[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Google
  Web www.spinics.net

Re: TargetAlias in SendTarget response



I must not have made myself clear. The expectation would be that the target would only send the TargetAlias if a key was sent first by the Intiator. For example Sun's initiator could send:

X-com.sun.sendaliases=true

during the login phase when the SessionType=Discovery. The Solaris target would then add the TargetAlias information to a SendTargets=All response only if that value were true. This wouldn't cause any other initiators or targets to break since they (initiators) either will not send the key or understand the key (targets).

If other folks feel that such a feature is worth prompting then I'll write up a proposal to create a new X#sendaliasinfo=true type key.

On May 31, 2006, at 9:52 AM, Jacob_Cherian@Dell.com wrote:

If the purpose is to find the alias of a target in a discovery session,
wouldn't it be better to define a new text key SendTargetAlias="Target
Name". This would not break existing implementations (either initiator
or target).

Legacy targets (those that do not support this key) would just reject
the request. Legacy initiators would never issue it.

Jacob Cherian



-----Original Message-----
From: Rick McNeal [mailto:Rick.McNeal@Sun.COM]
Sent: Wednesday, May 31, 2006 8:18 AM
To: Eddy Quicksall
Cc: ips@ietf.org; David Weibel; Sandars,Ken
Subject: Re:  TargetAlias in SendTarget response

During a discovery session the TargetName is not sent because it's
not known. The thought was that in the text response for the
SendTargets=All the target could return:

     TargetName=iqn.xxxx
     TargetAlias=fubar
     TargetAddress=yyyy (1 or more)

To do this without changing the specification means adding a new key
which is sent by the initiator.

On May 31, 2006, at 7:08 AM, Eddy Quicksall wrote:

Weren't you looking for a way to display the target alias of the
target that was used for discovery? If so then TargetName would not
need to be sent.

Or are you saying that for every target reported there would be a
TargetAlias?

Eddy

----- Original Message ----- From: "Rick McNeal" <Rick.McNeal@Sun.COM>
To: "Eddy Quicksall" <eddy_quicksall_iVivity_iSCSI@Comcast.net>
Cc: <ips@ietf.org>; "David Weibel" <David.Weibel@Sun.COM>;
"Sandars,Ken" <ken_sandars@adaptec.com>
Sent: Wednesday, May 31, 2006 8:40 AM
Subject: Re:  TargetAlias in SendTarget response


Since the TargetName is not required during the discovery session
how would a target know which TargetAlias to return? Not to
mention that during a discovery session there could be many
targets which are discovered.

It seems like using having an extra key sent by the initiator
during  the discovery session would be the simplest method.

On May 30, 2006, at 5:31 PM, Eddy Quicksall wrote:

TargetAlias would be allowed in a discovery session if it is not
in response to the SendTargets key.

Would that be OK? Then there would be no need to change the spec.

Eddy

----- Original Message ----- From: "Sandars, Ken"
<ken_sandars@adaptec.com>
To: "David Weibel" <David.Weibel@Sun.COM>; <ips@ietf.org>
Sent: Thursday, May 25, 2006 10:46 PM
Subject: RE:  TargetAlias in SendTarget response


Hi David,

It's certainly a nice idea, and I cannot remember why the
TargetAlias
was excluded from the SendTargets response. Since it would be
optional
the target can decide if there are any security issues or FUD with
sending it or not.

The benefit is it avoids the rather clumsy sequence of doing a
(discovery) login and logout to each target found in the original
unnamed discovery session just to provoke the target into sending
the
TargetAlias as part of the login phase.

However, it cannot just be added now without potentially breaking
some
initiators which are not expecting it, or are actively ensuring
it is
not sent.

Perhaps a new IO,LO,Declarative key could be introduced for use
by the
initiator to indicate its willingness to accept TargetAlias as
part of
the SendTarget's response?

Cheers
Ken

-----Original Message-----
From: David Weibel [mailto:David.Weibel@Sun.COM]
Sent: 26 May 2006 08:11
To: ips@ietf.org
Subject:  TargetAlias in SendTarget response


I was wondering what people would think about adding the
allowed and suggested use of TargetAlias during a SendTargets
response.  This would be an optional component of the response.
The goal would be to send the alias earlier in the discovery
process so the initiator software could also present it earlier
to the user.

This would require the ammendment of Appendix D. "No text
keys other than TargetName and TargetAddress are permitted
within a SendTargets response."  Along with some additional
text.

_______________________________________________
Ips mailing list
Ips@ietf.org
https://www1.ietf.org/mailman/listinfo/ips


_______________________________________________
Ips mailing list
Ips@ietf.org
https://www1.ietf.org/mailman/listinfo/ips


_______________________________________________
Ips mailing list
Ips@ietf.org
https://www1.ietf.org/mailman/listinfo/ips

----
Rick McNeal

A good friend will come and bail you out of jail...but, a true
friend will be sitting next to you saying, "Damn...that was fun! "




_______________________________________________
Ips mailing list
Ips@ietf.org
https://www1.ietf.org/mailman/listinfo/ips


----
Rick McNeal

A good friend will come and bail you out of jail...but, a true friend
will be sitting next to you saying, "Damn...that was fun! "




_______________________________________________
Ips mailing list
Ips@ietf.org
https://www1.ietf.org/mailman/listinfo/ips

_______________________________________________
Ips mailing list
Ips@ietf.org
https://www1.ietf.org/mailman/listinfo/ips

----
Rick McNeal

A good friend will come and bail you out of jail ... but, a true friend will be sitting next to you saying, "Damn ... that was fun!"



_______________________________________________
Ips mailing list
Ips@ietf.org
https://www1.ietf.org/mailman/listinfo/ips

[IETF]     [Linux iSCSI]     [Linux SCSI]     [Linux Resources]     [Yosemite News]     [IETF Announcements]     [IETF Discussion]     [SCSI]

Add to Google Powered by Linux