Hello again.
This week I have been playing with network monitoring. We currently monitor all of our network infrastructure and our customers connections using nagios and cacti. These are both fantastic open source tools and serve well however we are using nagios v 2 and v3 is now out, plus maintaining it involves editing about 16 files which with my fat fingers often means I make a mistake and it takes a while to fix.
I built myself a nice new box with some 500Gb Hard drives in to hold all our logging information and set about setting it up. I used the Cent-Os Linux distribution which is a red hat clone as we do on all of our hosting. I would recommend this distribution to anyone setting up enterprise services. While the updates were installing I had a look at other network monitoring solutions and I came across groundwork.
Groundwork offer a free community version and also a starter, professional and enterprise edition which come at a price, but support and extra features are included. They are not cheap ranging from £4,000 to £24,000+. The Community edition offers a wide range of features including Nagios, Management portal, Nagios plug-ins, Alert and escalation, SNMP monitoring console and Network health Overview.
After following the installation I got my web interface up and running and started to add my hosts. After trying the auto-discovery I realised that although very useful for an undocumented network, it was not required for a well documented network. I added all my host groups and started to build the network. It was amazingly easy and very satisfying to see it build up so quickly.
After adding all of my hosts I started to configure services. I added all of the Linux hosts to the system and set them up so that the plug-ins could SSH in to them and check the disk space, memory usage and process level. This should help maintain the servers without having to do so many weekly checks on them. I also added the smtp, http/s and dns services as appropriate. I was able to add snmp monitoring to all of my network devices and graph the interface status and the traffic passing on them. I haven’t got around to setting up the central syslog server yet but I will save that for a rainy day.
Having configured all of my hosts I was able to configure email notifications to let us know as soon as there is a service problem and the nagios map helps us diagnose faults quickly and efficiently. This lends itself to providing good service.
How can network monitoring help you
- Help you understand your network
- Locate any bottlenecks
- Understand your network requirements for the future
- Provide good service by allowing you to diagnose network faults more quickly
If you are interested in network monitoring then contact us and we can set up something like this for you.
Matt.
