Random "connection failed"

Setup Information

********** AllStarLink [ASL] Version Info **********

OS : Debian GNU/Linux 13 (trixie)
OS Kernel : 6.18.34+rpt-rpi-2712

Asterisk : 22.9.0+asl3-3.9.3-1.deb13
ASL [app_rpt] : 3.9.3

Inquiry

I have two Allstar nodes. One works flawlessly. The other has random connection failures. I have been trying to troubleshoot this for weeks with no luck.

Call accepted by 69.164.197.28:4569 (format ulaw)
-- Format for call is (ulaw)
-- Hungup 'IAX2/69.164.197.28:4569-3790'

I can wait a few minutes and the connection will work. This is is not specific to any one node I am connecting too and is totally random.

Also, when connecting the two nodes together. From the working node to the problem node; It shows to be connected but the connection does not appear on the bubble chart. When connecting from the problem node to the working node; the connection does show on the bubble chart.

I have checked and rechecked my configurations. The nodes are on separate ports. I have replaced my router, but the replacement cloned the old one, so I have not completely ruled out a router problem.

At this point I am wondering if I should just abandon the problem node and create a new one. Ideas on how to troubleshoot this? Thanks.

Are the two systems on the same local network? If so, have you looked at Multiple Nodes on the Same Network - AllStarLink Manual?

Yes, same network. Yes, I have followed the instructions for multiple nodes on the same network and they are on separate ports.

After a long conversation with the help desk for my router, the known issues were resolved and I have been able to rule out the router as the source of the problem.

Another copy of the log. Notice this time it shows port 4570, which is the correct port for this node. Five minutes before this I connected to 5 different nodes including this one with no issues.

Call accepted by 74.196.158.135:4570 (format ulaw)
-- Format for call is (ulaw)
-- Hungup 'IAX2/74.196.158.135:4570-15629'

-- Call accepted by 198.58.124.150:4569 (format ulaw)
-- Format for call is (ulaw)
-- Hungup 'IAX2/198.58.124.150:4569-11915'

Sounds like the other node doesn't like you. Either it thinks you are not registered, it has a whitelist (probably not that) or a codec mismatch is happening.

From asterisk CLI, if you type

iax2 set debug on

and then connect to that node, it should at least give you a code relative to the state when it drops, but you may not no more without seeing the log on the other node.

Immediately after one of these failures, does everything look OK when you run sudo asl-node-auth-check?

sudo asl-node-auth-check

Comes back with a result of "No problems detected with node 49406"

I agree it just does not like me? Other than hardware, the only major difference between my two nodes it that the problem unit has dvswitch installed.

debug log

-- Call accepted by 198.58.124.150:4569 (format ulaw)
-- Format for call is (ulaw)
Tx-Frame Retry[-01] -- OSeqno: 001 ISeqno: 001 Type: IAX Subclass: ACK
Timestamp: 00011ms SCall: 16155 DCall: 09093 198.58.124.150:4569
Rx-Frame Retry[ No] -- OSeqno: 001 ISeqno: 001 Type: IAX Subclass: HANGUP
Timestamp: 00046ms SCall: 09093 DCall: 16155 198.58.124.150:4569
CAUSE CODE : 0

iax2 set debug on

-- Call accepted by 69.164.197.28:4569 (format ulaw)
-- Format for call is (ulaw)
Tx-Frame Retry[-01] -- OSeqno: 001 ISeqno: 001 Type: IAX Subclass: ACK
Timestamp: 00007ms SCall: 07569 DCall: 05275 69.164.197.28:4569
Rx-Frame Retry[ No] -- OSeqno: 001 ISeqno: 001 Type: IAX Subclass: HANGUP
Timestamp: 00139ms SCall: 05275 DCall: 07569 69.164.197.28:4569
CAUSE CODE : 0

Is the node you're trying to connect to an old HamVOIP node? We've seen that from time to time with no explanation and no one in a position to debug why the HamVOIP node is sending HANGUP.

One of the nodes was upgraded, I honestly don’t remember which one. I’m going to start over with a fresh install (yay) and hopefully that will resolve my issue.

I have a question for when I rebuild the node. Should I start over completely from scratch? If I import a backup image, will the gremlins return? I feel like using a backup copy would be a bad idea, but it sure would save some time.

Are you running multiple nodes on one server, or one node on multiple servers?

Both nodes typically run on one server. The reliable unit is in my car and I have a few different ways to connect to the internet when mobile. I can remote into the desk unit when I'm out, and the problem persists.

I installed a fresh version of ASL3. along with Supermon, Allscan, and Node Remote. The hangup issue persists. At this point I am thinking I have a hardware problem. This is running on a Amazon Aursinc Shari Pi which does not have the best reputation.

I believe I have finally resolved the issue. There were multiple errors in the extensions.conf file. Fingers crossed, no hang ups for a couple of hours. Connecting the nodes together only worked one way. Troubleshooting that revealed the errors. AI is a big help once you ask the right questions.