post #1 of 1
Thread Starter 
So I finally am getting started on this.

About a week ago I had posted looking for a monitoring solution that could do both NetFlow and SNMP, but came up short. So I decided to stick with OpenNMS for SNMP monitoring and using Scrutinizer for NetFlow monitoring. OpenNMS is an open source application, and Scrutinizer has a free version that will only hold data for the past 24 hours. I wish the free version of Scrutinizer held unlimited data for a certain number of nodes, but beggars can't be choosers!

So anyway, let's get started with OpenNMS. I have used this software in the past, and things pretty much seem to look the same now. This is a java based application so it can run on pretty much any OS (Windows Server 2012, of course, being my choice) and can monitoring pretty much any kind of protocol or service. I have not finished configuring mine all the way, but here is a quick list of the services that I have configured for monitoring:
HTTP (various ports)
SQLServer (Microsoft SQL)

Most of these were automatically configured after a discovery scope was set up (no agent is needed for OpenNMS). OpenNMS is monitoring uptime on most of these by default, so nothing spectacular except knowing if something goes down. SNMP, on the other hand, is the one I wanted to use which can give me statistics such as HDD storage utilization and bandwidth monitoring for any NIC. As of right now I have SNMP configured on only two devices, the first being my firewall (Ubiqiuti EdgeRouter Lite) and second being my HyperV03 server (now my primary Hyper-V host). Here are some screenshots of my Firewall device (named FW01 -- public IP masked for obvious reasons):

This is the default page for any node that you view with OpenNMS.

From the above screenshot you can see what services are being monitored, the availability of those services in the past 24 hours, the most recent events, and some other information. The most important factor for me is in the node navigation bar, the "Resource Graphs" link. This is where I can view graphs showing certain information polled via SNMP. Lets take a look at eth0 (my WAN interface, using DHCP and connected to my Time Warner Cable modem -- 50 x 5 is my connection):

This is the default graph view showing bandwidth for the Last Day (24 hours).

You can see from this I have had an average of 668.14Kbps traffic inbound to the WAN interface (in to my network from the internet), with a minimum of 112.79Kbps, and a maximum of 38.26Mbps in the past 24 hours. Right below that we can see I have had an average of 4.77Mbps traffic outbound to the WAN interface (from my network to the internet), with a minimum of 2.67Mbps, and maximum of 5.34Mbps in the past 24 hours. Below that is the combined total, showing a overall total of 467.55Gbits of traffic in the past 24 hours. That's a total of 58.44GB of data in 24 hours being transferred. What is all of that traffic you ask? I have my cloud backups running (still running), to backup my TV Shows (I have my important data backed up, so now I am backing up less important data since AltDrive gives me unlimited storage for ~$50/year).

Here is the same graph showing bandwidth for the Last Week (7 days).

You can see from here I have had a total of 2.19 Terabits of transfer in the past 5 days (since I don't have a full week of data on here yet). 2.19 Terabits is an approximate total of 273.75GB of data!

This is the information I am wanting to see and track, and SNMP gives me the ability to see that. I can also, instead of stats from eth0 (WAN interface) I can look at the stats of eth1 (LAN) or eth2 (DMZ) to see more specific information. This is my main use of OpenNMS, so I am not going to go into too much more detail with this application. For fun, here are a few more graphs showing information from HyperV03:

Take note of the time period. Here I am using a custom time to show traffic over the NIC (2 Gigabit NICs in LACP but configured for Hyper-V) between 12AM and 7AM.

Here is the same graph, but showing bandwidth between 7AM and 3PM.

Two of the VMs on this Hyper-V host have iSCSI storage to another LAN attached storage server, and one is used as a temporary download directory for my Usenet VM, which must have downloaded something during this time showing the spike of traffic.

So yeah, that's pretty much it for OpenNMS. You can do so much more with it, such as custom monitoring using WMI, configuring alerts, email notifications, etc. There is also mobile apps available for OpenNMS to connect and monitor from a phone or tablet. Pretty cool stuff!

Check back in a few days, and I will eventually write something up about Scrutinizer!

Let me know if you have any questions or anything is specific you are wanting to see, and I will gladly show that off.