|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] RE: iSCSI - change proposal LUN field definition on every PDUI agree with all the folks who oppose this change. The approach should be to question usefulness of each and every field, not add more. > -----Original Message----- > From: owner-ips@ece.cmu.edu [mailto:owner-ips@ece.cmu.edu]On Behalf Of > Robert Snively > Sent: Tuesday, November 13, 2001 8:23 AM > To: 'Julian Satran'; ips@ece.cmu.edu > Subject: RE: iSCSI - change proposal LUN field definition on every PDU > > > Folks, > > I have followed this thread up to the current time, and I have > to agree with the people who think that this is probably > not necessary and has undesirable consequences. > > One reason that it is not necessary is that as we write, the > SCSI MIB people are defining the properties and statistic > gathering capabilities of SCSI logical units. These capabilities > are clearly going to require logical units to capture additional > performance information using SCSI logging commands (which > already capture a huge amount of error information). > > A second reason that it is not necessary is that the command > and data counts can be extracted from the command PDU's > CDB field, which is labeled with a logical unit. Given that > the command completes correctly, the counting is already done > and the count is in the command PDU. > > A third reason that it is not necessary is that any decent > instrumentation on the line has to already be capable of sorting > through the TCP/IP mapping, the security mappings, and the > basic PDU identification. All the monitors that we have today > in the SCSI environment apply a little simple postprocessing to > the data they have captured and presto, they not only identify > PDU's related to particular logical units, but to particular > commands, and they have interpreted the commands and verified > that the responses are appropriate to the commands. > > If the instrumentation is inline with the receiving code, > it is by definition aware of the states and identifications > already available to it and does not need this additional information. > > As for undesirable consequences: > > Paul Koning's objection to the additional functional steps > required to provide the LUN is a valid concern. > > Mallikarjun Chadalapaka's concern about the requirement for > additional consistency checking (and a whole new class of > non-SCSI protocol error checking) is also a valid issue. > > Let's talk to your colleague and explain to him/her how there > are better ways to achieve his/her goals. > > Bob Snively e-mail: rsnively@brocade.com > Brocade Communications Systems phone: 408 487 8135 > 1745 Technology Drive > San Jose, CA 95110 > > > > > -----Original Message----- > > From: Julian Satran [mailto:Julian_Satran@il.ibm.com] > > Sent: Monday, November 12, 2001 4:31 AM > > To: ips@ece.cmu.edu > > Subject: iSCSI - change proposal LUN field definition on every PDU > > Importance: High > > > > > > Dear colleagues, > > > > A colleague interested in instrumentation approached me with > > a question > > about stateless logging of specific LU activity. > > With the current iSCSI PDU formats this is not possible. > > We have consistently avoided having fields that are redundant > > and will > > need consistency checking. > > However I think we should consider including the LU field in > > all PDUs that > > are referencing a specific LU to simplify this type of > > instrumentation (as > > we did with the direction bit in the op-code). > > > > As I am already in "count-down" mode for version 09 I would > > appreciate > > your comments ASAP. > > > > Julo > > >
Home Last updated: Tue Nov 13 16:17:36 2001 7788 messages in chronological order |