ltpitt / bash-network-repair-automation Goto Github PK

This happens on rpi zero w, fixed by:
command -v /sbin/ifup >/dev/null 2>&1 || { echo >&2 "Sorry but ifupdown is not installed. Aborting."; exit 1; }

cat /etc/debian_version
10.1

also on rpi model b
cat /etc/debian_version
9.11

After a while the log file contained hundreds of /bin/sh: 1: Cannot fork lines.
The network connection is very spotty, so it's possible the script triggered a lot.

Running in a Pi Zero W 2 using 02c5fc5

5.10.63-v7+ #1496 SMP Wed Dec 1 15:58:11 GMT 2021 armv7l Raspbian GNU/Linux 10 (buster)

after 5 netcheck its successfully restarting wlan0, but in this case 23.26 watch this time in log file after this log next 6.24 in next day log is only there in log file,
what happen pi between 23.26 to 6.24 ?

note : i was tested this script in raspberrypi board runing 24/7.

Hi, Is there some way to use this script with multiple interfaces?

I have an environment when ethernet might go offline quite frequently and would like it to check using wifi then before doing anything.

I have set the reboot_server=true. However, as the /tmp/.last_net_autoboot is on /tmp it gets removed on reboot and does not exist ever after the reboot. Hence, the reboot_cycle=60 is not really effective. /tmp is cleaned on reboot at least on Raspberry Pi OS.

Another note. The network_check_threshold=5 can be quite short. On my environment there might be easily four failures after the network reset by the script. The script is looping in some 20 seconds those attempts.

The find command on line 73 in network_check.sh uses the -mtime option with a value of 60 for reboot_cycle. When I read the documentation for the find command it looks like it won't reboot unless the last boot was more than 60 days ago, rather than 60 minutes ago (which I assume was the intention). I think the option should be -mmin

Per your instructions, /etc/default/rcS does not appear to exist on a Raspberry Pi 2 running Raspbian 10 (Buster).