What can I monitor agentlessly fron hyperic open source

[hypermike]

I was asked what can I have Hyperic monitor if the client does NOT have a agent installed.
Seems some admin worried that having Hyperic agent Java (jre) on a system could
corrupt or confuse the system use of its own java or JRE.
Is there a list somewhere of some such items that can be monitored without ana agent . Is there any thing the unix ,linux or windows systems needs to be monitored agentlessly.??

[ama_hyperic]

The agent is kind of crucial for monitoring. Even with a SNMP device, its still proxied through an agent.

If your admin is concerned with 2 JRE's on a system, the agent usually uses its own JRE, its in the agent directory and normally not anywhere near a normal path or installed in such a way that it would conflict with another JRE on the system.

[BradFelmey · Accepted solution]

There are a few things, like pinging it, trying to connect to sockets, and so forth, but any worthwhile data is going to come from a resident agent.

I can confirm that having HQ on systems with other JREs/SDKs does not interfere in any way, shape, or form. We have lots of systems being monitored by HQ that have Weblogic, Oracle, Tomcat, JBoss, and so forth running on different JREs that are not affected whatsoever.

[hypermike]

I guess things like http,dns and other services can be monitored but an agent on some
server would be needed, but not an agent on the server hosting the web site etc. Thanks for your answer.