPC-4 are now available on T10 web site. These will be cited by the update draft in the form of "this revision or subsequent". x Key Name is an open issue from mail list o May need a numeric value that is increasing so we can do a direct comparison instead of having to invent new keys for every update o David Black – proposes PDUFormatLevel x Numeric key where: 1 = existing iSCSI, 2 = features in update draft x If key is not negotiated, 1 is default value x This is an IETF key with values assigned by IETF – This value is independent of SAM version number o No objection to David's proposal. Action: David to send proposal with all details to list RDMA-related drafts – Tom Talpey x iSER draft published – some comments made on list o David believes there is an open issue around fields that IB needs (but iWARP may not need in all cases). Need to review list comments. x MPA Peer Connect draft – published late Feb. iSCSI, Unicode and newprep – David Black x History – iSCSI IQN names use reversed Domain Names – uses stringprep to normalize Unicode input – stringprep RFC is based on Unicode 3.2 – Unicode is now at 5.2, so any characters added since Unicode 3.2 are not allowed by stringprep x newprep BOF held this week to discuss how to move beyond stringprep – Conclusion of BOF – "We need a working group, but are not sure exactly what it should do." – Problem requires specialized Unicode expertise, a WG is a good way to focus and apply that expertise. x One of the biggest concerns for iSCSI is that conversion to lower case is not workable for all languages. IQNs are always lower case. x Four characters cause backwards compatibility problems – German ezsett, Greak final sigma, and two zero-width characters used for languages such as Farsi (Persian). New iSCSI draft – Mallikarjun Chadalapaka (HP) x Draft revised throughout to allow combined Initiator/Target implementations. - Section will be added to summarize changes from consolidated RFCs, starting with this change. s Decision made to include UML diagrams in consolidated draft instead of update draft (will appear in next version of consolidated draft). x Fred Knight: Is this draft intended to obsolete old drafts? – In consolidated draft, statement made that this supersedes previous drafts x David Black: iSCSI features to take out, start with all authentication mechanisms except CHAP and SRP. Will discuss further on list – David to bring to list. Action Items: x David Black to send PDUFormatLevel proposal to mailing list for discussion. x David Black to bring list of features to remove from iSCSI to mailing list for discussion.