hyperic agent on vROps vApp

[rgcda]

I'm trying to install the Hyperic agent on the vROps vApp for the purpose of monitoring whether the machine is up or down via Hyperic so if the vROps machine fails I'll get a notification from Hyperic.  I've installed the agent and went to start it but it seems as if it fails to communicate with the Hyperic server.  I presume there is a firewall in vROps but I cannot find any documentation on it.  Is anybody familiar with if there is a firewall in vROps 6.x and how to edit / configure it?

[mark_j]

Yea.. don't do that. There is a firewall in vROps, and playing with it will compromise the security of the appliance. Installing additional agents within vROps is not supported nor recommended. The self-monitoring is pretty good for vR Ops, so anything app-level or indicative of impending doom would yield a notification. I wouldn't rely on Hyperic-originating alerts, as that isn't going to fit in to the vision of where we're heading with the suite.

Also, you said vApp, but v6.x isn't in a vApp container. It's a virtual appliance (VA").

[jomccon]

It is pretty disappointing that VMWare's agents don't run on VMWare's servers.  This is the same for VCM.  I'm not sure why anyone would be comfortable with a vApp monitoring itself.

[admin]

jomccon, what do you mean by "I'm not sure why anyone would be comfortable with a vApp monitoring itself."? what else would you expect in that case?

[admin]

Hi rgcda,

You can do it, but as Mark mentioned, that is a change to the firewall. You will need to open port 2144 to let the agent connect to the server and send its monitoring data, as well as resources and configuration.

I don't know why there are claims which this is not supported, or in the roadmap, but I can tell you that we have such solution and it works well enough, not for Hyperic but for vROps itself and we use it in QA environments. If you are OK with opening the port on the appliance, you will be able to monitor the machine's state and know whether the machine is up or down.

[jomccon]

I would expect it to be integrated to allow Hyperic to complete normal troubleshooting functions that Hyperic completes for resources where a Hyperic agent is installed.  I would also expect it to be compatible with VCM from a config management stand point. 

For example, Hyperic can complete actions such as restarting a service if it dies or a specific event is detected, or it can send an alert.  If this vApp has an issue sending alerts, no one could be notified.  VCM pulls attributes such as IP address and NIC information, which can be used for compliance and reporting.