|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] RE: iSCSI AutosenseSomesh, I have on objection on making Autosense mandatory and removing the bit from iSCSI. My issue (if it is an issue) is SCSI related. How are asynchronous events that are not error related handled without Contingent Allegiance - and what effect will it have on iSCSI. The case that was raised a while ago (by Costa I think) was that of a volume change - with the message reporting it "crossing" a write on the wire or in the OS. With CA the state synchronization can be easily enforced (command are rejected until a sense is received). Without it - even with autosense - state synchronization in the OS becomes tricky. Perhaps somebody on the list active also in T10 can give us a hint of how this type of sequence will be handled without CA. I understand that CA can be kept even with autosense and we can remove the bit but what if T10 decides to keep both? Regards, Julo somesh_gupta@hp.com on 29/08/2000 20:17:02 Please respond to somesh_gupta@hp.com To: ENDL_TX@computer.org, ips@ece.cmu.edu cc: (bcc: Julian Satran/Haifa/IBM) Subject: RE: iSCSI Autosense If we are looking for consensus on this issue, I agree with Ralph. Better to clean up now than implement something that is waste of energy. Somesh > -----Original Message----- > From: ralphoweber@compuserve.com [mailto:ralphoweber@compuserve.com] > Sent: Monday, August 28, 2000 8:03 PM > To: ips@ece.cmu.edu > Subject: FW: iSCSI Autosense > > > Note: Clause numbers reference draft-satran-iscsi-01.txt. > > There is nothing in the SCSI Architecture that requires the A > Flags bit (byte 2 bit 7) described in 3.2 and 3.2.1. Specifically, > the SCSI Architecture does not require a transport to have the > ability to disable Autosense. > > Contingent Allegiance (the target state that follows a CHECK > CONDITION in the absence of Autosense) is not a required SCSI > behavior. In fact, Contingent Allegiance is an out dated > SCSI behavior that continues to be documented in SAM for > backwards compatibility. Contingent Allegiance is wholly > superseded by the Autosense capability. > > Autosense is dramatically simpler to implement provided the > SCSI transport protocol allows for delivery of the sense data > in the same packet as the CHECK CONDITION status (as is the > case in iSCSI). > > IMHO iSCSI would be well served by following the precedent > set by FCP. Remove the A bit and state explicitly that all > iSCSI devices are required to support Autosense and that > all instances of a CHECK CONDITION status shall be accompanied > by Autosense data. > > Note: These changes will also affect 3.3.7, 5.3 and possibly > other clauses of the internet draft. > > Thanks. > > Ralph... > > >
Home Last updated: Tue Sep 04 01:07:38 2001 6315 messages in chronological order |