My basic (read simple) understanding of the SCSI Pass Through Interface (SPTI), is that all SCSI commands issued at the initiator will be transferred to the target device?
The corollary of this, I assume, is that any response from the target device should be returned to the initiator.
In addition, I would expect that this conversation through the SPTI should be transparent?
In the following example, I'm not sure that this is the case.
This screen shot shows three IO traces. The far right is the reference taken with all resources on the same machine. At count 12, after being told to do something, the Test Unit Ready enquiry is returned the SRBStatus = 0Ah.
However, the same operation carried out over the iSCSI nexus, gives an alternative SRBStatus to the Test Unit Ready SRB?
I don't think I should be worried by the "unusual" SRBStat in the middle trace (the iSCSI Target). This is because there is no SRBStatus in a SCSI_PASS_THROUGH_IOCTL data structure. See: http://msdn.microsoft.com/en-us/library/ms810301.aspx
So, my question is this: As the SCSI_PASS_THROUGH_IOCTL (in the middle trace) is "carrying" an SRB (with a SRBStat=0Ah) does it arrive at the initiator in an SCSI Request Block with the SRBStat= 84h?
Regards,
Martin
