Keyboard Shortcuts
Likes
- Direwolf
- Messages
Search
Re: Crash proofing
Skyler,
Responding through David's msg. - Assuming you're using PTY serial KISS to Xastir, it's unclear toYou don't have to restart Xastir, just go to interface control, select your down interface & click 'Start'. - Monitoring and restarting the various components of your APRSI have some scripts that install Direwolf & AX.25 and use systemd for start up here: The Direwolf script configs for a udrc hat but everything else is independent of that hardware. It also sets up the log & log rotate files for direwolf properly. Start with the README. For current Linux kernels systemd is the right approach for managing your daemons or any background process. Don't use cron. /Basil n7nix |
|||||||||
Re: Crash proofing
¿ªÔÆÌåÓýMe too ... outside of the first two tries to Tx are not correct ... i think thats already fixed in next release. From: direwolf_packet@... <direwolf_packet@...> on behalf of David Ranch dranch@... [direwolf_packet]
Sent: Wednesday, November 23, 2016 9:03 PM To: direwolf_packet@... Subject: Re: [direwolf_packet] Crash proofing ?
?
On 11/23/2016 08:33 PM,
wb2osz@... [direwolf_packet] wrote:
? |
|||||||||
Re: Crash proofing
David Ranch
¿ªÔÆÌåÓýJohn would know best but I've personally had reliable results with the -p option. --David On 11/23/2016 08:33 PM,
wb2osz@... [direwolf_packet] wrote:
? |
|||||||||
Re: Crash proofing
Don't use the "-p" pseudo terminal option if you can avoid it. ?Each time the device number could be different, making it difficult for automatic attachment by some other application. ?It has been problematic in other ways.
Xastir knows how to use the AGW TCPIP socket interface. ? Use that instead. ?User Guide section 5.5.1. ? |
|||||||||
Re: Crash proofing
David Ranch
¿ªÔÆÌåÓýHello Skyler, Giving a reliability report would depend on a huge number of factors.? To your other point, one item crashing and impacting the other components entirely depends on : ?? - You mentioned you're starting Direwolf with the -p option which I assume you're using with a PTY serial interface directly to Xastir.? Yes?? Maybe you're connecting it to the Linux AX.25 stack? ?? - Assuming you're using PTY serial KISS to Xastir, it's unclear to me how Xastir would deal with Direwolf going away and coming back.? Will it retry / re-initialize things? ? You'd have to test that ? ?? - Monitoring and restarting the various components of your APRS stack can be done but it gets complicated.? There are SystemD examples for Direwolf out there but if you're using the Linux AX.25 stack, how would that be managed.? It's also unclear how doing this for Xastir would work since it's a graphical application. The safe bet would be that if a daemon is found missing/crashed/etc, you tear down all other related processes and restart all of the APRS applications again.? You could do this very crudely using cron (that's what the Direwolf documentation recommends), you could try using tools like Monit, or you could try doing this all via SystemD.? Up to you. --David KI6ZHD On 11/23/2016 01:21 PM, Skyler F
electricity440@... [direwolf_packet] wrote:
? |
|||||||||
Crash proofing
I just did an install on a friends vehicle containing a raspberry pi running Direwolf, Xastir, and GPSD (BU-353 GPS). The Pi operates in WiFi Access Point mode for ease of wireless remote desktop to view Xastir window maps. The script I wrote for bootup contains code to start Xastir and Direwolf. A terminal window is open in the background while a GUI window is opened up for Xastir. I have a feeling that this is a lot of different processes that have to go right on bootup, and need to stay online for the duration that the vehicle is online.? If my direwolf -p window stops and then starts up for example, xastir will not recognize packets and have an interface error until I manually press the button to start the interface back up.? Can anybody give me a reliability report of all these applications? Is there a chance one will crash or do something funky and cause the system as a whole to stop? If so, is there a better way to run it, maybe a Daemon that detects crashing and restarts everything? 73 KD0WHB -- Skyler Fennell KD?WHB |
|||||||||
Re: recipe for target 'demod_afsk.o' failed
¿ªÔÆÌåÓýOdd .. i didnt see the previous post ... sorry if this is a repeat ...
did you verify all the dependencies ?
I just did a git pull on master (1.3) ... and was stil up to date and didint have that issue ...
Im using an odroid-X2 Ubuntu 14.04 ArmV8
odroid@odroid:~/Downloads/direwolf$ git pull
odroid@odroid:~/Downloads/direwolf$ git tagAlready up-to-date. 1.0
1.1 1.2 1.3 1.3-beta 1.3-dev-F 1.3-dev-I 1.3-dev-K odroid@odroid:~/Downloads/direwolf$ git branch * master odroid@odroid:~/Downloads/direwolf$ From: direwolf_packet@... on behalf of Skyler F electricity440@... [direwolf_packet]
Sent: Thursday, November 17, 2016 9:18 PM To: direwolf_packet@... Subject: [direwolf_packet] Re: recipe for target 'demod_afsk.o' failed ?
?
For got to say, I am running a raspberry pi with the latest release of jessie.?
On Fri, Nov 18, 2016 at 5:14 AM, Skyler F
<electricity440@...> wrote:
Skyler Fennell
KD?WHB
|
|||||||||
Re: recipe for target 'demod_afsk.o' failed
Precompiled version --?? Then apt-get update apt-get install direwolf On Thu, Nov 17, 2016 at 9:18 PM, Skyler F electricity440@... [direwolf_packet] <direwolf_packet@...> wrote:
--
John D. Hays K7VE PO Box 1223, Edmonds, WA 98020-1223 ![]() |
|||||||||
Re: recipe for target 'demod_afsk.o' failed
For got to say, I am running a raspberry pi with the latest release of jessie.? On Fri, Nov 18, 2016 at 5:14 AM, Skyler F <electricity440@...> wrote:
--
Skyler Fennell KD?WHB |
|||||||||
recipe for target 'demod_afsk.o' failed
I am trying to build direwolf on my raspberry pi 3.? I ran the command make, but got the following error:? make: *** [demod_afsk.o] Error 1 When I ran make again, here it the entire output: Any suggestions? gcc -O3 -pthread -Igeotranz -ffast-math -mfpu=neon -DUSE_ALSA -DENABLE_GPSD -DUSE_HAMLIB ? -c -o demod_afsk.o demod_afsk.c In file included from demod_afsk.c:51:0: demod_afsk.c: In function ¡®demod_afsk_process_sample¡¯: demod_afsk.c:884:28: error: ¡®FFF_PROFILE¡¯ undeclared (first use in this function) ? if (D->profile == toupper(FFF_PROFILE)) { ? ? ? ? ? ? ? ? ? ? ? ? ? ? ^ demod_afsk.c:884:28: note: each undeclared identifier is reported only once for each function it appears in : recipe for target 'demod_afsk.o' failed make: *** [demod_afsk.o] Error 1 Thanks, Skyler KD0WHB Skyler Fennell KD?WHB |
|||||||||
Re: Rx only iGate - stop Tx data
On Thu, Nov 17, 2016 at 1:08 PM, kq6di@... [direwolf_packet]?<direwolf_packet@...>?wrote:
Unless they're using the 0N 0W null position... I think the most correct way to implement the original objective is to use the send-only UDP port, where every packet is injected into APRS-IS as a UDP packet containing "[APRS-IS login credentials]\r\n[Packet contents]" I don't know what the exact bandwidth trade-off would be for adding the APRS-IS login on every packet vs not having an active TCP connection. Probably pretty good considering you no longer have the APRS-IS keep-alive lines either. I have no idea if Direwolf supports UDP mode, so I may be suggesting a solution which isn't possible. -- Kenneth Finnegan |
|||||||||
Re: Rx only iGate - stop Tx data
Fred, I'll give it a try. ?The worst I could do is break something I'm testing, so it's also easy to fix. Thanks, Arnold KQ6DI From: "'Fred Hillhouse' fmhillhouse@... [direwolf_packet]" To: "direwolf packet" Sent: Thursday, November 17, 2016 12:50:02 PM Subject: RE: [direwolf_packet] Rx only iGate - stop Tx data A filter like this might help. I think it won¡¯t receive anything that has TCPIP or TCPXX in the packet. I don¡¯t know it messages will pass if you are looking to pass them. Maybe there is a filter for the Q construct that might work better. [Not my information, just something I saw recently] ? FILTER IG 0 ! d/TCPIP/TCPXX ? Best regards, Fred N7FMH ? ? From: direwolf_packet@... [mailto:direwolf_packet@...] ? ? [Attachment(s) from kq6di@... included below] Yes, I have a RX-only IGate. ?Below is a snip (with the receiver off) of a bunch of data coming from the internet that wants to be transmitted. ?It's about 5 minutes worth. ?I realize it isn't a lot, but I intend to locate the RX-only IGate & digipeater in a location that has rather limited internet capabilities. ?Is there a filter to hopefully send to the server that will reduce this some? Believe it or not, this entire installation is solar powered, so even transmitting each packet requires power from somewhere... ? Arnold KQ6DI ? ? From: "Jim Alles kb3tbx@... [direwolf_packet]" <direwolf_packet@...> ? ? Arnold, ? It seems you want to undertake a RX-only IGate. ? AFAIK, it doesn't matter if you are using direwolf or anything else, if you are directly connected to an APRS-IS server TCP port 14580, you will be receiving packets _because_ you sent packets to the server. This is by design, to support messaging. ? The field called 'filter' that is sent to the server, is not really a filter, it is a subscription for additional packets to what I just described, above. ? The only alternative (theoretically) is to setup your own server, and modify the code to do what you want, and point your IGate to that, but there may be options I don't know about. ? Explain what you are trying to accomplish, and describe your local RF territory a bit. ? There may be direwolf config parameters that can get the incoming packet count down significantly, yet I am sorry that is not what you are looking for. ? 73, Jim A. ? On Thu, Nov 17, 2016 at 1:37 PM, kq6di@... [direwolf_packet] <direwolf_packet@...> wrote: ? ? I have a receive only IGate using DireWolf.? I would like to stop ALL traffic coming from the IS server that wants to be transmitted, but I don't know the best parameter(s) to enter.? Suggestions? ? I know it isn't much data, but I have a future location that will only have internet via satellite, and every little bit of data counts toward (against) the monthly limit. ? Thanks. Arnold KQ6DI ? ? ?
|
|||||||||
Re: Rx only iGate - stop Tx data
Jim, I already have a filter on what goes to APRS-IS, and that works well. ? Note to everyone, 50 km is a decent range, but definitely anything beyond 300 km is completely bogus. ?Just about everyone receives something from thousands of miles away from defective trackers, so a filter on the To IS is good. Arnold KQ6DI From: "Jim Alles kb3tbx@... [direwolf_packet]" To: "direwolf packet" Sent: Thursday, November 17, 2016 12:16:24 PM Subject: Re: [direwolf_packet] Rx only iGate - stop Tx data Yes, there are two issues here. I have recently been testing Direwolf for this. you can limit what you send to the APRS-IS, by putting in a filter like this in the config file: "FILTER 0 IG ? r/40.8/-77.8/50 ?| t/m " APRS-IS will only receive position packets within 50 km of some point near me, and messsages that might come from anywhere. This will trap the stray packets that you might get by ducting, for example. That will minimize what is sent to direwolf to decide what to do with messages. In my experience, you can also reduce the incoming data flow but logging into an APRS-IS server that uses aprsc software - it seems to be more conservative. YMMV. 73,? Jim A. On Thu, Nov 17, 2016 at 2:33 PM, kq6di@... [direwolf_packet] <direwolf_packet@...> wrote:
|
|||||||||
Re: Rx only iGate - stop Tx data [1 Attachment]
¿ªÔÆÌåÓýA filter like this might help. I think it won¡¯t receive anything that has TCPIP or TCPXX in the packet. I don¡¯t know it messages will pass if you are looking to pass them. Maybe there is a filter for the Q construct that might work better. [Not my information, just something I saw recently] ? FILTER IG 0 ! d/TCPIP/TCPXX ? Best regards, Fred N7FMH ? ? From: direwolf_packet@... [mailto:direwolf_packet@...]
Sent: Thursday, November 17, 2016 2:33 PM To: direwolf packet Subject: Re: [direwolf_packet] Rx only iGate - stop Tx data [1 Attachment] ? ? [Attachment(s) from kq6di@... included below] Yes, I have a RX-only IGate. ?Below is a snip (with the receiver off) of a bunch of data coming from the internet that wants to be transmitted. ?It's about 5 minutes worth. ?I realize it isn't a lot, but I intend to locate the RX-only IGate & digipeater in a location that has rather limited internet capabilities. ?Is there a filter to hopefully send to the server that will reduce this some? Believe it or not, this entire installation is solar powered, so even transmitting each packet requires power from somewhere... ? Arnold KQ6DI ? ? From: "Jim Alles kb3tbx@... [direwolf_packet]" <direwolf_packet@...> ? ? Arnold, ? It seems you want to undertake a RX-only IGate. ? AFAIK, it doesn't matter if you are using direwolf or anything else, if you are directly connected to an APRS-IS server TCP port 14580, you will be receiving packets _because_ you sent packets to the server. This is by design, to support messaging. ? The field called 'filter' that is sent to the server, is not really a filter, it is a subscription for additional packets to what I just described, above. ? The only alternative (theoretically) is to setup your own server, and modify the code to do what you want, and point your IGate to that, but there may be options I don't know about. ? Explain what you are trying to accomplish, and describe your local RF territory a bit. ? There may be direwolf config parameters that can get the incoming packet count down significantly, yet I am sorry that is not what you are looking for. ? 73, Jim A. ? On Thu, Nov 17, 2016 at 1:37 PM, kq6di@... [direwolf_packet] <direwolf_packet@...> wrote: ? ? I have a receive only IGate using DireWolf.? I would like to stop ALL traffic coming from the IS server that wants to be transmitted, but I don't know the best parameter(s) to enter.? Suggestions? ? I know it isn't much data, but I have a future location that will only have internet via satellite, and every little bit of data counts toward (against) the monthly limit. ? Thanks. Arnold KQ6DI ? ? ? |
|||||||||
Re: Rx only iGate - stop Tx data [1 Attachment]
Yes, there are two issues here. I have recently been testing Direwolf for this. you can limit what you send to the APRS-IS, by putting in a filter like this in the config file: "FILTER 0 IG ? r/40.8/-77.8/50 ?| t/m " APRS-IS will only receive position packets within 50 km of some point near me, and messsages that might come from anywhere. This will trap the stray packets that you might get by ducting, for example. That will minimize what is sent to direwolf to decide what to do with messages. In my experience, you can also reduce the incoming data flow but logging into an APRS-IS server that uses aprsc software - it seems to be more conservative. YMMV. 73,? Jim A. On Thu, Nov 17, 2016 at 2:33 PM, kq6di@... [direwolf_packet] <direwolf_packet@...
|
|||||||||
Re: Rx only iGate - stop Tx data
Yes, I have a RX-only IGate. ?Below is a snip (with the receiver off) of a bunch of data coming from the internet that wants to be transmitted. ?It's about 5 minutes worth. ?I realize it isn't a lot, but I intend to locate the RX-only IGate & digipeater in a location that has rather limited internet capabilities. ?Is there a filter to hopefully send to the server that will reduce this some? Believe it or not, this entire installation is solar powered, so even transmitting each packet requires power from somewhere... Arnold KQ6DI From: "Jim Alles kb3tbx@... [direwolf_packet]" To: "direwolf packet" Sent: Thursday, November 17, 2016 10:59:11 AM Subject: Re: [direwolf_packet] Rx only iGate - stop Tx data Arnold, It seems you want to undertake a RX-only IGate. AFAIK, it doesn't matter if you are using direwolf or anything else, if you are directly connected to an APRS-IS server TCP port 14580, you will be receiving packets _because_ you sent packets to the server. This is by design, to support messaging. The field called 'filter' that is sent to the server, is not really a filter, it is a subscription for additional packets to what I just described, above. The only alternative (theoretically) is to setup your own server, and modify the code to do what you want, and point your IGate to that, but there may be options I don't know about. Explain what you are trying to accomplish, and describe your local RF territory a bit. There may be direwolf config parameters that can get the incoming packet count down significantly, yet I am sorry that is not what you are looking for. 73, Jim A. On Thu, Nov 17, 2016 at 1:37 PM, kq6di@... [direwolf_packet] <direwolf_packet@...> wrote:
|
|||||||||
Re: Rx only iGate - stop Tx data
Arnold, It seems you want to undertake a RX-only IGate. AFAIK, it doesn't matter if you are using direwolf or anything else, if you are directly connected to an APRS-IS server TCP port 14580, you will be receiving packets _because_ you sent packets to the server. This is by design, to support messaging. The field called 'filter' that is sent to the server, is not really a filter, it is a subscription for additional packets to what I just described, above. The only alternative (theoretically) is to setup your own server, and modify the code to do what you want, and point your IGate to that, but there may be options I don't know about. Explain what you are trying to accomplish, and describe your local RF territory a bit. There may be direwolf config parameters that can get the incoming packet count down significantly, yet I am sorry that is not what you are looking for. 73, Jim A. On Thu, Nov 17, 2016 at 1:37 PM, kq6di@... [direwolf_packet] <direwolf_packet@...> wrote:
|
|||||||||
Rx only iGate - stop Tx data
I have a receive only IGate using DireWolf. ?I would like to stop ALL traffic coming from the IS server that wants to be transmitted, but I don't know the best parameter(s) to enter. ?Suggestions? I know it isn't much data, but I have a future location that will only have internet via satellite, and every little bit of data counts toward (against) the monthly limit. Thanks. Arnold KQ6DI |
|||||||||