Customer Case Study: New Edge Networks

New Edge Networks provides managed IP-based networks through a mix of private and public connections to deliver secure, managed network solutions. Connecting business sites at virtually any address in the United States, New Edge allows businesses to choose any blend of available access technologies – from DSL and T1 lines to fiber-optic connections – for building private and secure MPLS networks with performance guarantees managed over an award-winning portal.

OpenNMS manages the entire New Edge client network which currently includes more than 160,000 interfaces and a million RRD files. New Edge evaluated several network management platforms and chose the OpenNMS solution after judging OpenNMS to be superior on performance, flexibility and total investment costs.

New Edge Networks and the OpenNMS Solution

New Edge Networks began using the OpenNMS platform in 2004 when New Edge concluded that a legacy system that came with one of its acquired businesses was no longer sufficient for its expanding needs. In the five years of rapid growth that have followed, New Edge has used the OpenNMS platform and the OpenNMS Group to gain:

  • Exceptional network stability in a Sun/Solaris environment.
  • Easy NMS scalability to more than 160,000 interfaces.
  • Custom monitoring applications and other applications built quickly to solve critical New Edge needs.
  • Incorporation of New Edge-required functionality into new releases of the OpenNMS Platform.
  • High ROI from a very cost-effective NMS that allows a small staff to manage a large network.

Five Years on the OpenNMS Platform at New Edge Networks

Scott McLaughlin, Director Infrastructure Operations at New Edge Networks, summarizes his experience with the OpenNMS platform and the OpenNMS Group:

In 2004, we received some new client requests that our legacy network monitoring system (NMS) could not support, so we started looking around at other options. At the time, the decision really came down to what we thought we could do for the money, which gave OpenNMS an advantage that helped us make our choice. But as we’ve grown, we’ve realized that we made a good choice for far more reasons than acquisition cost. The big thing was finding something we could grow with really fast. OpenNMS has proven to be the most scalable option

Our internal support costs have dropped dramatically... and [we've] acquired better management tools.

In the last five years we’ve gone from 10,000 interfaces to 160,000 interfaces. It’s always going up and now there are more than a million RRD files on our Net App. And yet, our internal support costs have dropped dramatically as we got things more stable and acquired better management tools. Today we have just one person who takes care of it all and it’s not even a half-time job.

We don’t need new customizations with every OpenNMS upgrade because they’ve built our previous customizations into the platform.

One key to our ongoing success with OpenNMS is that it keeps evolving to match our needs. The OpenNMS product that’s out today has a lot of pieces that were driven by New Edge requirements several years ago. We don’t need new customizations with every OpenNMS upgrade because they’ve built our previous customizations into the platform.

This never would have happened if the platform wasn’t open source and if we didn’t have the OpenNMS Group. Personally, I’m very glad that the source code for OpenNMS now reflects our needs and innovations, and that other companies are using the technology that we requested for New Edge. It’s really cool, and you would never see that happen with any black box application.

The OpenNMS Group response time is amazing, and we need that because our customers expect New Edge to be working 24/7 for them... Once a problem is in their hands, it’s never very long before it’s solved.

Support speed is one of the things we’ve really appreciated with the OpenNMS Group, and an important reason we’ve been able to justify buying our 24/7 contract. The OpenNMS Group response time is amazing, and we need that because our customers expect New Edge to be working 24/7 for them.

For example, somebody from OpenNMS is logged into their IRC server almost all the time. The support teams here stay there all day – that’s how they communicate. There have been times when we’ve had problems and we’ve just set up a chat on the IRC channel, “Hey Tarus” or “Hey Ben, this is what’s going on” and they say “OK we’ll look at it.” They’re just right there. That’s not something that’s in the contract. That’s just something they do because of the geek in them. Those are the kinds of things we love getting from them that just don’t exist anywhere else.

When we’ve had an issue and asked “what is going on here?” we’ve usually received an answer from one of the people who wrote the piece of code or is an expert in PostgreSQL. They’ve responded, “Oh yeah, I know what that is” and it’s that quick. We’ve never had a major outage that’s lasted long. Once a problem is in their hands, it’s never very long before it’s solved. We’ve never really had a multi-day or multi-hour outage that I can remember.

Advice for the CTO considering OpenNMS

McLaughlin has some detailed observations to share with the CTO responsible for large networks.

First, they need to understand that a smart solution can combine software that’s free plus support that you pay for. People didn’t understand that model in the old days, but it’s getting easier now.

If I didn’t have the OpenNMS Group, I’d have to hire more people.

It’s hard to describe the biggest selling point for the OpenNMS solution. The platform is great, of course. You can take all the arguments for open source and apply them to the OpenNMS platform. Beyond that, though, it’s really about the OpenNMS Group and the way they support us. They really care about our environment. They’re going to do their best to understand it and to make sure it stays up.

If I didn’t have the OpenNMS Group, I’d have to hire more people . I’m getting a lot more value for what I pay OpenNMS because I get a Postgres guy, I get the folks who wrote the code, I get Tarus, I get SNMP experts – I get all the stuff that I need for the NMS and save $250,000 a year or more in salaries. It’s like outsourcing – it’s one of those times where it makes a lot more sense to buy that contract than to staff it on your own.