Last week I was honored to be invited to the famous Velocity 2012 conference in Santa Clara California, hosted by O’Reilly. Monitis was one of the silver sponsors. It was a great opportunity for me to meet the wonderful team I’ve been working with for the last year.
Velocity is all about web and cloud computing. A truly great place to learn about new technologies and keep up with bleeding edge developments.
Getting to the Monitis booth I must admit I was slightly disappointed to find out there was no demonstration of any Custom Monitors capabilities.
Custom Monitors when tailored as turnkey solutions are truly a very important aspect of monitoring server and application uptime.
Enough talking, lets get some work done.
A custom monitor in a real-life situation
Lets show people what Custom Monitors in Monitis can really do. Let’s count the number of wireless hosts connected to the conference network. Yes!
Time to implement, test and deploy: 10 minutes. I have witnesses.
I quickly connected my laptop with some USB tethered 3G internet so the wireless card is free to go sniffing.
Did anyone say nmap?
Oh yes. How do we find our IP address and subnet mask?
# ip -f inet addr show dev wlan0 | grep inet | tr -s " " | cut -d' ' -f3
Scanning with nmap a whole subnet is easy:
# nmap -sP ip_addr/subnet_mask
Since we’ve already found the IP address and subnet mask, it’ll be easy to scan the network:
# nmap -sP `ip -f inet addr show dev wlan0 | grep inet | tr -s " "
| cut -d' ' -f3` | grep 'Host is up' | wc -l
We’ve just counted all the hosts in the wireless network!
A sample file containing this configuration can be found here.
Proper disclosure: in the conference the network was a full class B network (65535 addresses). I scanned just a subnet out of it, a class C network (255 addresses). The results are in the picture above.
Along the way another feature appeared in M3 3.9 – additional metrics.
In the wireless hosts example you can see that the IP address on the network will be added as an ‘additional’ metric.
Additional metrics will not be graphed, but will be viewable if you double click on a result.
In the wireless hosts monitor it’ll provide the IP address of the wlan0 interface.
In terms of implementation, additional metrics are handled almost the same as regular metrics in M3, they’ll just be uploaded to Monitis as additional metrics rather than regular ones. Parsing and compute plugins will work exactly the same with them – familiar ground for the M3 warrior.
Additional metrics are very handy when reporting errors or just data that shouldn’t or doesn’t have to be graphed.
Velocity 2012 was a great experience for me. I hope I get another change to prove again that anything can really be monitored with Monitis and easily!
In addition – we’ve got a lot of ideas for steering Monitis (and M3) towards the future, being and staying a lead player in the very competitive market of web based monitoring solutions.
See you in Velocity 2013!