November 3, 2013 1:03 a.m.
Ok, every smart sysadmin should monitor their network. This way the clients and staff do not have to inform you if something is down before you know yourself. Common sense, right? Hopefully so! I make sure that when things go down I am in the know. If I'm not, then I am not doing my job effectively. Just to state... I always know!
At one point in my life, I was part of a department in a company which monitoring was our complete existence. Our job was to watch the entire network 24/7 for an ISP that had presence in multiple South East US states, holding 375k+ customers. We were known as "Netmon" and did our job with pride very well!
Left on our own to figure out the proper tools for such a task, we looked to opensource solutions that could fulfill our needs. We chose a collection of multiple different software products to monitor everything we could. Big Brother, MRTG, Netsaint, etc.
Speaking of Netsaint, which was the purpose of this post. This program alone has been one of my most used automated monitoring tools over the years. As some of you may know, it's now called Nagios, but underneath it's pretty much the same program even with it's evolution over the last 10+ years.
Configuring netsaint/nagios entails a huge number of edits to *.cfg files to accomplish your needs. Not to mention that most of them play on and depend on others to make the entire configuration work.There are tools out there to help with this administration. I first started on command line and doing everythihng by hand, but shortly realized that it's not very effective or efficient for someone else other than myself.
At one time I decided on NagiosSQL for internal edits on a server that doesn't really depend on much. I also used centreon to handle the external things that I need to know about when things happen.
Jump to today... Nagios is still around and used heavily among the world's IT departments. However, now there are better tools to configure/admin your Nagios server. Not that I was opposed to command line, but c'mon! Newer tools automagically make associations that would have to be done by hand in the years past. Things like Check_MK and OMD make the nagios monitoring work so much easier.
If you plan to use nagios centric monitoring, please check out all the current tools. Nagios by itself with manualy configurations has served me well over the years. Check_MK is a server/client for nagios that allows automagic configuration of hosts and services by digging through /proc on the client side. OMD is another option that allows you to deploy different nagios monitoring environments. It could be used for different clients or dev vs prod monitoring to give an example. Total flexibility to configure multiple monitoring situations independent of each other.
Give 'em a look and see what you think. Honestly, there are probably more out there that I do not know about or haven't tried for various reasons. Look at all of your options, but make sure that monitoring is something that you care about and utilize.