- App_rpt-users - lists.keekles.org

Node becomes unregistered after many weeks (no reply from registration server)
by Mike Hugo 21 Jan '15

21 Jan '15

I've seen this problem on various nodes now for a while and finally had a minute to actually grab ethernet packets from the WAN side of the firewall to see if I could dig deeper to figure out what is breaking after sometime. This is not a major problem since a reboot fixes it but more of a curiosity. I tried to search the forums to see if this had been asked before but didn't find anything of help. The issue is that after a node sits unused for a long time it becomes unregistered. I'm not sure how long a long time is but it is in weeks. A reboot of the box always bring is back. Today I went looking on the WAN side of the firewall and I can see the box sending IAX2 packets out to the registration server but the registration server never replying. I used the WAN side so as to rule out local network issues. As far as I can tell the IAX registration UDP packet looks fine to me. The contents of the failed one look the same to a successful one. It appears to be sending the REGREQ message and waiting 2 seconds before resending the same REGREQ message again and waiting 10 seconds and repeating once more after 10 seconds. Finally waiting 28 seconds and starting again with a new REGREG message with a new Source Call. When it works the REGREQ is replied to from the registration server with a CALL TOKEN, etc and the rest of the IAX handshake and authentication happens. Not sure if helps in figuring this out but when the node is in this state the call to get nodes.pl is rejected. Call to uhandler are replied with "OK!". Since it takes a long time to reproduce this issue I'm wondering if I should be looking for something else the next time this happens. The fact that a reboot always bring it back implies that there is something else on the local node that is causing this to break and the registration server to ignore the IAX requests on purpose. Mike

2 2

Re: [App_rpt-users] More than 4 nodes on a PC
by Bryan Fields 20 Jan '15

20 Jan '15

On 1/20/15, 1:30 PM, Cole Cunningham wrote: > We’re looking at app_rpt for sites with up to eight radios, all currently > controlled by a RLC-3. Our initial approach is to use a PC and try to run all > eight radios off of it. For testing we’re using modified fobs running on USB > 2.0 ports and when we try to use more than four, things break up, sound bad, > etc. I’ve found that CentOS 5 that comes on the ACID disk doesn’t support USB > 3.0, which we were thinking might offer more bandwidth and handle the added fobs. The usb2 devices on the usb3 bus will slow the bus to usb2 speeds (assuming it's the first connected device, no way to control this upon cold boot). So you'll not gain much. If you have squelch/COR control in the radios, two of the quad pci cards would work well. You can still find motherboards with pci slots, or use a converter to pcie. The cards are under $75/port, it's hard to go wrong with this. The other option if you don't have squelch is RTCM's all feeding over IP to the server. That way the server would not need anything special (other than a multi-core cpu) The USB FOB's are nice, but they don't scale nicely (IMO). As to getting the Allstar version of asterisk working on a newer distro, I believe all you need to do for this is use dahdi and the dahdi patches and then compile from source. I have it running under debian squeeze with out dahdi, but I plan to build another box running wheezy and try the dahdi patches. Can anyone correct me if I'm wrong (or a +1 if I'm right) about this? -- Bryan Fields 727-409-1194 - Voice 727-214-2508 - Fax http://bryanfields.net

2 1

USB fob to Baofeng cable
by geoffrey mendelson 20 Jan '15

20 Jan '15

Anyone have a suggestion for the wiring of a USB fob to Baofeng radio cable? TIA. Geoff. -- Geoffrey S. Mendelson 4X1GM/N3OWJ Jerusalem Israel.

2 1

More than 4 nodes on a PC
by Cole Cunningham 20 Jan '15

20 Jan '15

We're looking at app_rpt for sites with up to eight radios, all currently controlled by a RLC-3. Our initial approach is to use a PC and try to run all eight radios off of it. For testing we're using modified fobs running on USB 2.0 ports and when we try to use more than four, things break up, sound bad, etc. I've found that CentOS 5 that comes on the ACID disk doesn't support USB 3.0, which we were thinking might offer more bandwidth and handle the added fobs. Has anyone used this PC based approach for 4+ nodes using USB fobs? Or could I get some suggestions as to how to approach the implementation of eight nodes without using fobs or the discontinued 4 port PC card? Or is it a viable solution to use a BBB for every radio and connect them all with Ethernet switches and treat each BBB like a separate server. Any ideas are appreciated Cole AA7RD

3 4

New Thin Client Centos Loading Lock Up
by Chris Hill - K7AZ 19 Jan '15

19 Jan '15

I am attempting to load Centos on a pair of 5740 HP Thin Clients. The load from ISO image CD goes well, but after the restart/reboot, the loading hangs up (just shows a curser) after the following lines: Starting auditd: [OK] Starting system logger: [OK] Starting kernel logger: [OK[ ___ (blinking underscore here) Rebooting and it goes through start up and hangs at same place…… Any suggestions would be very welcome at this point. Thank you, Chris K7AZ

2 1

Oops
by Steve Zingman 19 Jan '15

19 Jan '15

Sorry about the TV post. Not paying attention I guess 73, Steve N4IRS

1 0

changing funcchar in rpt.conf messed up echolink
by Elden Fenison 18 Jan '15

18 Jan '15

I thought I would change the command prefix of my private simplex node to be D rather than *. Using the funcchar parm in rpt.conf seemed to work in general. But for some reason echolink would no longer complete outbound connections. I'm betting that there is a separate parameter for the echolink funcchar value. However I wasn't able to find it documented. Any ideas?

2 2

Get the same USB port assigned to a node at startup each time?
by Cole Cunningham 18 Jan '15

18 Jan '15

I'm a pure newbi to Asterisk/app_rpt and am trying to learn enough to see if it will work for an RLC-3 replacement for my club. I have four nodes in test mode running on one server using USB FOBs and every time Asterisk starts the four USB FOBs get assigned to different nodes than they were on when Asterisk was shut down. Lsusb shows the FOBs are split with two on each internal USB bus, corresponding to the hardware USB controllers. I get good audio out of all four FOBs every time, just from different nodes than last time. The PC is a home built Intel G2130 of fairly recent vintage. Is there a way to tie a USB hardware port to a particular node so I get the same node on the same hardware every time? This is particularly critical in case of power failure so the site can come back up without operator intervention. Thanks, Cole, AA7RD

3 5

12 monkeys
by Steve Zingman 18 Jan '15

18 Jan '15

Just watched the first episode. Pretty good. Any, yes, I have seen the original before.

1 0

System Interaction
by Mike M 18 Jan '15

18 Jan '15

I've been trying to understand how all the pieces fit together. I'm more of a visual learner so I've been drawing this diagram. Anyone care to comment/correct/complete it? I'm interested in completely understanding the "flow" of how an incoming connection is routed out to a particular USB port, as well as what configuration parameters/files are involved. I've attached it but I think this list strips attachments so you can get it here. https://drive.google.com/open?id=0BynhvyBxwd-IWEJkQnVWRE5wWW8&authuser=0 On a side note.. One thing I've see in extensions.conf is exten => ${NODE},1,rpt,${NODE} This would seem like it would keep you from having to write lines for every extension on the local server but things stop working if I just have that line in my radio-secure staza. Is there a secret to this? Thanks! Regards, Mike Montalvo, KF7MBK

1 0