This is Darren Mallette from Quest. I'm a Solutions Architect on the Foglight team.
I recently had a proof of concept where there were firewall configurations in place that wouldn't let the monitored HP-UX hosts communicate with the Foglight Management Server (FMS). However, we got lucky as there was an intermediate host that could accept connections from the monitored hosts, and also communicate with the Foglight Management Server.
We started by installing Fglam to the intermediate host. It is listed as the "Concentrator Host" in the diagram above. It was installed to /opt/quest/fglam. We configured the URL to point to the FMS on the http port it is listening on, in this case port 8080. Checking that it is shutdown, we edit the fglam.config.xml file in /opt/quest/fglam/state/default/config. In the section, we add an element at the end. In most cases you will just need to specify a port. In this case we added:
This is shown in the following screenshot.
We saved the file and started fglam as a daemon (/opt/quest/fglam/bin/fglam -d). Fglam on this host is now forwarding to the FMS on port 8080, and listening for downstream requests on port 8081.
Next we proceeded to the first monitored host on our list and installed Fglam as normal. When we added the URL for the FMS, we specified the URL for the concentrator host, with the port we configured as the downstream port (8081). An example URL would look like
http://concentrator_host:8081 . We tested the connection and it was successful. Once the installation was finished we started fglam on the monitored host and then we were able to deploy agent packages and create agents for it from the Foglight console.