19 May
2014
19 May
'14
7:17 p.m.
Hi Jeroen,
Thanks! for taking care of my ping not responded situation.
Will try to comment on your answer.
You're right, the first trace on tun0 shows my 44 IP inside a packet sent
by amprgw
IP: len 104 169.228.66.251->44.153.0.1 ihl 20 ttl 39 prot IP
IP: len 84 184.72.226.23->44.153.0.1 ihl 20 ttl 38 DF prot ICMP
ICMP: type Echo Request id 58952 seq 1
Answer come back, but without being encaped
IP: len 84 44.153.0.1->184.72.226.23 ihl 20 ttl 175 prot ICMP
ICMP: type Echo Reply id 58952 seq 1
But this doesn't satisfy the original ping as is not acknowledged by
https://www.wormly.com/test_remote_ping
I intend that tun0 interfase, defined both in jnos and in linux, act as
link between the two systems, it does his work as all tcp sessions are
working ok and encaped between my 44 system and other 44s and viceversa.
44.153.0.3 is Linux, 44.153.0.10 is jnos which has 44.153.0.1 address.
Router gateway is 44.153.0.2, all this is tried to be related thru the
commands both thru shell and inside jnos. I have only one ethernet card on
the PC, so jnos doesn't have an ethernet interfase, and relies on this tun0
to comunicate to linux and to the internet.
Perhaps some of these commands are doing harm, and/or causing this strange
no ping answer from no 44 systems, but I was not able to found which one.
Commands I used are result of many test & try and result of many months of
experimenting.
What wonder me is that many 44 systems (not only mine) are not reachable
thru internet, but perfectly reachable encapsulating within 44 land.
I.E. try ping gb7cip.ampr.org , or ve3mch.ampr.org you won't reach via
Internet, but if you ping from within a 44 systems will respond ok.
You can try telnet to lu7abf.org.ar where you get into my bbs and then ping
above systems and see they do respond.
Note that you won't reach my system telneting to lu7abf.ampr.org as this
ping not working is avoiding this.
I am more than willing to change wathever is suggested to try this work, so
our 44 systems could be reachable.
Once more appreciate time and actual comments already given, I am looking
deeply to see what to change.
Best 73,
lu7abf, Pedro Converso
On Mon, May 19, 2014 at 6:48 PM, Jeroen Massar <jeroen(a)massar.ch> wrote:
(Please trim inclusions from previous messages)
_______________________________________________
On 2014-05-19 23:26, Pedro Converso wrote:
(Please trim inclusions from previous messages)
_______________________________________________
Hello fellows 44 crew
I have a ping not returning syndrome from my 44 address via Internet.
From any ampr system ping to 44.153.0.1 are answered ok, but
from internet don't receive answer to pings sent to 44.153.0.1
To try thru Internet I use https://www.wormly.com/test_remote_ping
setting 44.153.0.1 as destination for pings.
System is an ubuntu 10.03 and jnos2.0i, locally pings returns ok.
I made some traces that follows, some related autoexec.nos are at end.
Your traces are tcpdumps; you might want to try traceroute (or tracepath
when on Linux) instead.
[..]
Mon May 19 16:36:37 2014 - tun0 recv:
IP: len 104 169.228.66.251->44.153.0.1 ihl 20 ttl 39 prot IP
IP: len 84 184.72.226.23->44.153.0.1 ihl 20 ttl 38 DF prot ICMP
ICMP: type Echo Request id 58952 seq 1
That is a IPv4-in-IPv4-ICMPv4 packet.
Mon May 19 16:36:37 2014 - tun0 sent:
IP: len 84 44.153.0.1->184.72.226.23 ihl 20 ttl 175 prot ICMP
ICMP: type Echo Reply id 58952 seq 1
That is a IPv4-ICMPv4 packet.
That should match the former, IPv4-IPv4-ICMPv4
ifconfig tun0 ipaddress 44.153.0.10
ifconfig tun0 netmask 255.255.255.0
ifconfig tun0 mtu 1500
shell ifconfig tun0 44.153.0.3 pointopoint 44.153.0.10 mtu 1500 up
Please note that you are mixing up typical 'tunX' usage.
Typically 'tunX' is the tun/tap interface. The above makes this a
IPv4-in-IPv4 tunnel.
Interresting that you see the encapsulated result on tun0 though, one
would expect to see that on the underlying interface, not here.
That 1500 MTU is likely bad, unless you run jumboframes towards UCSD.
shell iptables -t nat -A PREROUTING -d
44.153.0.10/32 --proto 4 \-j DNAT
--to 44.153.0.1
This makes all packets going to 44.153.0.10 of type ipencap
(IPv4-in-IPv4) be send to 44.153.0.1 instead of the real destination.
Is that really what you want to do, and why?
shell iptables -t nat -A PREROUTING -d
44.153.0.10/32 --proto 94 \-j
DNAT
--to 44.153.0.1
Same but for the 94 variant of IPv4-in-IPv4. (though in theory one is
allowed to use any IP version inside of the other there)
Same Q: why do you do that?
shell iptables -t nat -A POSTROUTING -s
44.153.0.1/32 -o eth0 -p 4
shell iptables -t nat -A POSTROUTING -s 44.153.0.1/32 -o eth0 -p 94
Except for counting packets, these two rules have no effect.
shell sudo -s iptables -t nat -A PREROUTING -d
44.153.0.3/32 -p tcp
--dport
23 \-j DNAT --to 44.153.0.1:23
This causes everything towards .3 to go to .1
shell sudo -s iptables -t nat -A PREROUTING -d
44.153.0.3/32 -p tcp
--dport
6300 \-j DNAT --to 44.153.0.1:23
Same but redirecting port 6300.
shell sudo -s iptables -t nat -A PREROUTING -d
lu7abf.org.ar -p tcp
--dport
6300 \-j DNAT --to 44.153.0.1:23
shell sudo -s iptables -t nat -A POSTROUTING -s
44.153.0.1/32 -o eth0 -j
MASQUERADE
shell sudo -s iptables -t nat -A POSTROUTING -s 44.153.0.10/32 -o eth0
-j
MASQUERADE
This causes these two 44 net addresses to become NATted when they try
and communicate to your local network.
As those addresses are local though, it is unlikely there are used to
ever communicate with services living on eth0, unless you force programs
to do so.
ifconfig encap ip 44.153.0.1
ifconfig encap mtu 1500
ifconfig encap netmask 0xffffffff
ifconfig encap broadcast 255.255.255.255
ifc encap descript "IPIP Encapulation interface"
And here you configure a secondary encapsulation interface. Are you sure
you need two?
Greets,
Jeroen
_________________________________________
44Net mailing list
44Net(a)hamradio.ucsd.edu
http://hamradio.ucsd.edu/mailman/listinfo/44net