Hi:
Long time happy user. Just did a Fedora 37 update using dnf to upgrade the
kernel from 6.3.4 to 6.3.8-100. Upon reboot, none of my frontend clients
could connect to the backend. Netstat reflected that there was nothing
listening on port 6543.
Backend was running and the logs showed it interacting with the database.
But when I tried to stop / restart the service, it wouldn't complete the
shutdown, just hung (systemctl stop mythbackend). If I got the PID with ps
-aux | grep mythbackend, and then tried to kill the process, it wouldn't
kill. Even "kill -9 XXXX" wouldn't kill the mythbackend process.
Only way to kill the process was a reboot, which was lengthy because the
kill timers all had to timeout before rebooting.
Anyway - I selected the prior 6.3.4 kernel from the grub2 boot menu,
everything booted right up, mythbackend started and functioned fully
normally as it has for many years.
I ran backend in the foreground with -v all, but no errors popped.
Any suggestions or ideas? It is reproducable. if I boot 6.3.8-100,
mythbackend is a non-responsive zombie. Boot 6.3.4, everything is fine.
Thanks for any ideas.
Long time happy user. Just did a Fedora 37 update using dnf to upgrade the
kernel from 6.3.4 to 6.3.8-100. Upon reboot, none of my frontend clients
could connect to the backend. Netstat reflected that there was nothing
listening on port 6543.
Backend was running and the logs showed it interacting with the database.
But when I tried to stop / restart the service, it wouldn't complete the
shutdown, just hung (systemctl stop mythbackend). If I got the PID with ps
-aux | grep mythbackend, and then tried to kill the process, it wouldn't
kill. Even "kill -9 XXXX" wouldn't kill the mythbackend process.
Only way to kill the process was a reboot, which was lengthy because the
kill timers all had to timeout before rebooting.
Anyway - I selected the prior 6.3.4 kernel from the grub2 boot menu,
everything booted right up, mythbackend started and functioned fully
normally as it has for many years.
I ran backend in the foreground with -v all, but no errors popped.
Any suggestions or ideas? It is reproducable. if I boot 6.3.8-100,
mythbackend is a non-responsive zombie. Boot 6.3.4, everything is fine.
Thanks for any ideas.