You are here
5 Tips to Prevent Network Failures and Keep Your Sanity
Networks crash, it’s a fact and the vigilant network admin needs to be primed to spring into action should the worst occur. What you may not be aware of however are these 5 simple ways to help prevent network failures to keep you from losing your sanity.
1) Monitoring with Network Analyzer Software
Network analyzers are designed to provide you with all the key statistics and data concerning your network and internet activities. These tools will help you in identifying problems before they become a serious issue impacting on productivity and performance. For example; a simple check on the bandwidth use on your network will help you pinpoint bandwidth hogs and others abusing the network. Better still, network analyzers allow you to report on the data. This includes the analysis of network traffic. When the lights go off and all hope seems lost even the simplest PC can be turned into a network tester to help diagnose problems.
2) Use SNMP Traps
Simple Network Management Protocol Traps (to give them their long name) are a way for an agent to send a monitoring station ad-hoc notifications about conditions the monitor should know about. Let’s say you lose connection with one of your routers. The SNMP trap allows the management station to be notified of any or all significant events by way of an unsolicited SNMP message. For those wishing to persist with SNMP traps and for the more experimentally minded there is the possibility to customise and configure your network monitoring software’s response to different traps. This could range from a full scale power shut down to running a piece of script should the trap be hit.
3) SNMP Polling
SNMP polling differs from SNMP traps in that you have the functionality to poll or ‘ask’ your network devices for their status and statistics regularly. This polling can occur constantly or at set integers; you will generally poll for either ‘status’ when you want information on an up/down or own/off function or ‘statistics’. These statistics encompass a variety of information such as traffic status, memory usage and error reports. The timescales to which you poll your network are dependent on your individual preferences. Anything from constantly to five minute intervals is not un-common.
Utilising a ping utility will enable you to test the reachability of a host on an IP network. The idea being is that you enable pings to single or multiple addresses to get various information on the availability and latency of network attached devices. When might one ping? If you want to know whether a remote device such as a web server can be reached across a network you could use a ping to determine whether this can be reached. Failure to reach the server will return you with a notification while successful connection will provide you with statistics such as the latency of said server (particularly useful for high volume sites such as games servers).
5) Process maps for visualization
Process maps used in IT monitoring software give you that quick at a glance visual to see where the errors are. All you need to do is upload your process mapped image, size it as desired, and then overlay with the various checks you wish to set up. The process maps enable you to have a visual representation of your network while creating high level information diagrams. For the case of internet activity, you can set a ping to check if ‘google.com’ is reachable, if anything happens to break that link from being reached you will be notified on your process map with orange, red and green visualisations.
These five simple tips will help you make sure a catastrophe doesn’t occur with your network. Hopefully you'll keep your sanity, and maybe even be able to grow out your hair a little bit faster than the rate you’re pulling it out at. It’s really a win-in.
Why not try out Opsview free for 14 days? With our easy to use installation method your trial can be up and running in minutes!
More like this
We have various articles already in our documentation for setting up SNMPv2 trap handling in Opsview Monitor, but SNMPv3 traps are a whole new ball
More often than not, we enable all alerts to all devices and expect to not get frustrated with the volume of alerts we receive.