monitor
monitor copied to clipboard
monitor.sh stops working without error message
Hi,
I am using monitor.sh for home assistant. For this I created a docker container out of monitor.sh The scripts runs fine for a while, then it just stops doing anything. No further logging, no updates within mqtt. If I check the .pids file and compare it with the running processes it looks that a few of the internal startet processes are no longer running. Any idea what I can do?
Lukas
Here is a detailed snapshot, when it fails:
root@raspberrypi2:~/monitor# cat .pids
> log listener pid = 164
> btle scan pid = 165
> btle text pid = 166
> btle listener pid = 168
> mqtt listener pid = 169
> packet listener pid = 171
> beacon database time trigger pid = 174
root@raspberrypi2:~/monitor# ps -ef
UID PID PPID C STIME TTY TIME CMD
root 1 0 0 Sep20 pts/0 00:15:26 bash monitor.sh -b -a
root 164 1 0 Sep20 pts/0 00:01:36 bash monitor.sh -b -a
root 165 1 0 Sep20 pts/0 00:00:19 bash monitor.sh -b -a
root 169 1 0 Sep20 pts/0 00:01:15 bash monitor.sh -b -a
root 171 1 0 Sep20 pts/0 00:07:25 bash monitor.sh -b -a
root 174 1 0 Sep20 pts/0 00:01:30 bash monitor.sh -b -a
root 188 169 0 Sep20 pts/0 00:00:00 bash monitor.sh -b -a
So missing are the pids 166 (btle text) 168 (btle listener)
Any idea how to go on?
Is this package still beeing supported? Does not seem so.
Theres a couple of Dockerized Monitor rl's here on Github, Im sure by now you've compared them to your setup...?
1st I have build my own one and after that I switched over to one mentioned github containers here. But the error situation is still the same. Today I restart the container every 3 hours :(
You can also pull the beta, then your up to date. I believe there's a memory leak and the creator is working on it and has had a baby in the last few months. So things are in a standstill for the moment.
It's been about six months since I've made changes, due to a newborn. One of the big issues with docker containerized versions of monitor (and certain USB dongles) is flooding of adv messages. What BT hardware are you using with your container?
I'm having the same problem.
Funny thing is that I had monitor running in docker for more than a year without problems. But after trying to update from monitor 1.x.x to 2.x.x, I was forced to recreate my docker image with debian buster, because of mosquitto>1.5. And now I'm facing 2-3 hours of working time, then I need to reboot.
No error or anything indicating a problem, only thing is that i seem CMD-RAND messages dosn't seem to come trough.
Using RPI3B.
I had a similar problem on rpi zero.
got it working again by using python-bluez. It is described here how I did it: #288 and #295
Had the same problems running on rpi zeroes. Tried to switch to rpi3b's and have been running fine for a couple of days now.