I have a VM running OEL 7.3, inside Oracle VM 3.4.1.
This system is running a database ported from SPARC. The application is old, depends on some scripts that would run serverside.
Some scripts are KSH, many are Perl.
These system (dev, validation, prod) hang on every "shutdown -r ". A "reboot -f" works, so I suspect it is not a kernel or device problem but something in the service scripts that do an orderly shutdown.
I suspect it is something we have monkeyed with to make the application's scripts run with little to know modification.
So I would like to debug the systemd shutdown processes and scripts and see what might be hanging it up.
I've watched from the OVM console and observed it hanging a very long time at the Network Manager and then at the firewall.
Stopped target Network is Online.
Stopped target Network.
Stopping LSB: Bring up/down networking... (long pause here, several minutes)
Stopping firewalld
Stopped Rollback uncommitted netcf network config change transactions (followed quickly by)
Stopping Rollback uncommitted netcf network config change transactions ... (with a long pause, one minute)
Many lines will then proceed normally
Stopping Laod/Save Random Seed
Stopped target encrypted Volumes
Stopped Apply Kernel Variables...
Stopping Apply Kernel Variables... (long pause, one minute)
Another 25+ lines scroll through quickly
Reached target Shutdown.
Going away and coming back after lunch shows:
[ 1938.333051] systemd[1]: Timed out starting Reboot.
And there is hangs until it is killed in the VM Manager. It does not execute the reboot.
So, I'm really leaning towards the shell/perl changes on the system. So I'd like to get some debug output from what these shutdown scripts are doing? The startup up is doing ok, I think...