On Tue Dec 11 20:43:20 2012, LAJANDY wrote:
Show quoted text> Client had issue where two instances of a watch script ran at the same
> time and successfully started two Helios daemons for the same service on
> the same host.
The helios.pl daemon does not try to lock its PID file currently, so
that is a (remote) possibility in a small number of circumstances.
We should be able to fix this bug by exclusive flocking the pid file; it
should only be a few extra lines of code in the write_pid_file()
function in helios.pl.