Clarification on ENCSERV bit in the standard INQUIRY

[Ralph O. Weber]

* From the T10 Reflector (t10 at t10.org), posted by:
* "Ralph O. Weber" <roweber at ieee.org>
*
Kevin,

I cannot speak to question number one.

Regarding question number two... Once upon a time there was just
one LUN in a SCSI target device. Even today, many folks see the
world in this context. In such cases, logical unit and target
device can be (and are) viewed as synonymous.

Unfortunately, I do not view changing 'SCSI target device'
to 'logical unit' as an editorial change, owing to lingering
sensitivities on the subject and the numerous opportunities
to get it wrong.

I need some kind of specific direction from T10 before making
any changes.

All the best,

.Ralph

Kevin_Marks at Dell.com wrote:

> All,
>
> I recall several months ago at a CAP meeting, a discussion of the=20
> ENCSERV, MCHNGER and SCCS bits in the standard INQUIRY and when they=20
> are set or not set. To my recollection from the discussion, these =
bits=20
> are used to indicate that a device (LUN, further question) supports=20
> these command sets in addition to the command set generally =
associated=20
> with the peripheral device type.
>
> So, for example if the block device with a peripheral device type of=20
> 00h (SBC-2 - Direct access block device) contained an attached=20
> enclosure services process (one of the SES models) that it would have =

> the ENCSERV bit set to one in the standard INQUIRY to indicate that =
it=20
> contained an embedded enclosure services processor. For a standalone=20
> enclosure services processor, such as one that would be contained in =
a=20
> JBOD as an accessible LUN, the ENCSERV bit would be set to zero,=20
> because the peripheral device type is already set to 0Dh (SES -=20
> Enclosure services device.)
>
> So is my recollection correct. If so, then I have to additional =
questions:
>
> 1. SES-2r12 currently states that for both the attached enclosure=20
> services process model and the standalone enclosure services process=20
> model that the ENCSERV shall be set to one. Is this correct?
>
> 2. SPC-3 and SPC-4r02 for the ENCSERV bit say =93An Enclosure =
Services=20
> (ENCSERV) bit set to one indicates that the SCSI target device=20
> contains an embedded enclosure services component. See SES for =
details=20
> about enclosure services, including a device model for an embedded=20
> enclosure services device. An ENCSERV bit set to zero indicates that=20
> the SCSI target device does not contain an embedded enclosure =
services=20
> component.=94 Should the SCSI target device actually be Logical Unit=20
> instead? This question actually applies to several of these bits.
>
> Thanks
>
> Kevin
>

*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org