[arm-allstar] Permanent Allstar Connections

Doug Crompton wa3dsp at gmail.com
Fri Feb 24 00:24:32 EST 2017


We have gotten a couple of messages recently relating to permanent
connections and problems with them not working properly. We are looking at
the code and have not been able to duplicate this as yet but we are
continuing to look at the problem. It is possible this is a long standing
Allstar problem that only shows up in certain scenarios. If anyone is
having this or any problem it is important to document the exact scenario
in which it is happening. Giving more information makes it easier to track
down a problem.

Also VERY IMPORTANT - since we have had an influx of new users since the
V1.5 release please be aware that since the beginning hamvoip images have
used *73xxxxx and *71xxxxx are the permanent connect and permanent
disconnect codes and *76 as the disconnect all code. This is different than
the ACID release. So make sure you are actually doing a permanent connect
before reporting a problem with it. Also if for some reason the 7 was
missed in the DTMF *73xxxxx sequence it would be a normal connect or
disconnect. The entire function list is documented in the functions howto
on the hamvoip.org web page.

If you make a permanent connect to another node it means that if anything
between you and that node including that node rebooting causes a circuit
failure your system will try a reconnect. This reconnect essentially goes
on forever until it gets connected or you do a permanent disconnect
(*71xxxxx) of that node, you reboot or restart Asterisk or that node forces
a disconnect from its end. Permanent connects are often used on circuits
that you want to stay connected. A normal *3 connect will fail if anything
along the path is disrupted even for a relatively short period of time.
While this does not happen often it does happen especially if a connection
is left up for an extended period or it is a marginal circuit.


*73 Doug*

*WA3DSP*

*http://www.crompton.com/hamradio <http://www.crompton.com/hamradio>*


More information about the arm-allstar mailing list