Improving Network Efficiency
Tuesday, March 3rd, 2009There are many free or low cost tools available today to help a Network Administrator run and manage an efficient and well documented network. I’m going to discuss a few of those and how they assist me in this endeavor.
Most network devices have the ability to log system messages to a syslog server. For Cisco and other vendors, these messages will range in severity. For example, a notification message with a severity = 5 is a ’notification’event where severity = 2 is a ‘critical’. The syslogging server will collect these messages from all your network devices (even if they are in multiple locations) and store them in a central location for review and archival. When reviewing these logs you may avoid a network situation which could result in a loss of service. Kiwi Enterprises (now Solarwinds) makes a popular syslogging server called Kiwi Syslog. This software can also be used to parse the logs and alert you via email or SMS based on the severity of the syslog message. This is an especially convenient time saving tool.
CatTools also offers change management features. When using their activity to backup the running-config, CatTools will create an archive copy of configs that have been changed. You can even get it to send HTML email reports detailing any lines that have changed. Pretty cool feature I’d say!
For devices that support Netflow or other flow protocols, I use Scrutinizer, which is a free flow collector, and setup the devices to send ‘flows’ to the collector. A network flow is a unidirectional sequence of packets containing IP information. When setup with a collector, you can see detailed information about your net work traffic including ingress interface, source IP, destination IP, and Source/Destination port for TCP and UDP. The screen shot here shows you some of that reporting functionality.
These are just a few simple and low cost systems you can apply to your network to immediately improve overall efficiency, helping you proactively learn about unwanted network events and save critical time when troubleshooting.