I've been digging into issues with an Engine running on a separate host in a distributed setup. Service looks to be running but is no longer chatting with Core and is not longer showing up under Engines in the web UI.
After ruling out some of the obvious causes on this troubleshooting page. I've started logging running processes and ports being used on that Host to understand at what point the service goes silent or inactive. That's when I noticed what seems like large number of TCP connections FMEProcessMonitorEngines uses with multiple ports open.
The Architecture Diagram and Documentation are really helpful. So I recognize the 7070 for FMEEngines services and the 7082 for Database Server but then there are a bunch of ports that I assume are what FME dynamically picks. If I only have one job executing on that engine at a given time, how many of those ports/connections will be in play? Is having 3-5 or so that I don't recognize an indication that something may not be working right? I'd like to rule out that there are some AV/AM or other tools blocking ports or services.
Thanks for any insights or references to Safe articles about this that I haven't discovered yet.