Re: iscsi : DataPDULength can differ in each direction.

["Julian Satran" <Julian_Satran@il.ibm.com>]

Dave,

Implementation specific? I assume you wanted to say connection specific.

No the parameters you quote (except MaxOutstandingR2T) are not connection
specific.
They reflect multiplexing capability that the transport allows for SCSI.
Making them connection specific will
only complicate things with no added benefit.
MaxOutstandingR2T might be viewed as a way to compensate for RTT bu depend
also on handling capabilities
of the end-points.

Julo


                                                                                                     
                    Dave Sheehy                                                                      
                    <dbs@acropora.rose.ag       To:     ips@ece.cmu.edu (IETF IP SAN Reflector)      
                    ilent.com>                  cc:                                                  
                    Sent by:                    Subject:     Re: iscsi : DataPDULength can differ in 
                    owner-ips@ece.cmu.edu        each direction.                                     
                                                                                                     
                                                                                                     
                    05-10-01 22:44                                                                   
                    Please respond to                                                                
                    Dave Sheehy                                                                      
                                                                                                     
                                                                                                     




> Having now accepted that DataPDULength is a "maximum receive pdu length"
> limitation that each end point is allowed to express, this key needs to
> be on a per connection basis, and not a per session basis.
>
> The reason for this being that limitations on DataPDULength (more aptly,
> max receive pdu length) stem from implementation specific quirks. (One
> of the reasons for allowing this feature, in the first place.).

Following this same line of reasoning the keys:

           MaxBurstSize
           FirstBurstSize
           MaxOutstandingR2T
           DataPDUInOrder
           DataSequenceInOrder
           ErrorRecoveryLevel

are all potentially implementation specific and should not be categorized
as LO.

Dave