RE: Minor issue with MN advertisement sequence number handling

Maybe I'm missing something here.
At step 5, when FA restarts, why is the advertisement sent
with sequence
number 6? Wouldn't it be 1?
The window when this error can happen is that FA sends
advertisement
with seq num 1, restarts and again sends aa with seq num =1.
Then the
two are AAs have equal seq num and MN will not register. I
think in any
other case, restart would not cause a problem.  

Regards,
Arun


  

-----Original Message-----
From: Ahmad Muhanna [mailto:amuhannanortel.com] 
Sent: Friday, March 23, 2007 3:46 PM
To: Bjorn Andersson
Cc: mip4ietf.org
Subject: RE: [Mip4] Minor issue with MN advertisement
sequence number
handling


Now I understand what you meant by "If the FA is
restarted before it has
sent a second advertisement"
I also agree that this case is very unlikely.

Regards,
Ahmad
 

> -----Original Message-----
> From: bjonderssongmail.com [mailto:bjonderssongmail.com] On Behalf 
> Of Bjorn Andersson
> Sent: Friday, March 23, 2007 8:27 AM
> To: Muhanna, Ahmad (RICH1:2H10)
> Cc: mip4ietf.org
> Subject: Re: [Mip4] Minor issue with MN advertisement
sequence number 
> handling
> 
> On 3/23/07, Ahmad Muhanna <amuhannanortel.com> wrote:
> > Bjorn Andersson [mailto:bjorniki.fi]:
> > > If the FA is restarted before it has sent a
second
> advertisement the
> > > MN will not reregister.
> >
> > Why? Can you please explain.
> 
> 1. FA starts
> 2. FA sends advertisement with sequence number 0 3. MN
registers via 
> FA 4. FA restarts 5. FA sends advertisement with
sequence number 0 6. 
> Sequence number is equal to, not less than, previous
sequence number. 
> MN does not register.
> 7. MN sends packet to FA for tunneling, but MN is not
registered to FA

> (since the FA restarted after the registration), so the
FA does not 
> tunnel the packet to the HA. This will be the case
until the 
> registration expires.
> 
> But as I said, it is minor issues since it quite
unlikely to occur, 
> but it could still be improved and fixed in the RFC.
> Or am I missing something?
> 
>   Bjorn
> 

--
Mip4 mailing list: Mip4ietf.org
    Web interface: https://w
ww1.ietf.org/mailman/listinfo/mip4
     Charter page: h
ttp://www.ietf.org/html.charters/mip4-charter.html
Supplemental site: http://www.mip4.org/

--
Mip4 mailing list: Mip4ietf.org
    Web interface: https://w
ww1.ietf.org/mailman/listinfo/mip4
     Charter page: h
ttp://www.ietf.org/html.charters/mip4-charter.html
Supplemental site: http://www.mip4.org/

On 3/23/07, Arun SG <arunsgairvana.com> wrote:
> Maybe I'm missing something here.
> At step 5, when FA restarts, why is the advertisement
sent with sequence
> number 6? Wouldn't it be 1?

Your, or someone's mail reader has lost the new line after
the zero.
The number six you see if from step 6. Step 5 says it will
be sent
with sequence number zero, the first sequence number (it
could be 1 if
it always starts with 1).

> The window when this error can happen is that FA sends
advertisement
> with seq num 1, restarts and again sends aa with seq
num =1. Then the
> two are AAs have equal seq num and MN will not
register. I think in any
> other case, restart would not cause a problem.

Yes, this is exactly what I am saying. The time period
between the two
first advertisements. If the FA advertises once per 30s then
the FA
should not be restarted for 30s or MNs may lose their
connection.

It's a minor issue, but has already caused confusion in a
real life
setup (that's how I noticed it). So can't we just change the
text to
"less or equal to" ? Or are there protocolwise any
negative
implications of changing the text?

  Bjorn

-- 
Mip4 mailing list: Mip4ietf.org
    Web interface: https://w
ww1.ietf.org/mailman/listinfo/mip4
     Charter page: h
ttp://www.ietf.org/html.charters/mip4-charter.html
Supplemental site: http://www.mip4.org/