Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.os.linux.networking > #1769

Re: 192.168

Show all headers | View raw

On 20 Oct 2012, in the Usenet newsgroup comp.os.linux.networking, in article
<slrnk85cgf.1d3.grahn+nntp@frailea.sa.invalid>, Jorgen Grahn wrote:

>orcus@pwr.wroc.pl wrote:
>> Jorgen Grahn <grahn+nntp@snipabacken.se> wrote:
>>> orcus@pwr.wroc.pl wrote:

>>>> Nope - a lot of ISPs use privete adresses to connect their routers. 

IP addresses are valuable commodities - as of Monday, 93.66 percent of
the non-RFC5735 IPv4 addresses have been assigned or allocated.  Why
should an ISP waste them on hosts that normal users (never mind those
out on the Big Bad Internet) won't be accessing?  See the "Category 1:"
definition on the bottom of page 2 of RFC1918.

>>>> Sometimes we can see "non-routable" adsresses in output of
>>>> traceroute - somewere in "Internet".

>>> If an address shows up in traceroute output, I should be able to
>>> access it.

Why?  What gives you ANY reason to expect that "backbone.com" is going
to allow you to connect to their router?   It's not a web server, and
it's offering NO services to Internet users other than forwarding or
routing packets.

>> Which fragment fragment of RFC they violate?

Second paragraph on page 5 of RFC1918:

   Because private addresses have no global meaning, routing information
   about private networks shall not be propagated on inter-enterprise
   links, and packets with private source or destination addresses
   should not be forwarded across such links.

HOWEVER - note that RFC1918 is a "Best Current Practice" (see RFC1818
which is historic, but explains that a BCP is not, and should not be
considered "a standard").  Against this, you have RFC1812
(Requirements for IP Version 4 Routers) section 4.3.2.4 which says:

   Except where this document specifies otherwise, the IP source address
   in an ICMP message originated by the router MUST be one of the IP
   addresses associated with the physical interface over which the ICMP
   message is transmitted.

and RFC1812 _is_ a standard.

>You say you can see e.g. 192.168.1.2 in traceroute output.  That
>means you have actually received an ICMP time exceeded (or whatever
>it's called) message from that host.  Surely then it's not just
>"inside their network"?

True - but an ICMP error message is "one way" - you don't reply to it
(RFC0792, fourth paragraph of "Introduction" on bottom of page 1), so
a little common sense would be helpful.  Read the second paragraph of
section 7 of RFC5735.

>I may have a host 192.168.1.2 in my local network already. I cannot
>think right now of any bad effects of getting ICMP from a doppelganger
>(except for confusion in the logs) but I'd prefer for it never to
>happen!

See RFC2827 (and RFC3704).  Your router or firewall should not be
forwarding packets from the "outside" that have a source address of
the LAN on the "inside".  To avoid confusion, you may wish to filter
all of the addresses listed in RFC5735 (and RFC5156 if you are using
IPv6) as recommended by the first paragraph of of section 7 of RFC5735.

        Old guy

Back to comp.os.linux.networking | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

192.168 bob@coolfone.comze.com - 2012-10-17 08:37 -0700
  Re: 192.168 Richard Kettlewell <rjk@greenend.org.uk> - 2012-10-17 16:40 +0100
    Re: 192.168 ein <ein@no.spam> - 2012-10-17 20:04 +0200
      Re: 192.168 Marc Haber <mh+usenetspam1118@zugschl.us> - 2012-10-17 21:23 +0200
        Re: 192.168 ein <ein@no.spam> - 2012-10-17 22:37 +0200
          Re: 192.168 Marc Haber <mh+usenetspam1118@zugschl.us> - 2012-10-18 18:29 +0200
            Re: 192.168 ein <ein@no.spam> - 2012-10-19 07:54 +0200
  Re: 192.168 Maurizio Loreti <mlo@foobar.it> - 2012-10-17 20:37 +0200
  Re: 192.168 unruh <unruh@invalid.ca> - 2012-10-18 01:43 +0000
    Re: 192.168 Richard Kettlewell <rjk@greenend.org.uk> - 2012-10-18 09:22 +0100
      Re: 192.168 ein <ein@no.spam> - 2012-10-19 07:51 +0200
        Re: 192.168 Noob <root@127.0.0.1> - 2012-10-19 12:16 +0200
          Re: 192.168 orcus@pwr.wroc.pl - 2012-10-19 18:54 +0000
            Re: 192.168 Jorgen Grahn <grahn+nntp@snipabacken.se> - 2012-10-20 06:51 +0000
              Re: 192.168 orcus@pwr.wroc.pl - 2012-10-20 07:46 +0000
                Re: 192.168 Jorgen Grahn <grahn+nntp@snipabacken.se> - 2012-10-20 08:00 +0000
                Re: 192.168 orcus@pwr.wroc.pl - 2012-10-20 13:43 +0000
                Re: 192.168 Jorgen Grahn <grahn+nntp@snipabacken.se> - 2012-10-20 14:15 +0000
                Re: 192.168 orcus@pwr.wroc.pl - 2012-10-20 14:54 +0000
                Re: 192.168 Moe Trin <ibuprofin@painkiller.example.tld.invalid> - 2012-10-20 15:21 -0500
                Re: 192.168 Moe Trin <ibuprofin@painkiller.example.tld.invalid> - 2012-10-20 15:20 -0500
          Re: 192.168 Pascal Hambourg <boite-a-spam@plouf.fr.eu.org> - 2012-10-20 13:55 +0200
        Re: 192.168 unruh <unruh@invalid.ca> - 2012-10-21 00:33 +0000
          Re: 192.168 orcus@pwr.wroc.pl - 2012-10-21 10:22 +0000
            Re: 192.168 Markus Koßmann <mkossmann_n1@gmx.de> - 2012-10-21 15:53 +0200
          Re: 192.168 GangGreene <GangGreene@example.com> - 2012-10-21 07:06 -0400
  Re: 192.168 Vanni <Vanniz@nospam.thanks> - 2012-10-18 07:56 +0200
    Re: 192.168 Johann Klammer <klammerj@NOSPAM.a1.net> - 2012-10-18 06:40 +0200

csiph-web