|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: iSCSI : DataRN issues.I originally made this request on 11/7/00 and therefore agree that DataRN should be removed. I especially take exception to the statement on page 87, refering to "18 DataNumber" that "An initiator MUST support data numbering if requested". An initiator should not be required to support a feature that might help recover an error faster. -Matt Wakeley Agilent Technologies Santosh Rao wrote: > > Hi Julian, > > Some further comments on this. > > Thanks, > Santosh > > julian_satran@il.ibm.com wrote: > > > Section 1.2.2.3. Data PDU Numbering > > ============================ > > What is the scope of DataRNs ? Is this maintained per task or per > > connection or per session ? This should be explicitly stated in this > > section as is done in earlier sections for CmdRN and StatRN. > > [js] per command and it says so [/js] > > I believe that the DataRN feature and all its related overheads add a level of > complexity that is going to discourage both targets and initiators from using > this. I am yet to see anyone state sufficient benefits of this feature that > can justify this complexity. Moreover, the benefit it may provide [if any, and > if somebody does use it] is in the error path context and its overheads hurt > the performance path. > > Therefore, in the interests of keeping iSCSI simple, I would like to request > removal of all references to DataRN and DataNumber. > > Thanks, > Santosh
Home Last updated: Tue Sep 04 01:05:55 2001 6315 messages in chronological order |