SORT BY:

LIST ORDER
THREAD
AUTHOR
SUBJECT


SEARCH

IPS HOME


    [Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

    Re: draft-ietf-ips-scsi-mib-04 comments



    Hi Robert,
    
    	This is regarding the technical comment given by you with respect to devices having 
    multiple ports. In this case my opinion is that it is possible to know the target port identifier
    ( relative port identifier according to SPC-2 ) to access the Logical Unit.
    
    Please refer  Table 174 and Table 175 under the section section 8.4.4 Device identification page.
    Since, if the ASSOCIATION value is 1h, the IDENTIFIER value contains a four-byte binary number
    identifying the port relative to other ports in the device according to the following (Table 175) values.
                    ^^^^
    
    
    Table 175  Relative port identifier values
    --------------------------------------------
    0h                            Reserved
    1h                            Relative port 1, historically known as port A
    2h                            Relative port 2, historically known as port B
    3h-7FFFFFFFh                  Relative port 3 through 2 147 483 647
    80000000h-FFFFFFFFh           Reserved
    
     In this case, the CODE SET field shall be set to 1h and the IDENTIFIER LENGTH field shall be set to 4. If
    the ASSOCIATION value is not 1h, use of this identifier type is reserved.
    
     So we can keep the value 1 (hex) for scsiLuIdAssociation field in the population example for 
    parallel scsi.
    
    
    Thanks,
    -Narsimha
    
    
    "Elliott, Robert (Server Storage)" wrote:
    > 
    > Comments on draft-ietf-ips-scsi-mib-04:
    > 
    > Technical:
    > =========
    > COMMENT: For a logical unit in a multiple ported device, the ScsiLuIdentifier cannot return a scsiLuIdAssociation of type "port" (01 binary) because the target port being used to access the logical unit is not known.  (The MIB is probably being retrieved through a non-SCSI channel... which target port does the logical unit pick?)
    > 
    > For this version of the MIB, I'd just require that scsiLuIdAssociation be set to 0 (or possibly 2) and figure out how to report target-port specific information in the future.
    > 
    > This occurs in the ScsiIdAssociation textual-convention and in the parallel SCSI example:
    > "   ScsiIdAssociation ::= TEXTUAL-CONVENTION
    >      STATUS    current
    >      DESCRIPTION
    >           "This Textual Convention specifies what the identifier is
    >           associated with (e.g, with the addressed physical/logical
    >           device or with a particular port) for the identifier contained
    >           in an Identification Descriptor returned in a Logical Unit's
    >           Device Identification Page, and is formatted as defined in T10
    >           SPC-2 (see REFERENCE) table 178 û Association."  "
    > and:
    > "   11.14 scsiLuIdentifer:
    > 
    >       Attribute                       Value
    >       ----------                      ------
    >       scsiInstIndex                   1
    >       scsiDeviceIndex                 2
    >       scsiLuIndex                     1
    >       scsiLuIdIndex                   1
    >       scsiLuIdCodeSet *1              2
    >       scsiLuIdAssociation *2          1
    >       scsiLuIdType *3                 1
    >       scsiLuIdValue                   IETFsl318203-0004
    > 
    >      *1 - The identifier field will have ASCII graphic codes.
    >      *2 - The identifier is associated with the port that received the
    >      request. "
    > 
    > Change the example value to 0 and remove note 2.
    > 
    > Also note that SPC-3 is changing 0 to mean "logical unit" rather than
    > "physical/logical device" and 1 to mean "target port".  It is also
    > adding 2 as "target device" (which could be reported through this
    > MIB).
    > 
    
    -- 
    
    Please find us at http://san.hcltech.com/
    and http://cdn.hcltech.com/
    


Home

Last updated: Tue Dec 10 12:18:57 2002
12073 messages in chronological order