[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [MIP-QOS] Re: MIP-QOS -- Scope of discussions




Well, the only question I have is whether we
should be writing down requirements for which
there isn't actually a problem in the current crop
of protocols. This capability as written is
already available with RSVP. I sure hope we're not
headed down the path of writing requirements for a
new protocol from a clean slate here. IMO, we
should concentrate on the inadequacies of the
current crop with respect to mobility. Even if
there were a new protocol, we wouldn't be writing
its general requirements, after all.

ObReq: * QoS signaling protocols should
         differentiate between intentional
         topology changes and flapping with their
         damping strategies.

		   Mike

ext Glenn Morrow writes:
 > This reads very good to me.
 > 
 > -----Original Message-----
 > From: Hemant.Chaskar@nokia.com [mailto:Hemant.Chaskar@nokia.com]
 > Sent: Wednesday, April 18, 2001 10:54 AM
 > To: mip-qos@research.nokia.com
 > Subject: RE: [MIP-QOS] Re: MIP-QOS -- Scope of discussions
 > 
 > 
 > Can we condense the discussion about this issue of not being able to get old
 > QoS on the new path as the following requirement:
 > 
 > "The new end-to-end path may not support the old QoS agreement
 > because one or more routers or links are already at capacity. The QoS
 > mechanism MUST have provisions to handle this situation. It MAY also
 > be required to inform the MN about the impending QoS degradation. The
 > action taken in response to this situation (adapt, gracefully degrade,
 > re-negotiate or tear down) and the decision point (MN or service
 > provider policy) of the action are solution specific issues."
 > 
 > > -----Original Message-----
 > > From: marc.greis@nokia.com [mailto:marc.greis@nokia.com]
 > > Sent: 18. April 2001 11:29
 > > To: mip-qos@research.nokia.com
 > > Subject: RE: [MIP-QOS] Re: MIP-QOS -- Scope of discussions
 > > 
 > > 
 > > Mike,
 > > 
 > > > -----Original Message-----
 > > > From: ext Michael Thomas [mailto:mat@cisco.com]
 > > > Sent: 18 April, 2001 9:28 AM
 > > 
 > > >    As far as requirements goes, are there any
 > > >    cases where this is not purely a local policy
 > > >    decision of the mobile node? If there are, then
 > > >    there should be a requirement that the mobile
 > > >    node be the final arbiter of what is "acceptible
 > > >    QoS" based on local policy. 
 > > 
 > > I agree that the mobile node should make the final decision 
 > > (though you
 > > probably can not prevent the service provider from making the 
 > > decision for
 > > the mobile node). However, I think that creates the 
 > > additional requirement
 > > that there needs to be a mechanism for informing the mobile 
 > > node of the fact
 > > that the QoS has been downgraded. I don't think that it is a 
 > > good solution
 > > (in all cases) to wait for the user to notice that the QoS has become
 > > inacceptable.
 > > 
 > > > Note that this may actually be a requirement on MIP/SEAMOBY.
 > > 
 > > Actually, this may be an even broader requirement for QoS 
 > > mechanisms in
 > > general. Mobility is not the only case where QoS can degrade.
 > > 
 > > Marc
 > > 
 > <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
 > <HTML>
 > <HEAD>
 > <META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
 > <META NAME="Generator" CONTENT="MS Exchange Server version 5.5.2654.59">
 > <TITLE>RE: [MIP-QOS] Re: MIP-QOS -- Scope of discussions</TITLE>
 > </HEAD>
 > <BODY>
 > 
 > <P><FONT SIZE=2>This reads very good to me.</FONT>
 > </P>
 > 
 > <P><FONT SIZE=2>-----Original Message-----</FONT>
 > <BR><FONT SIZE=2>From: Hemant.Chaskar@nokia.com [<A HREF="mailto:Hemant.Chaskar@nokia.com">mailto:Hemant.Chaskar@nokia.com</A>]</FONT>
 > <BR><FONT SIZE=2>Sent: Wednesday, April 18, 2001 10:54 AM</FONT>
 > <BR><FONT SIZE=2>To: mip-qos@research.nokia.com</FONT>
 > <BR><FONT SIZE=2>Subject: RE: [MIP-QOS] Re: MIP-QOS -- Scope of discussions</FONT>
 > </P>
 > <BR>
 > 
 > <P><FONT SIZE=2>Can we condense the discussion about this issue of not being able to get old</FONT>
 > <BR><FONT SIZE=2>QoS on the new path as the following requirement:</FONT>
 > </P>
 > 
 > <P><FONT SIZE=2>&quot;The new end-to-end path may not support the old QoS agreement</FONT>
 > <BR><FONT SIZE=2>because one or more routers or links are already at capacity. The QoS</FONT>
 > <BR><FONT SIZE=2>mechanism MUST have provisions to handle this situation. It MAY also</FONT>
 > <BR><FONT SIZE=2>be required to inform the MN about the impending QoS degradation. The</FONT>
 > <BR><FONT SIZE=2>action taken in response to this situation (adapt, gracefully degrade,</FONT>
 > <BR><FONT SIZE=2>re-negotiate or tear down) and the decision point (MN or service</FONT>
 > <BR><FONT SIZE=2>provider policy) of the action are solution specific issues.&quot;</FONT>
 > </P>
 > 
 > <P><FONT SIZE=2>&gt; -----Original Message-----</FONT>
 > <BR><FONT SIZE=2>&gt; From: marc.greis@nokia.com [<A HREF="mailto:marc.greis@nokia.com">mailto:marc.greis@nokia.com</A>]</FONT>
 > <BR><FONT SIZE=2>&gt; Sent: 18. April 2001 11:29</FONT>
 > <BR><FONT SIZE=2>&gt; To: mip-qos@research.nokia.com</FONT>
 > <BR><FONT SIZE=2>&gt; Subject: RE: [MIP-QOS] Re: MIP-QOS -- Scope of discussions</FONT>
 > <BR><FONT SIZE=2>&gt; </FONT>
 > <BR><FONT SIZE=2>&gt; </FONT>
 > <BR><FONT SIZE=2>&gt; Mike,</FONT>
 > <BR><FONT SIZE=2>&gt; </FONT>
 > <BR><FONT SIZE=2>&gt; &gt; -----Original Message-----</FONT>
 > <BR><FONT SIZE=2>&gt; &gt; From: ext Michael Thomas [<A HREF="mailto:mat@cisco.com">mailto:mat@cisco.com</A>]</FONT>
 > <BR><FONT SIZE=2>&gt; &gt; Sent: 18 April, 2001 9:28 AM</FONT>
 > <BR><FONT SIZE=2>&gt; </FONT>
 > <BR><FONT SIZE=2>&gt; &gt;&nbsp;&nbsp;&nbsp; As far as requirements goes, are there any</FONT>
 > <BR><FONT SIZE=2>&gt; &gt;&nbsp;&nbsp;&nbsp; cases where this is not purely a local policy</FONT>
 > <BR><FONT SIZE=2>&gt; &gt;&nbsp;&nbsp;&nbsp; decision of the mobile node? If there are, then</FONT>
 > <BR><FONT SIZE=2>&gt; &gt;&nbsp;&nbsp;&nbsp; there should be a requirement that the mobile</FONT>
 > <BR><FONT SIZE=2>&gt; &gt;&nbsp;&nbsp;&nbsp; node be the final arbiter of what is &quot;acceptible</FONT>
 > <BR><FONT SIZE=2>&gt; &gt;&nbsp;&nbsp;&nbsp; QoS&quot; based on local policy. </FONT>
 > <BR><FONT SIZE=2>&gt; </FONT>
 > <BR><FONT SIZE=2>&gt; I agree that the mobile node should make the final decision </FONT>
 > <BR><FONT SIZE=2>&gt; (though you</FONT>
 > <BR><FONT SIZE=2>&gt; probably can not prevent the service provider from making the </FONT>
 > <BR><FONT SIZE=2>&gt; decision for</FONT>
 > <BR><FONT SIZE=2>&gt; the mobile node). However, I think that creates the </FONT>
 > <BR><FONT SIZE=2>&gt; additional requirement</FONT>
 > <BR><FONT SIZE=2>&gt; that there needs to be a mechanism for informing the mobile </FONT>
 > <BR><FONT SIZE=2>&gt; node of the fact</FONT>
 > <BR><FONT SIZE=2>&gt; that the QoS has been downgraded. I don't think that it is a </FONT>
 > <BR><FONT SIZE=2>&gt; good solution</FONT>
 > <BR><FONT SIZE=2>&gt; (in all cases) to wait for the user to notice that the QoS has become</FONT>
 > <BR><FONT SIZE=2>&gt; inacceptable.</FONT>
 > <BR><FONT SIZE=2>&gt; </FONT>
 > <BR><FONT SIZE=2>&gt; &gt; Note that this may actually be a requirement on MIP/SEAMOBY.</FONT>
 > <BR><FONT SIZE=2>&gt; </FONT>
 > <BR><FONT SIZE=2>&gt; Actually, this may be an even broader requirement for QoS </FONT>
 > <BR><FONT SIZE=2>&gt; mechanisms in</FONT>
 > <BR><FONT SIZE=2>&gt; general. Mobility is not the only case where QoS can degrade.</FONT>
 > <BR><FONT SIZE=2>&gt; </FONT>
 > <BR><FONT SIZE=2>&gt; Marc</FONT>
 > <BR><FONT SIZE=2>&gt; </FONT>
 > </P>
 > 
 > </BODY>
 > </HTML>