Tuesday, October 18, 2011

Minimizing data gaps when troubleshooting network devices

Traditional NOC-based network management uses SNMP polling over the network to send requests for device information. This traffic slows down both the network and the devices, so the goal is to minimize the polling by extending the time periods in between requests. Depending on importance of the site, traffic volume and other factors, the time between polls is measured in minutes -- often 15, 30 or even more.

When there is a problem with a device it is likely to occur between polls, and often data useful for troubleshooting is lost during the event and both before and after the event. This diagnostic information is often an important tool for determining exactly what lead up to the event and caused the issue. How many times have you known what the problem was, but not the cause? It's not a good feeling.

With Uplogix, an intelligent appliance is connected directly to managed devices over the console ports. Syslog and other information are collected locally, without relying on the network and without taxing the devices. Data collection intervals are configurable, with every 30 seconds being standard.

Also, since the data collected is held locally on the appliance, it's always accessible for troubleshooting -- even through an out-of-band connection if the network is still down. This can be a handy thing to have in your toolbox when it comes to isolating the problem.

See the one-minute excerpt below from the webinar, Lowering Network Infrastructure OPEX, for more information.