question about reservation and persistent reservation

Knight, Frederick Frederick.Knight at netapp.com
Tue Jan 2 08:54:43 PST 2007


* From the T10 Reflector (t10 at t10.org), posted by:
* "Knight, Frederick" <Frederick.Knight at netapp.com>
*
> In the process of defining persistent reservations is was not clear in
> the standards as to what the reservations were associated with. At the
> same time the SCSI architecture was moving from a parallel based 
> architecture to a serial based architecture. That change required a 
> clear delineation between ports and devices that was not previously 
> necessary. The end result is that the I_T nexus became a port to port 
> definition and that made all reservations port to port.
> 
> I would suggest the only reasonable solution is the use of persistent 
> reservations in environments that contain multi-ported SCSI devices.
but if all reservations (even persistent reservation) port to port, then
this issue is still not solved in a reasonable way, why a target should
reject a request from _same_ initiator who holds reservation but
different port?
>> This is exactly what you want.  Each host O/W must control all the
paths
>> from the host to the device.  There may be times when the host needs
to
>> stop I/O on only one path.  The way the host does that is via a
PREEMPT
>> PR function that impacts only that one path.
>>
>> To do that, each path must register with a unique key.  When the same
key
>> is used, the PREEMPT will impact all paths using that key.  So the
spec
>> allows the host to have whatever level of control it wants (such as a
>> common key for many paths, or a unique key for each path).
>>
even the reservation key is to identified the _only_ nexus and can not
be borrowed by another nexus between _same_ initiator and target right?
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org



More information about the T10 mailing list