Question on Ipv6 address

Keith Medcalf kmedcalf at dessus.com
Wed Mar 27 05:40:42 UTC 2013


 The "default" mtu of 576 is because, well, 2400 baud signaling is pretty
 darn slow and interactive performance (or any kind of multileaving of more
 than a single connection packet stream) is, what do we call it, laggy.
 Sort of like trying to telnet while doing a bulk transfer if you have
 bloated buffers, and do not use a decent QoS scheduler -- only with echo
 times in the order of seconds per character.
 
 I believe LCP uses a signed two-octet integer for frame size negotiation,
 so you can negotiate quite large frames if you so desire and so configure
 your endpoints.
 
 ---
 ()  ascii ribbon campaign against html e-mail
 /\  www.asciiribbon.org
 
> 
> > -----Original Message-----
> > From: Frank Bulk (iname.com) [mailto:frnkblk at iname.com]
> > Sent: Tuesday, 26 March, 2013 22:46
> > To: 'Mark Jeremy'; Justin Wilson
> > Cc: nanog at nanog.org
> > Subject: RE: Question on Ipv6 address
> >
> > My understanding is that because IPv6 has a minimum MTU of 1280 and
> dial-
> > up
> > maxes out at 576, that special measures must be taken for IPv6 to work
> > over
> > a dial-up connection.
> >
> > Please correct me if someone has this working out of the box.
> >
> > Frank
> >
> > -----Original Message-----
> > From: Mark Jeremy [mailto:mejndp at rit.edu]
> > Sent: Tuesday, March 26, 2013 2:10 PM
> > To: Justin Wilson
> > Cc: nanog at nanog.org
> > Subject: RE: Question on Ipv6 address
> >
> > Justin,
> >
> > Dial-up modem is just a layer 2 device with no IP address. Just think of
> > it
> > as a converter, its sole function is to convert the telephone line to
> > something your PC can use, in this case, Ethernet. Both IPv4 and IPv6
> > operate on the layer 3 of the OSI model which is taken care of by the
> RAS.
> > So basically any dial-up modem support IPv6.
> >
> > -MJ
> >
> > -----Original Message-----
> > From: Justin Wilson [mailto:lists at mtin.net]
> > Sent: Tuesday, March 26, 2013 12:06 PM
> > To: NANOG
> > Subject: Re: Question on Ipv6 address
> >
> > I don't mean to hijack the thread so if someone wants to open a new one
> > that¹s cool.  But my question is what dial-up hardware supports v6? I am
> > *assuming* Cisco does.
> >
> >
> > Justin
> >
> > --
> > Justin Wilson <j2sw at mtin.net>
> > Aol & Yahoo IM: j2sw
> > http://www.mtin.net/blog ­ xISP News
> > http://www.zigwireless.com ­ High Speed Internet Options
> > http://www.thebrotherswisp.com ­ The Brothers Wisp
> >
> >
> >
> > -----Original Message-----
> > From: Joe <sj_hznm at hotmail.com>
> > Date: Tuesday, March 26, 2013 11:39 AM
> > To: NANOG <nanog at nanog.org>
> > Subject: Question on Ipv6 address
> >
> > >I'm new to Ipv6 and trying to understanding something about IPv6 in
> > >service provider network.
> > >I've got the following questions , could anybody do some helps?
> > >1. In a dial-up network (Q-in-Q for each customer who dials in ) Should
> > >each customer be assigned to ipv6 subnet prefix like /64 unique
> > >universily?  I've read   a rfc which stated point-to-point like should
> be
> > >assigned /64. But to my understanding, in dial-up   network , each user
> > >should only needed to be assigned a single ipv4 address, with wich
> > >customer   could used in his PC or his home router.
> > >2. In dial-up network,  could each vlan's ipv6 link-id  be planned with
> > >its vlan number?     if so,  IP v6 address confliction could be avoided
> > >when  BAS is assigned a /64 or longer prefix.
> > >3. we are testing some BAS with IPv6 accessing, in radius accouting
> > >packets, there is     IP-v6-prefix, Ip-v6-link-id,
> > >Ip-v6-delegated-prefix.    how could dial-up PC's  IPv6    address be
> > >calculated with above information?
> > >4. should it be necessary to plan  different
> > >IP-v6-prefix(IP-v6-delegated-prefix) for each dial-up customers  in
> BAS?
> > >5. How could delegated IPv6 prefix be used in service provider's
> network?
> > >is this useful in dial-up access network?
> > >
> > >each word will be highly appreciated.
> > >Joe
> > >
> >
> >
> >
> >








More information about the NANOG mailing list