|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] RE: iSCSI : Command Ordering Proposal.Santosh, Rather than being clever with a special case setting the sequence number to zero, the limits imposed on the sequence for things like Head of Queue or NOP could be ignored and not be treated as violations. This would allow all such activity between the initiator and the target to remain in sequence. Allowing this exception to a basic scheme of ensuring sequential communication has several drawbacks such as lack of acknowledgement. What drawback is there in having the target making allowances for some commands such as Head of Queue or NOP to exceed a maximum? Doug > Douglas Otis wrote: > > > Santosh, > > > > If there is a task attribute, the complexity is when you allow > CmdSN = 0; > > > > Perhaps CmdSN = 0 should not be allowed or treated as just > another in the > > sequence. > > The (CmdSN==0) does have its uses. One example is the use of NOP-OUT for > a "connection ping." > > > > If a retry flag is set, the CmdSN below the > > ExpCmdSN should be viewed as an exception much as Head of Queue > should be an > > exception for MaxCmdSN. Both of these cases imply exceptional > behavior by > > the initiator in response to a problem. > > Accepting a command outside the (ExpCmdSN, MaxCmdSN) window with the retry > bit set then opens up windows for other types of stale and > duplicate commands > to be > processed. > > Here's the deal (as I view it) : > > Commands are sent with the "retry" bit under 2 conditions : > - Connection Failures > - Digest Errors > > In both these cases, an explicit handshake is required regarding the > (ExpCmdSN, MaxCmdSN) window, prior to starting to send "retry" > commands. This ensures that genuine "retry" commands do not slip > behind the (ExpCmdSN, MaxCmdSN) window. > > In the case of a connection failure, the Logout Response provides this > handshake. For digest errors, there is no such hand-shake. > Hence, a command that encountered a digest error and was retried could > likely slip behind the (ExpCmdSN, MaxCmdSN) window and then be never > processed. > > Regards, > Santosh >
Home Last updated: Tue Sep 04 01:05:38 2001 6315 messages in chronological order |