Re: iSCSI - do we need anything in the mode pages?

[Santosh Rao <santoshr@cup.hp.com>]

Julian,

I fully support this move. 

The "Enable CRN" bit should be a SCSI ULP feature. The EPDC bit of FCP-2 
is a layering mess-up in FCP-2 since it is not clear if
they are testing for CRN support in their target FCP layer's FCP_CMD IU,
or whether they are testing their device server or both.

A new bit can be introduced in the "Control Mode Page" which is bag of
bits that verify SCSI ULP features currently. This "Enable CRN" bit should
purely test the device server since this is a ULP peer to peer check. In
addition, the FCP-2 folks would check the EPDC bit to test their target
SCSI LLP's capability to handle the CRN in its FCP_CMD IU.

Perhaps, this can be one of the iscsi requests for change in T10, along
with the rest of the changes that iscsi is driving. iSCSI would then
attach no special semantics to any of its mode pages.

Thanks,
Santosh

> 
> Dear colleagues:
> 
> We have now only two scraps of information in the iSCSI mode page:
> 
>    EMDP
>    CRN
> 
> 
> I suggest we take EMDP out (for simplicity) and reintroduce
> DataSequenceInOrder
> 
> Unfortunately we can't do anything about CRN - but since it is not strictly
> an iSCSI issue
> we can just as for it to be placed honorably in commonly used per-LU page
> by T10.
> 
> Please comment.
> 
> Julo
> 
> 


-- 
##################################
Santosh Rao
Software Design Engineer,
HP-UX iSCSI Driver Team,
Hewlett Packard, Cupertino.
email : santoshr@cup.hp.com
Phone : 408-447-3751
##################################