Can anyone recommend the best software for monitoring Linux AND Win2k servers/workstations? I'm interested in both network AND system monitoring capabilities.

Thanks.

MRTG via http://people.ee.ethz.ch/~oetiker/webtools/mrtg/ is excellent for monitoring servers and clients, as well as switches and routers.

Also, for management Nagios via http://www.nagios.org/ rivals Cisco's management/HP Openview for switches/servers/etc.

There are many more, but those are the main package's for remote management and monitoring.

Appreciate the open source recommendations, but if cost were no object, would there be other alternatives that you'd suggest?

Thanks again.

If cost is no object.. Then I would hire a guy to code the server and client parameters into mrtg and nagios. HP Openview is nice/easier, but the GPL one is more versatile IMHO.

I've never configured either, so I don't know, but aren't nagios and mrtg overlapping products or is one better suited for systems and the other for networks/routers?

I'm seriously considering taking the nagios plunge and wondering if it will be worth the effort it is going to require. Right now we have a crappy configuration of sitescope that I'm on the verge of taking a shotgun to.

Look at RRDTOOL, which is a much needed upgrade to MRTG. It's at

http://people.ee.ethz.ch/~oetiker/we...rld/index.html

in particular, Cacti is an easy to set up monitoring tool that uses RRDTOOL

http://www.raxnet.net/products/cacti/

It costs nothing to try it out. You need to do alot of testing before you make a tool production anyways. Take the time to try out a couple of the "bigger" packages.

OpenView, NG and Opennms offer a ton of features. The problem is if you only want to look at system stats and or uptime they are complete over kill.. They also require clients to be loaded on the machines for some of the information to be polled.

The people suggesting this stuff are obviously stat hounds (not to mention openview is completely out of the scope of this discussion).

Use mrtg to graph uptime and resource usage, and call it a day.