weird issue with a node
13 views
Skip to first unread message
Patrick Perdue
Apr 20, 2024, 1:19:31 PM4/20/24
to repeaterphone-users
Greetings:
I have a node (508425) which is a remote base I set up in my old home town.
Since it regularly sits on repeaters with long hangtimes, I like
accessing it with Repeaterphone, so I can TX while the radio is still
receiving.
Recently, though, my parents' ISP, where the node lives, has gone to an
odd sort of CG-NAT model. I say sort of CG-NAT, because no inbound TCP
connections work, the router's WAN IP is a 100.x.x.x, apparently CG-NAT,
but the IP address everything else sees is in the 72.x.x.x range.
Furthermore, inbound UDP connections still get routed.
I have another machine on their network connected to a Wireguard VPN so
I can access the console of the node, but the node itself has nothing to
do with any VPN connection at the moment.
Ever since then, if I try connecting to this node through Repeaterphone,
it breaks in a fun and interesting way. I can still connect to this node
through the [radio-secure] context, I.E. the way you would normally
connect to a node through authenticated and registered nodes.
This isn't Repeaterphone's fault, but I find it really interesting, and
not sure exactly what's going on.
Here's what it looks like when something tries to establish a connection
through WT to that node. See the http:400. I have also seem http:500
internal server error.
I can connect to other nodes from Repeaterphone.
-- Accepting AUTHENTICATED call from <IP_Redacted>:
> requested format = ulaw,
> requested prefs = (ulaw),
> actual format = ulaw,
> host prefs = (ulaw|gsm),
> priority = mine
-- Executing [s@allstar-public:1]
Ringing("IAX2/allstar-public-412", "") in new stack
-- Executing [s@allstar-public:2] Set("IAX2/allstar-public-412",
"RESP=<html>
<head><title>400 Request Header Or Cookie Too Large</title></head>
<body>
<center><h1>400 Bad Request</h1></center>
<center>Request Header Or Cookie Too Large</center>
<hr><center>nginx/1.17.6</center>
</body>
") in new stack
-- Executing [s@allstar-public:3] Set("IAX2/allstar-public-412",
"NODENUM=508425") in new stack
-- Executing [s@allstar-public:4] GotoIf("IAX2/allstar-public-412",
"0?hangit") in new stack
-- Executing [s@allstar-public:5] GotoIf("IAX2/allstar-public-412",
"0?hangit") in new stack
-- Executing [s@allstar-public:6] GotoIf("IAX2/allstar-public-412",
"1?hangit") in new stack
-- Goto (allstar-public,s,14)
-- Executing [s@allstar-public:14]
Answer("IAX2/allstar-public-412", "") in new stack
-- Executing [s@allstar-public:15] Wait("IAX2/allstar-public-412",
"1") in new stack
-- Executing [s@allstar-public:16]
Hangup("IAX2/allstar-public-412", "") in new stack
== Spawn extension (allstar-public, s, 16) exited non-zero on
'IAX2/allstar-public-412'
-- Hungup 'IAX2/allstar-public-412'
linkbox*CLI>
Anyone have a clue what could be happening here?
73
N2DYI
I have a node (508425) which is a remote base I set up in my old home town.
Since it regularly sits on repeaters with long hangtimes, I like
accessing it with Repeaterphone, so I can TX while the radio is still
receiving.
Recently, though, my parents' ISP, where the node lives, has gone to an
odd sort of CG-NAT model. I say sort of CG-NAT, because no inbound TCP
connections work, the router's WAN IP is a 100.x.x.x, apparently CG-NAT,
but the IP address everything else sees is in the 72.x.x.x range.
Furthermore, inbound UDP connections still get routed.
I have another machine on their network connected to a Wireguard VPN so
I can access the console of the node, but the node itself has nothing to
do with any VPN connection at the moment.
Ever since then, if I try connecting to this node through Repeaterphone,
it breaks in a fun and interesting way. I can still connect to this node
through the [radio-secure] context, I.E. the way you would normally
connect to a node through authenticated and registered nodes.
This isn't Repeaterphone's fault, but I find it really interesting, and
not sure exactly what's going on.
Here's what it looks like when something tries to establish a connection
through WT to that node. See the http:400. I have also seem http:500
internal server error.
I can connect to other nodes from Repeaterphone.
-- Accepting AUTHENTICATED call from <IP_Redacted>:
> requested format = ulaw,
> requested prefs = (ulaw),
> actual format = ulaw,
> host prefs = (ulaw|gsm),
> priority = mine
-- Executing [s@allstar-public:1]
Ringing("IAX2/allstar-public-412", "") in new stack
-- Executing [s@allstar-public:2] Set("IAX2/allstar-public-412",
"RESP=<html>
<head><title>400 Request Header Or Cookie Too Large</title></head>
<body>
<center><h1>400 Bad Request</h1></center>
<center>Request Header Or Cookie Too Large</center>
<hr><center>nginx/1.17.6</center>
</body>
") in new stack
-- Executing [s@allstar-public:3] Set("IAX2/allstar-public-412",
"NODENUM=508425") in new stack
-- Executing [s@allstar-public:4] GotoIf("IAX2/allstar-public-412",
"0?hangit") in new stack
-- Executing [s@allstar-public:5] GotoIf("IAX2/allstar-public-412",
"0?hangit") in new stack
-- Executing [s@allstar-public:6] GotoIf("IAX2/allstar-public-412",
"1?hangit") in new stack
-- Goto (allstar-public,s,14)
-- Executing [s@allstar-public:14]
Answer("IAX2/allstar-public-412", "") in new stack
-- Executing [s@allstar-public:15] Wait("IAX2/allstar-public-412",
"1") in new stack
-- Executing [s@allstar-public:16]
Hangup("IAX2/allstar-public-412", "") in new stack
== Spawn extension (allstar-public, s, 16) exited non-zero on
'IAX2/allstar-public-412'
-- Hungup 'IAX2/allstar-public-412'
linkbox*CLI>
Anyone have a clue what could be happening here?
73
N2DYI
Reply all
Reply to author
Forward
0 new messages