Raspberry Pi Feed "Dies" every 5 days

[MartyMain]

Brand new installation, just a month active. And I'm pretty much a newbie at Raspberry Pi, although I some limited and distant past experience with Linux.

I get a notice every five days that the feed has stopped in the wee hours of the morning. It's a 20-minute drive for me to go to the installation site, so I'm going to install TeamViewer in order to remotely access/reset the Pi. But I wonder why it's happening? As near as I can tell, there is no power problems at the Fire Station where it is installed.

I read somewhere if the network connection (modem/router) goes down, that the connection must be restarted in the Pi (manually or by a automatic script). We're using Starlink Gen 2 via WiFi (literally 12 inches away as I saw no reason to buy the Starlink Gen 2 ethernet adapter).

Suggestions and advice welcomed.

 

[sallen07]

More of a workaround than a solution, but you could add a cron entry so the Pi reboots every morning.

15 3 * * * /sbin/reboot

That will reboot the box at 3:15 every day, but of course you can adjust the hour and minute as you wish.

I'd look in the logs to see if the wi-fi connection is being lost. It could be an upstream network issue too, and if that's the case there is probably little you can do to fix it. It's also possible to put together a script that can run out of cron and reboot the Pi if there is a connectivity issue.

 

[jtwalker]

You get an email from Broadcastify? You sure it isn’t just an auto notice after 2 hours because Broadcastify hasn’t received any radio traffic due to things being slow in the middle of the night? If it is this, you can increase time in Broadcastify setup before it will send this notice.

Obviously if your pi is really restarting/crashing then this isn’t the answer.

Starlink does update and restart every few days in wee hours of the morning.

 

[MartyMain]

I just got to the installation site and found the Service running with the WiFi still connected. I tried restarting the Service, but no joy. Then I restarted the Pi via the BCFY Shell and viola! it started working again.

Concerning the Broadcastify inactivity time out and email notices from Broadcastify, our County has medical calls as well as fire calls throughout the night, so I'm confident that is not the issue. But thanks for the suggestion and I'll look into the Starlink restarting "feature," too.

So, I'll create a script to restart the Pi nightly as suggested. And I'll still add TeamViewer for remote access.

 

[vagrant]

Three Pi‘s needed to reboot daily.
Another Pi reboots weekly.
A third Pi I login and reboot monthly.

Cause = memory leak

My Pi’s are running different things, so the loss of available memory differs, thus the time between rebooting differs. I recommended the crontab edit as well. It had me baffled at first, but the timing of the issue was consistent and led me to the memory issue/cause.

Once I enabled the consistent reboots, no “more” trouble for years now.

 

[Hit_Factor]

Set every piece of equipment at the remote site to reboot once a week (or so).

If the device can't do it on it's own. Use something like a light timer to power cycle the device.