How can you monitor the overall status and health of your network? ~ Ask The Admin

Monday, June 02, 2008

How can you monitor the overall status and health of your network?

So you've just finished rolling out 500 new desktops using disk imaging, and you're keeping them updated using WSUS. How are you going to monitor the overall status and health of your network?

There is a wealth of excellent network monitoring software available, both commercial and open source. One problem with many of them is that they are really geared towards the very large network. Monitoring 1000 servers, 300 switches, 100 routers and 15 firewalls on 3 continents is very different from monitoring 10 servers, 2 switches, 1 router and 1 firewall in a single office. Commercial monitoring software is probably going to be prohibitively expensive for a small network.

JFFNMS (Just For Fun Network Management System) is an excellent open source network monitoring package you can run on any spare Windows or Linux server. Don't let the name fool you, it is a full-featured piece of software which includes autodiscovery, fully configurable alerts, performance graphing, reporting and network mapping.

Installation and configuration is pretty typical for an open source project (meaning a bit more complex than a typical Windows installer package) but I'm sure any experienced administrator can handle it. You will also need to install and configure SNMP on any machine you want to monitor. (Full disclosure and shameless plug: I wrote the new version of the Windows installation instructions.)

After installation, the rest of the configuration is done from a remote browser interface. You add your individual machines and interfaces to the monitoring system and set what parameter changes you want to be notified about. You can monitor pretty much any part of the system that can be queried by SNMP, such as free disk space, network utilization, processor usage, reachability, or if a specific application is running. You can then be alerted when a specific threshold is met or event occurs. You can create pretty graphs to better show trends and create reports of system uptime and availability.

Even in a small network a good NMS allows the administrator to keep on top of the network and be alerted to any potential problems before they result in downtime. It's much better to receive an email telling you the mail server is running out of disk space than to start getting angry calls from users complaining the mail server is down.

Try a working demo.

Download JFFNMS.