[Sis-csi] Telecon today, 10/12 @ 3:30PM EST
Keith Hogie
Keith.Hogie at gsfc.nasa.gov
Thu Oct 12 14:53:16 EDT 2006
Greg,
I think the concept of a single subnet or a small number of subnets
sounds good. That would allow static configs in lots of places that
just say that since this packet is going to space, it must go in
a certain direction. I think we will have many scenarios that are
somewhat stable and that will work fine.
For more mobile nodes I'm not as sure about the multiple route
definitions. Also I'm not sure the C&DH will always know in advance
who it is going to be talking to. For standard spacecraft, links Hogie
are pretty scheduled and that may be true, but for manned nodes
wandering around the moon in proximity environments, I think links
may be less scheduled and more dynamic.
As Marc mentioned, I think we need to get our scenarios defined
better. Then we may see links that are fairly stable (e.g. 10's
of minutes to hours) versus other links that are fairly dynamic
(e.g. minutes). We may end up with solutions where more comprehensive
but chatty protocols like OSPF are used on the more stable links
and MANET or Mobile IP protocols get used on the more mobile links.
Still lots of work to do in this area.
Keith Hogie
gregory.menke at gsfc.nasa.gov wrote:
> In the CCA we have been tending towards the single subnet idea at the
> moment, though our scenarios are relatively simplistic. The idea was to
> avoid addressing irregularly end-to-end so routing tables can be
> established to do something useful regardless of which interface a
> packet arrives through. This leads to the manual interface & routing
> table config problem, but I don't think there is an easy way to avoid
> it.
>
> I agree that any numbering procedure requiring a bidirectional protocol
> is likely going to become problematic. Such a situation doesn't
> automatically exclude snazzy protocols to handle the situation but by
> themselves they are insufficient to handle all operational cases-
> something as simple as a long-haul link where the round-trip time is
> long is enough to exclude the usual router discovery protocols. Thus it
> seemed to us preferable in general to assign subnets statically and use
> forwarding rules to make traffic flow.
>
> The mobility issue then raises its head, and I think the technical
> response is to have a variety of predefined routing table and interface
> configs that may be imposed as connectivity warrants; flight software
> knows its talking to spacecraft X because thats in the contact schedule,
> so in addition to setting up the antenna & radio, it loads the
> appropriate IP interface config and routing table entries.
>
> If the CCA were in the mix, then it would probably receive the C&DH
> command to set up the IP routing table & interface config stuff but
> might or might not be involved in setting up the radio & antenna.
>
> Since the flight software is already facing the complexity of handing
> the antenna and radio, the additional work setting up the IP interface &
> routing table is fairly minimal.
>
> Greg
>
>
> Scott, Keith L. writes:
> > Right, and this was in fact the genesis for some rather odd addressing
> > schemes that have been proposed in other venues, such as placing all of
> > the 'space-facing' interfaces into a single subnet. Things get sticky
> > when you want to communicate from the ground directly to S/C1 or if you
> > want to use S/C2 as a relay to get to S/C1. I'd prefer some solution
> > that didn't involve renumbering interfaces on the fly (ICK!).
> >
> > The unnumbered interfaces idea might be really good, especially since I
> > haven't heard anybody proposing anything except serial data links
> > between space and ground or between spacecraft. I'm a bit leery of
> > anything that requires bidirectional connectivity (like negotiating
> > link-local addresses) as bidirectional connectivity's not necessarily a
> > given.
> >
> > I also don't think we're necessarily tied to what commercial routers
> > allow or don't allow; though knowing when we're in/out of the bounds of
> > commercial implementations is VERY useful.
> >
> > --keith
> >
> >
> >
> > ________________________________
> >
> > From: Ivancic, William D. (GRC-RCN0)
> > [mailto:william.d.ivancic at nasa.gov]
> > Sent: Thursday, October 12, 2006 9:51 AM
> > To: Scott, Keith L.; sis-csi at mailman.ccsds.org
> > Subject: RE: [Sis-csi] Telecon today, 10/12 @ 3:30PM EST
> >
> >
> > Traditionally routers have to be on the same network. This
> > creates problems for space-based networks - particularly if one wishes
> > to share infrastructure. We are seeing similar issues in aeronautics
> > network centric operations. Some thoughts that need investigation
> > are:
> >
> > Can one use unnumbered serial interfaces to connect routers?
> > This would allow router interfaces to not have to be on the same
> > network. I think we tried this and it did not work, but we haven't
> > tried to hard due to manpower and money.
> >
> > Can IPv6 link-local addressing help? I haven't had the
> > opportunity to see how IPv6 can help. Auto-configuration may also be
> > a solution. I know of one company that was looking at dynamic NEMO
> > techniques that use some IPv6 capability to some interesting dynamic
> > things (no home agent required). I will see if I can find the
> > Internet Draft is one exists.
> >
> >
> > Will
> >
> >
> > ******************************
> >
> > William D. Ivancic
> >
> > Phone 216-433-3494
> >
> > Fax 216-433-8705
> >
> > Lab 216-433-2620
> >
> > Mobile 440-503-4892
> >
> > http://roland.grc.nasa.gov/~ivancic
> > <http://roland.grc.nasa.gov/~ivancic>
> >
> >
> >
> > ________________________________
> >
> > From: sis-csi-bounces at mailman.ccsds.org
> > [mailto:sis-csi-bounces at mailman.ccsds.org] On Behalf Of Scott, Keith L.
> > Sent: Thursday, October 12, 2006 8:44 AM
> > To: sis-csi at mailman.ccsds.org
> > Subject: [Sis-csi] Telecon today, 10/12 @ 3:30PM EST
> >
> >
> > 12:30 Pacific
> > 3:30 Eastern
> >
> > NEW NUMBER: 703.983.6338 (703.983.MEET) [It's not my fault]
> >
> > I'd like to talk about plans to get the first set of
> > specifications going, starting with addressing and routing plans. Come
> > armed with your thoughts on what routing protocols might be appropriate
> > where, how we might want to break things into CIDR blocks, etc.
> >
> > --keith
> >
> >
> > <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
> > <HTML><HEAD>
> > <META http-equiv=Content-Type content="text/html; charset=us-ascii">
> > <META content="MSHTML 6.00.2900.2963" name=GENERATOR></HEAD>
> > <BODY>
> > <DIV dir=ltr align=left><SPAN class=875285213-12102006><FONT face=Arial
> > color=#0000ff size=2>Right, and this was in fact the genesis for some rather odd
> > addressing schemes that have been proposed in other venues, such as placing all
> > of the 'space-facing' interfaces into a single subnet. Things get sticky
> > when you want to communicate from the ground directly to S/C1 or if you want to
> > use S/C2 as a relay to get to S/C1. I'd prefer some solution that didn't
> > involve renumbering interfaces on the fly (ICK!).</FONT></SPAN></DIV>
> > <DIV dir=ltr align=left><SPAN class=875285213-12102006><FONT face=Arial
> > color=#0000ff size=2></FONT></SPAN> </DIV>
> > <DIV dir=ltr align=left><SPAN class=875285213-12102006><FONT face=Arial
> > color=#0000ff size=2>The unnumbered interfaces idea might be really good,
> > especially since I haven't heard anybody proposing anything except serial data
> > links between space and ground or between spacecraft. I'm a bit leery of
> > anything that requires bidirectional connectivity (like negotiating link-local
> > addresses) as bidirectional connectivity's not necessarily a
> > given.</FONT></SPAN></DIV>
> > <DIV dir=ltr align=left><SPAN class=875285213-12102006><FONT face=Arial
> > color=#0000ff size=2></FONT></SPAN> </DIV>
> > <DIV dir=ltr align=left><SPAN class=875285213-12102006><FONT face=Arial
> > color=#0000ff size=2>I also don't think we're necessarily tied to what
> > commercial routers allow or don't allow; though knowing when we're in/out of the
> > bounds of commercial implementations is VERY useful.</FONT></SPAN></DIV>
> > <DIV dir=ltr align=left><SPAN class=875285213-12102006><FONT face=Arial
> > color=#0000ff size=2></FONT></SPAN> </DIV>
> > <DIV dir=ltr align=left><SPAN
> > class=875285213-12102006> <FONT
> > face=Arial color=#0000ff size=2>--keith</FONT></SPAN></DIV>
> > <DIV dir=ltr align=left><SPAN class=875285213-12102006><FONT face=Arial
> > color=#0000ff size=2></FONT></SPAN> </DIV><BR>
> > <BLOCKQUOTE dir=ltr style="MARGIN-RIGHT: 0px">
> > <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left>
> > <HR tabIndex=-1>
> > <FONT face=Tahoma size=2><B>From:</B> Ivancic, William D. (GRC-RCN0)
> > [mailto:william.d.ivancic at nasa.gov] <BR><B>Sent:</B> Thursday, October 12,
> > 2006 9:51 AM<BR><B>To:</B> Scott, Keith L.;
> > sis-csi at mailman.ccsds.org<BR><B>Subject:</B> RE: [Sis-csi] Telecon today,
> > 10/12 @ 3:30PM EST<BR></FONT><BR></DIV>
> > <DIV></DIV>
> > <DIV dir=ltr align=left><SPAN class=875074513-12102006><FONT face=Arial
> > color=#0000ff size=2>Traditionally routers have to be on the same
> > network. This creates problems for space-based networks -
> > particularly if one wishes to share infrastructure. We are seeing
> > similar issues in aeronautics network centric operations. Some
> > thoughts that need investigation are:</FONT></SPAN></DIV>
> > <DIV dir=ltr align=left><SPAN class=875074513-12102006><FONT face=Arial
> > color=#0000ff size=2></FONT></SPAN> </DIV>
> > <DIV dir=ltr align=left><SPAN class=875074513-12102006><FONT face=Arial
> > color=#0000ff size=2>Can one use unnumbered serial interfaces to connect
> > routers? This would allow router interfaces to not have to be on the
> > same network. I think we tried this and it did not work, but we haven't
> > tried to hard due to manpower and money.</FONT></SPAN></DIV>
> > <DIV dir=ltr align=left><SPAN class=875074513-12102006><FONT face=Arial
> > color=#0000ff size=2></FONT></SPAN> </DIV>
> > <DIV dir=ltr style="MARGIN-RIGHT: 0px" align=left><SPAN
> > class=875074513-12102006><FONT face=Arial color=#0000ff size=2>Can IPv6
> > link-local addressing help? I haven't had the opportunity to see
> > how IPv6 can help. Auto-configuration may also be a
> > solution. I know of one company that was looking at dynamic NEMO
> > techniques that use some IPv6 capability to some interesting dynamic things
> > (no home agent required). I will see if I can find the Internet
> > Draft is one exists.</FONT></SPAN></DIV>
> > <DIV dir=ltr style="MARGIN-RIGHT: 0px" align=left><SPAN
> > class=875074513-12102006><FONT face=Arial color=#0000ff
> > size=2></FONT></SPAN> </DIV>
> > <DIV dir=ltr style="MARGIN-RIGHT: 0px" align=left><SPAN
> > class=875074513-12102006><FONT face=Arial color=#0000ff
> > size=2></FONT></SPAN> </DIV>
> > <DIV dir=ltr style="MARGIN-RIGHT: 0px" align=left><SPAN
> > class=875074513-12102006><FONT face=Arial color=#0000ff
> > size=2>Will</FONT></SPAN></DIV>
> > <DIV> </DIV>
> > <P align=left><FONT face=Courier
> > size=2>******************************</FONT></P>
> > <P align=left><FONT face=Courier size=2>William D. Ivancic</FONT></P>
> > <P align=left><FONT face=Courier size=2>Phone 216-433-3494</FONT></P>
> > <P align=left><FONT face=Courier size=2>Fax 216-433-8705</FONT></P>
> > <P align=left><FONT face=Courier size=2>Lab 216-433-2620</FONT></P>
> > <P align=left><FONT face=Courier size=2>Mobile 440-503-4892</FONT></P>
> > <P align=left><A href="http://roland.grc.nasa.gov/~ivancic"><U><FONT
> > color=#0000ff><FONT face=Courier
> > size=2>http://roland.grc.nasa.gov/~ivancic</FONT></U></FONT></A><FONT
> > face=Courier size=2> </FONT></P>
> > <DIV> </DIV><BR>
> > <DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left>
> > <HR tabIndex=-1>
> > <FONT face=Tahoma size=2><B>From:</B> sis-csi-bounces at mailman.ccsds.org
> > [mailto:sis-csi-bounces at mailman.ccsds.org] <B>On Behalf Of </B>Scott, Keith
> > L.<BR><B>Sent:</B> Thursday, October 12, 2006 8:44 AM<BR><B>To:</B>
> > sis-csi at mailman.ccsds.org<BR><B>Subject:</B> [Sis-csi] Telecon today, 10/12 @
> > 3:30PM EST<BR></FONT><BR></DIV>
> > <DIV></DIV>
> > <DIV><SPAN class=453124112-12102006><FONT face=Arial size=2>12:30
> > Pacific</FONT></SPAN></DIV>
> > <DIV><SPAN class=453124112-12102006><FONT face=Arial size=2>3:30
> > Eastern</FONT></SPAN></DIV>
> > <DIV><SPAN class=453124112-12102006><FONT face=Arial
> > size=2></FONT></SPAN> </DIV>
> > <DIV><SPAN class=453124112-12102006><FONT face=Arial size=2>NEW NUMBER:
> > 703.983.6338 (703.983.MEET) [It's not my
> > fault]</FONT></SPAN></DIV>
> > <DIV><SPAN class=453124112-12102006><FONT face=Arial
> > size=2></FONT></SPAN> </DIV>
> > <DIV><SPAN class=453124112-12102006><FONT face=Arial size=2>I'd like to talk
> > about plans to get the first set of specifications going, starting with
> > addressing and routing plans. Come armed with your thoughts on what
> > routing protocols might be appropriate where, how we might want to break
> > things into CIDR blocks, etc.</FONT></SPAN></DIV>
> > <DIV><SPAN class=453124112-12102006><FONT face=Arial
> > size=2></FONT></SPAN> </DIV>
> > <DIV><SPAN class=453124112-12102006>
> > <FONT face=Arial size=2>--keith</FONT></SPAN></DIV>
> > <DIV><SPAN
> > class=453124112-12102006></SPAN> </DIV></BLOCKQUOTE></BODY></HTML>
> > _______________________________________________
> > Sis-CSI mailing list
> > Sis-CSI at mailman.ccsds.org
> > http://mailman.ccsds.org/cgi-bin/mailman/listinfo/sis-csi
>
>
> _______________________________________________
> Sis-CSI mailing list
> Sis-CSI at mailman.ccsds.org
> http://mailman.ccsds.org/cgi-bin/mailman/listinfo/sis-csi
--
----------------------------------------------------------------------
Keith Hogie e-mail: Keith.Hogie at gsfc.nasa.gov
Computer Sciences Corp. office: 301-794-2999 fax: 301-794-9480
7700 Hubble Dr.
Lanham-Seabrook, MD 20706 USA 301-286-3203 @ NASA/Goddard
----------------------------------------------------------------------
More information about the Sis-CSI
mailing list