I’m all of a sudden having a lot of problems lately…
I have a node that has been working perfectly for weeks on a custom non port 4569
configuration.
I have not made any changes that should affect this aside of I have ran updates recently and think this may have something to do with it.
Everything is properly configured on my custom port and I register with allstarlink using this custom port.
But allstarlink is reporting to the world that I am on port 4569 and I an not.
What could I possibly be missing?
I think I have checked and double checked all of the obvious things on my end.
Also checked with and without static NAT port mapping in place.
It had been working for a long time.
Doing a DNS lookup on it returns port 4569 and this is causing obvious problems for people trying to connect.
Thanks!
-Steve
dig 2586.nodes.allstarlink.org TXT
-there is no stated way to specify the port (here).
I DO have this specified correctly in my allstarlink.org account “server” portion-
and this node is associated to that “server”.
This has been working fine as it is until now.
I am only guessing that the HTTP registration process either picks up on the bindport setting in iax.conf -OR it uses the server/node specified port which is set in the allstarlink account as mentioned above
If you could clarify this (which method is used to tell allstarlink.org the port) that would be great.
Meanwhile I do still have this problem.
My friends can’t connect because the DNS/Node lookup is providing the incorrect default port.
Thanks!
-Steve
Indeed it is.
Also for my own understanding can you confirm that the http registration process
simply checks the bindport setting in iax.conf when it updates the node.
Or some other source?
Thanks!
The registration process does not affect the UDP port registration regardless of using HTTP or IAX. It’s driven by the server configuration. If unset, the default of UDP/4569 is used. Otherwise the value set there is used for the node database. I would like to make that more dynamic at some point. But that requires a big change to the registration server code.
The problem this morning was caused by a code bug that was using the wrong column in a database to populate the udp port and I didn’t catch that during testing. I was kinda in a rush to get the final code done and cut over to it because the interim solution I rolled out a few months ago wasn’t scaling as I’d hoped and as causing some intermittent problems.
Oh cool, good to know how it works! Thank you.
I’m having another issue this evening.
Node 27298 is registering fine and allstarlink shows it’s registered.
But it is showing up in allstarlink DNS as the old IP address it was on hours ago and will not update. (I moved it to a different Internet connection)
I’ve tried rebooting and such to make sure it’s re-registering outside of the periodic timer.
Still stuck at the old IP address. Per DNS.
Right now I’m trying shutting it down for 10 minutes or so and waiting for it’s registration status to show unregistered at allstarlink.org then will fire it back up.
Looks like DNS update might be stuck for some reason.
I was also thinking to open a separate post/topic on this.
Still very broken.
I suspect something on the backend is not syncing.
I left the node off for an hour or so.
Powered it back up it is registering using a Verizon connection.
Still showing my old IP and port from the completely different Comcast connection it was using 4-hours ago.
dig @ns-436.awsdns-54.com27298.nodes.allstarlink.org txt
; <<>> DiG 9.11.4-P2-RedHat-9.11.4-9.P2.el7 <<>> @ns-436.awsdns-54.com27298.nodes.allstarlink.org txt
; (1 server found)
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 48224
;; flags: qr aa rd; QUERY: 1, ANSWER: 1, AUTHORITY: 4, ADDITIONAL: 1
;; WARNING: recursion requested but not available
I opened a new Topic over here.
Unless I am terribly mistaken or being really stupid, I think I am seeing a major service affecting problem with the backend.
Was fixed for a few minutes but problem came right back.
Follow the new thread for any updates from me if you want to.
I’ll stop updating this one as it has went off topic.
I suspect it’s related to the new changes and being ironed out.
I was kinda in a rush to get the final code done and cut over to it because the interim solution I rolled out a few months ago wasn’t scaling as I’d hoped and as causing some intermittent problems.