Ross wrote: > Hi John, > > If it's in 103, does that mean it's likely to be included in OpenSolaris > 2009.04 when that's released next year? > Likely, yes. > And am I right in thinking that with COMSTAR we actually can set IQN's now? > It sounds like we now have that plus aliases which to my (rather > inexperienced) eye sounds like a big step forward. > Yes, IQNs can be set. > With that in place, could I ask if there is anything show stopping with my > other idea - adding a flag to the iSCSI initiator asking it to use the IQN or > Alias as the basis for the device name? Would it be worth me filing a RFE > for this? > > > No, I don't believe it is. While the target name, in this case the iqn, could possibly be used for a portion of a device path, I still want to point out that it is not the device identifier nor should it be used as such. It is only a target port identifier. But what you're suggesting, if I'm understanding you correctly, just isn't possible in the current /dev/[r]dsk/cNtNdN namespace. And since the iqn (or the alias) is representing a target port, this basis for naming would not be practical for a multipathed device as you could be looking at multiple iqns/aliases being representative of a single logical unit.
- John _______________________________________________ storage-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/storage-discuss
