Today I have a big problem, I’m in charge of a Linux network of 100 computers distributed in 4 rooms of 25 computers each, I have a DNS server and I want to have the current state of every machine at real time, I have been developing a script that pings each computer one by one and changes a file every time the state changes (on to of; off to on) and prints the date, like a regular LOG file (a daemon, pretty much). and a few days ago I found this in the manual for the command ping:
This program is intended for use in network testing, measurement
and management. Because of the load it can impose on the network, it
is unwise to use ping during normal operations or from automated
scripts.
If this is not the right way to check the Status of my computers, what would you suggest?
Use Cacti or Nagios for network monitoring.
They’re great, open-source, and will allow for you to measure MUCH more than just pings.
These tools are built upon the RRDTool package (RRDTool is great as well.. check it out), and they offer a GUI for it, with graphing. Now the graphic that you get are images, so they’re not interactive like the graphing that we’re seeing now-a-days with JavaScript. But, you can still modify the constraints and have what you’re looking for represented properly.
You can also set up alerts. So you’ll know when something is awry.