Questions on Persistent Reservations...retry
Gerry_Houlder at notes.seagate.com
Mon Aug 11 09:37:44 PDT 1997
* From the T10 (formerly SCSI) Reflector (t10 at symbios.com), posted by:
* Gerry Houlder <Gerry_Houlder at notes.seagate.com>
>1st... What about persistent reservations to LUN0 if it is the controller
>device? Any thoughts on practicality of such a thing and any restrictions that
>might want to be placed on it (by comment or spec treatment?)...in general
>the reservations are for media access, but hosts ought to know what the right
>thing to do with the controller device...
I don't think LUN 0 of a controller device should act any differently than LUN 0
of a disk device with respect to reservations. Reservations (even persistent
reservations) are used for reasons other than reserving data access rights.
>2nd... How persistent to reservations need to be with respect to power-off
>situations. The definition of this case could be power-failure and planned
>power shutdowns. I can see where the reserations should be kept across
>power-failures, but I am not sure it applies when someone decides to turn
>their computer room for the weekend to save energy...I suppose it could
>be the fact that no other access is allowed providing some type security,
>but its not much...
Looking at this from the device's point of view, EVERY power off is a
"power failure". If the system doesn't want a persistent reservation to last
through a scheduled power off cycle, then the system should issue a
release command before doing the power off. Even Windows 95 has
an "orderly power off" procedure built in now to allow for things like
Persistent Reserve was designed to last through power failure so that
an inadvertent (or even intentional) power cycle couldn't be used to
get access to something that is supposed to remain inaccessible
unless you know the magic password....
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at symbios.com
More information about the T10