TwoToneDetect71 Released

[aaknitt]

I am also getting TTD crashing after a certain period of time. When I start TTD, I get two process running in task manager. Is this correct or is there something wrong. I generally run two instances of Two Tone Detect, when I start the second instance, two more processes show in task manager? I would run headless, but I prefer to see the screen(s) on my monitor to make sure it is running. I would run the script to check the status, but I do not want to reboot, as I have other programs running that I do not want to stop without user input. I would prefer an email letting me know it has crashed. I do not get duplicate notifications.

Yeah now that I look closer it is spawning two processes for each instance...one using more resources than the other. Not sure what that's about, but it seems normal. Does minimizing the GUI seem to help with reducing crashes for you? Even if you run headless you can still access the GUI via web browser to check status.

Andy

 

[johnmac]

Yeah now that I look closer it is spawning two processes for each instance...one using more resources than the other. Not sure what that's about, but it seems normal. Does minimizing the GUI seem to help with reducing crashes for you? Even if you run headless you can still access the GUI via web browser to check status.

Andy

I just put the two instances minimized. Will see how that does, may take a week or more to see if it crashes or not. If I go headless, what address do I put in chrome to view the GUI?

John

 

[aaknitt]

I just put the two instances minimized. Will see how that does, may take a week or more to see if it crashes or not. If I go headless, what address do I put in chrome to view the GUI?

John

If you're using all of the default settings, it would be http://localhost:8080/index.html

Andy

 

[Kingmobileaudio]

If you're using all of the default settings, it would be http://localhost:8080/index.html

Andy

I tried this and also ip:8080/index.html and get an error page.


Sent from my iPhone using Tapatalk Pro

 

[aaknitt]

I tried this and also ip:8080/index.html and get an error page.


Sent from my iPhone using Tapatalk Pro

Do you have the "Allow Remote Access" box checked? If so, take a look at the "Remote Access Port". If it's not 8080, you'll need to use that number instead of 8080 in your URL.

I also just noticed a couple bugs that cause the "Remote Access Port" input box to be disabled when it shouldn't be. You can fiddle with the "Start Headless" and "Allow Remote Access" check boxes to get it enabled to change it if needed until I get it fixed in the next release.

Andy

 

[MikeOrlando02]

Andy,

Working on a new TTD monitoring script, heartbeat is going to be really useful so thank you for adding that. I wanted to better understand how it works. It appears that it doesn't run based on clock as it doesn't update on the minute. Can you confirm it is a 60 second timer that starts when TTD opens and doesn't correlate with clock/time?

Also, what is in the heartbeat.txt? It appears just to be a second counter. Is there a date encoded in there that I am not seeing? Does the number ever reset or just grow longer?


Thanks,
Mike

 

[DC31]

Hi @MikeOrlando02 good job on the scripts. I am not a Windows person and run my TTD’s on the Raspberry Pi platforms. Just for informational purposes, let me give a quick description of how I monitor my two Pi’s, each running two instances of TTD. I don’t know if some of this migh5 also apply to Windows installations.

The pi is running the full Raspbian Buster OS with gui. The two instances of TTD are running headless. There is a program included with the OS called Node-Red (nodered.org) used for wiring the internet of things. There is a Watch node that watches a file for changes. I have this set to watch the heartbeat.log file. Every time this files changes (60 sec) it triggers a flow to an MQTT Output node.

MQTT is a publish/subscribe protocol. I run a MQTT server (mosquitto.org) on another pi at my house where the output from above is published.

The pi at my house also runs node-red and subscribes to the MQTT server using an MQTT Input node. This then flows to a Trigger node set up to trigger a Pushover Notification only if it DOES NOT receive input within 90 sec.


Yes, the heartbeat signal available from v71 is what makes this all possible and relatively easy to set up.

All my setup was drag and drop in node-red with the exception of setting the mqtt server up. I think that could also be done in node-red, but I already had that running.

 

[xxbryantd]

Not sure if its just me or other are having this issue on v71, but I cant seem to get the email body or subject to stay changed even saving and restarting the program. Also seems that I cant find the options in the config file to attempt to change manually. Also Im having issues with the email sending

 

[aaknitt]

Andy,

Working on a new TTD monitoring script, heartbeat is going to be really useful so thank you for adding that. I wanted to better understand how it works. It appears that it doesn't run based on clock as it doesn't update on the minute. Can you confirm it is a 60 second timer that starts when TTD opens and doesn't correlate with clock/time?

Also, what is in the heartbeat.txt? It appears just to be a second counter. Is there a date encoded in there that I am not seeing? Does the number ever reset or just grow longer?


Thanks,
Mike

That's correct, it updates once per minute from the time TTD started. The value in the file is a UNIX timestamp of when it last updated.

Andy

 

[aaknitt]

Not sure if its just me or other are having this issue on v71, but I cant seem to get the email body or subject to stay changed even saving and restarting the program. Also seems that I cant find the options in the config file to attempt to change manually. Also Im having issues with the email sending

It looks like you may have something corrupt in your tones.cfg file. I'd suggest manually editing it in a text file to clean it up and remove all of the comments (comments start with a semicolon).

Andy

 

[xxbryantd]

Maybe this might help. This is what shows up when hitting save

window closed or reloaded,checking for exit criteria in 2 seconds
[(u'index.html', <geventwebsocket.websocket.WebSocket object at 0x0509F458>)]
{u'Squad1': {u'pushbullet_channel': u'', u'text_emails': u'', u'pushover_alert_app_token': u'', u'btone': u'1433.4', u'atone': u'1153.4', u'exclude_to': u'', u'playback_during_record': 0, u'pushover_record_group_key': u'', u'amr_emails': u'"Bryant"<@mms.att.net>', u'gaplength': u'', u'btonelength': u'.3', u'ignore_after': u'60', u'exclude_from': u'', u'description': u'Highland Fire', u'mp3_emails': u'', u'record_delay': u'4', u'__name__': u'Squad1', u'post_email_command': u'', u'pushover_record_app_token': u'', u'pushover_alert_group_key': u'', u'email_body': u'Squad 1', u'atonelength': u'.1', u'pushbullet_api_key': u'', u'email_subject': u'[d] Page Received at [t]', u'alert_command': u''}}
Traceback (most recent call last):
  File "C:\Python27\lib\site-packages\gevent\greenlet.py", line 536, in run
  File "C:\Python27\lib\site-packages\eel\__init__.py", line 195, in _process_message
  File "C:\Users\Knitt\Google Drive\Python\TwoToneDetect71c.py", line 785, in receive_tones_json
ValueError: invalid literal for int() with base 10: ''
Mon Jul 15 12:22:16 2019 <Greenlet at 0x500a3f0: _process_message({u'args': [u'{"Squad1":{"__name__":"Squad1","atone, <geventwebsocket.websocket.WebSocket object at 0x0)> failed with ValueError

 

[aaknitt]

Maybe this might help. This is what shows up when hitting save

window closed or reloaded,checking for exit criteria in 2 seconds
[(u'index.html', <geventwebsocket.websocket.WebSocket object at 0x0509F458>)]
{u'Squad1': {u'pushbullet_channel': u'', u'text_emails': u'', u'pushover_alert_app_token': u'', u'btone': u'1433.4', u'atone': u'1153.4', u'exclude_to': u'', u'playback_during_record': 0, u'pushover_record_group_key': u'', u'amr_emails': u'"Bryant"<0@mms.att.net>', u'gaplength': u'', u'btonelength': u'.3', u'ignore_after': u'60', u'exclude_from': u'', u'description': u'Highland Fire', u'mp3_emails': u'', u'record_delay': u'4', u'__name__': u'Squad1', u'post_email_command': u'', u'pushover_record_app_token': u'', u'pushover_alert_group_key': u'', u'email_body': u'Squad 1', u'atonelength': u'.1', u'pushbullet_api_key': u'', u'email_subject': u'[d] Page Received at [t]', u'alert_command': u''}}
Traceback (most recent call last):
  File "C:\Python27\lib\site-packages\gevent\greenlet.py", line 536, in run
  File "C:\Python27\lib\site-packages\eel\__init__.py", line 195, in _process_message
  File "C:\Users\Knitt\Google Drive\Python\TwoToneDetect71c.py", line 785, in receive_tones_json
ValueError: invalid literal for int() with base 10: ''
Mon Jul 15 12:22:16 2019 <Greenlet at 0x500a3f0: _process_message({u'args': [u'{"Squad1":{"__name__":"Squad1","atone, <geventwebsocket.websocket.WebSocket object at 0x0)> failed with ValueError

It looks like you may have something corrupt in your tones.cfg file. I'd suggest manually editing it in a text file to clean it up and remove all of the comments (comments start with a semicolon).

Andy

 

[Forts]

I'm also still seeing 2 entries in task manager but I'm no longer getting duplicate notifications, so I think that was a fluke. Since I've switched to headless I haven't had any crashes/lockups as of yet, but I hear ya... I liked having the GUI open too to keep an eye on things.

 

[MikeOrlando02]

That's correct, it updates once per minute from the time TTD started. The value in the file is a UNIX timestamp of when it last updated.

Andy

Got it thanks, I think I will just compare current time stamp from when it last was checked, if it changed TTD is running. If it is the same, it is not. More to come.

 

[xxbryantd]

It looks like you may have something corrupt in your tones.cfg file. I'd suggest manually editing it in a text file to clean it up and remove all of the comments (comments start with a semicolon).

Andy

Got it fixed, thanks. Iv noticed though when trying leave the "Email body" blank like I did with previous messages, it will send a blank box for that spot instead of not removing it from the mms text

 

[webstar22]

Anyone having issues with 71c crashing/locking up? I'm seeing it almost weekly, sometimes more frequently than that. Either it's completely frozen or Chrome itself has crashed and just shows the "oh snap" crash screen. I've also noticed during these situations that according to task manager Chrome is using ignorant amounts of ram. This is on a PC dedicated to TTD, Unitrunker and DSD+, there isn't any everyday browsing done on this machine.

So... this becomes a problem when it's crashed and I'm unaware. Is there a way to close and restart the program daily or something like that? At least then my downtime would be limited if I'm tied up and haven't checked in on things in a while.

I am seeing this.

 

[kdetmer]

any ideas on 71c, it looks like there's a heartbeat on the audio level while monitoring 1 channel. if a tone matches, it will continue to record heartbeat sound after transmission is complete until it times out

disregard, was a radio charging issue

 

[xxbryantd]

Any thoughts about being able to send admin messages to a select few that are in specify group? Say I add someone new to "page group 1", but I wanna make sure I typed everything correctly and they get it. I could then select that group, then the email/number for that person then send a test message

 

[johnmac]

I just put the two instances minimized. Will see how that does, may take a week or more to see if it crashes or not. If I go headless, what address do I put in chrome to view the GUI?

John

I set one instance as open on the screen, it crashed after about a week. The other instance was set to headless at the same time. was able to access via the browser at first, but can no longer access via localhost, but it appears to still be running. Next step will restart and set both to minimized. See you in a week!! Could be some glitch on my end maybe?

John

 

[MikeOrlando02]

Just got my new laptop running for TTD 71c. I did see a few times "Failed to execute script", but it is only my first day. I am running two instances, one for EMS and other for Fire. I was not aware that you could run two instances on different channels using the same sound card. We have always used the on-board sound card and then a USB sound card to make that work. Both of mine are set to left, but they are on separate devices. Could this be my issue? I had a MVC and fire and EMS were both paged and I got both recordings respectively.

The biggest issue I am seeing is that I seem to only be capturing the first page in some circumstances. I thought it might be the case of two strung together as one recording, but it is only the one page. The MVC example above, both first and EMS only had one page recorded each. Any suggestions? Adjust tone tolerance?

I just watched this again. It started recording on the first page and then the second paged ended up being recorded all as one long file, but still didn't detect the second tone set and create a second file just with the second dispatch like previous version did.