RE: iSCSI: NOPs on discovery session

["Ayman Ghanem" <aghanem@cisco.com>]

That is true. However, this is an implementation issue. My point is there is
nothing in the spec that requires either side to drop the discovery session
after the initial discovery is completed. So, the question is: if this
session can stay active long enough, should this command be disallowed?. The
key "SendTargets" is also mandatory and there is no restriction on when/how
it can be issued.

-Ayman


> -----Original Message-----
> From: owner-ips@ece.cmu.edu [mailto:owner-ips@ece.cmu.edu]On Behalf Of
> Mark S. Edwards
> Sent: Thursday, December 06, 2001 9:31 AM
> To: ips@ece.cmu.edu
> Subject: RE: iSCSI: NOPs on discovery session
>
>
> At 09:08 AM 12/6/2001 -0600, Ayman Ghanem wrote:
> >If the initiator choose to keep the discovery session up, the
> target could
> >use it to inform the initiator of new targets or LUNs it can access.
> >
> >-Ayman
>
>
> SLP and iSNS will already do this and since SLP is mandatory to
> implement,
> this feature already exists.
>
> Discovery sessions should be short lived, they tie up network
> resources.  Can you imagine 1000 initiators with long lived discovery
> sessions on a target ?  Any sensible target will probably
> implement a timer
> on a discovery session anyway.
>
>
> Mark.
>
>