|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] RE: mailing listNo - I did not see it earlier - probably wwhen I was taken of the list... comments in text. Thanks, Julo "Ayman Ghanem" <aghanem@cisco.com> on 30-06-2001 23:45:13 Please respond to "Ayman Ghanem" <aghanem@cisco.com> To: Julian Satran/Haifa/IBM@IBMIL cc: Subject: RE: mailing list Julian, I sent this to you last week. Not sure if you got it. Also, in the Task Management Response, I believe the MSB of the reserved field (bit 24) should be set to 1 like other PDUs going from the target side. +++ fixed +++ -Ayman ---------------------------------------------------------------------------- ------ Julian, I have the following comments on draft-6.90: 1- I am not sure if you saw this posting, but it is not clarified in 6.90 http://www.pdl.cmu.edu/mailinglists/ips/mail/msg05089.html 2- I would prefer, for this draft and any future drafts, that any new response codes added to be assigned new values and not values assigned earlier to still-existing return codes. This will help backward compatibility with earlier drafts. For example, Task Mgt Rsp, Reject, and logout have responses which existed in earlier drafts but are now given new codes. ++OK (for the future and if regularity does not require otherwise) +++ 3- Section 7.1, there is no reject response code for "Unsupported Replay". +++ yes there is - code 7 - i'll make wording clearer +++ 4- In sec. 4.1, version number 0x'01' should be 0x'02' +++ fixed ++++ 5- Sec. 2.16.5, no Additional Runs field in the SNACK PDU +++ fixed +++ 6- Bit-7 of byte 1 of the logout command should be set to 0, or as an F-bit to be consistent with text of section 2.2.2.4 +++ Logout is always single and last. That is why I set the F bit to 1 +++ 7- In section 2.16.1, "targets MUST support Status SNACK". However, in section 1.2.2.2, "To enable command recovery the target MAY maintain enough state information to enable data and status recovery after a connection failure". These seem to be inconsistent. Also, I could not see the scenario if the target can not support a Status SNACK request. If dropping the connection is the option, then I think it should be clarified. +++ the current text reads: An iSCSI target that does not support recovery within connection MAY discard status SNACK. If the target supports command recovery within session it MAY discard the SNACK after which it MUST issue an Asynchronous Message PDU with an iSCSI event indicating "Request Logout". ++++ -Ayman > -----Original Message----- > From: owner-ips@ece.cmu.edu [mailto:owner-ips@ece.cmu.edu]On Behalf Of > julian_satran@il.ibm.com > Sent: Wednesday, June 27, 2001 8:29 AM > To: ips@ece.cmu.edu > Cc: bassoon@YOGI.PDL.CMU.EDU > Subject: mailing list > > > > > Dear colleagues, > > I do not receive mail from the list (since mid last week). > Please address mail that you think should reach me adding me explicitly > until the issue is fixed. > > Regards, > Julo > > >
Home Last updated: Tue Sep 04 01:04:22 2001 6315 messages in chronological order |