Address watchdogd service problem with generic script > 1s runtime #39

[senthilnthangaraj]

This commit addresses a problem related to the watchdogd service. When trying to
activate a generic monitoring script (in this example it's /usr/sbin/my-script.sh) with a runtime exceeding one second,
it triggers an unintended system reboot.

Our configuration is as follows:

generic {
enabled = true
interval = 60
timeout = 20
warning = 1
critical = 10
monitor-script = "/usr/sbin/my-script.sh"
}

The error message we're encountering reads as below (even though my-script returns 0):

sv-isup my-service
watchdogd[661]: Monitor script PID 1056 still running after 20 sec
exit 0
watchdog: watchdog0: watchdog did not stop!

Upon further investigation, it was determined that the problem arises
from the fact that 'gs->script_runtime' is measured in milliseconds,
while 'gs->script_runtime_max' is maintained in seconds, as indicated by the source code here: link to source code. This commit rectifies the issue.

Unit test results

With the fix in this PR there is no failure in watchdog service and it works as expected, please see below the traces.

sv-isup my-service
exit 0

[troglobit]

Please update the commit message to state what it does, and more importantly why. Also mention how you have tested it to verify that it fixes the problem. Don't assume access to the GitHub issue tracker for this.

[senthilnthangaraj]

Please update the commit message to state what it does, and more importantly why. Also mention how you have tested it to verify that it fixes the problem. Don't assume access to the GitHub issue tracker for this.

@troglobit I have updated the commit message and added the UT results to the PR. Let me know if this looks good, thanks.