Trying to troubleshoot Zenoss core 5.1.1 sporadic issues

We have an environment where we have a Perl script that sends events to Zenoss Core using RPC::XML . For a year or so, this has worked most times.  However, every so often, the RPC::XML returns without a response, and the event that was passed never appears in the event queue or archive.

When we discover this happening, we usually can go into control center to restart the processes, or, if that fails, reboot the machine, and things begin working again.  Sometimes the length of time is weeks between an issue.  Sometimes it is minutes.

We are seeing nothing in the logs except in zenhub.log.   All it shows is ongoing health check warnings which we have yet to figure out how to resolve. However, these warnings have been going on since day one, so it seems less likely they are directly related, since the events are showing up more often than not.

What gems of wisdom might we be pointed to that would lead us to resolving such issues?

Thank you