10 Tips For Efficient Network Management
Most major network faults are often the result of many smaller ones; fix the little issues and the big ones won’t be as numerous and your network will run more smoothly.
Tips For Keeping Your Network Healthy And Trouble-free
- Naming standards. It’s important to get your naming standards sorted out. The name should convey useful info; what type of device, where it actually is physically in the network and where it is logically.
- Basic alerts and monitoring. You should have basic availability monitoring in place for all your devices. Add CPU, throughput monitoring and memory monitoring for all key interfaces. It’s not practical to have reports for every single interface, but this just isn’t practical for most administrators. Collecting historical data is also very useful for troubleshooting.
- Automated configuration backups. There are plenty of inexpensive tools to do this for you. This not only helps when recovering device failure, it also shows you what’s changed in your network. Once all your configs are in one place, you can analyse them to make sure your settings are correct.
- NTP. Enable NTP; it costs nothing but gives you a lot; remember to set consistent time zones across every device. This can be invaluable when trying to track down precisely when events happened.
- Logs and traps. Make sure all devices send syslogs and traps to one central location. Log reviews are really useful exercises to carry out although you may need several iterations to boil down to the handful of ‘unknown’ entries; these are the ones you need to look at that are probably causing your users frustration. Look for the most common entry pattern and most frequently occurring source and fix what you find amiss.
- Duplex mismatches. Learn what signs to watch out for and keep an eye on your interfaces so that you can act immediately if necessary. Most networks suffer from duplex mismatches; fix these issues and you’ll get fewer complaints that the network is slow.
- Standardise firmware versions. All your 2920s should run the same ProVision version and likewise your 3750s should run the same IOS. It’s sensible to keep them up to date too and check the release notes for the new versions. Use the same versions throughout your network and all your commands and features will work consistently.
- Interface descriptions. It’s a good idea which will save you hassles in the future if you put descriptions on some of your ports; inter-network device links, server-facing switchports etcetera. If you’re dealing with a fault, it’s really useful to have the relevant information right there on the switchport in front of you.
- Templates and fancy features. Modern equipment has a host of useful features that can not only improve service quality, but help protect the network too. Take advantage of this by developing some useful base templates for your configurations using the key features and roll out the templates everywhere.
- Be disciplined. Make it a priority to fix stuff as you find it. If it’s not possible to fix it immediately, make a note of it so it’s not forgotten about. Little issues that are shoved to the back of the cupboard and left to fester can suddenly emerge as major problems; so be disciplined and fix stuff as you find it.
When it comes to the health of your network; the buck stops with you. Put in steady work to keep improving things and keep on top of small faults and things are far more likely to run smoothly. Pay attention to the basics, and the big issues seldom arise.