Monitoring a Nuage Network-Based Infrastructure


Dave Hegenbarth introduces the alliance partnership with Nuage Networks, and describe the benefits of integrating SevOne and Nuage technology.



Hi, I'm Dave Hegenbarth. Director of Systems Engineering for Global Strategic Partnerships at SevOne, thanks for joining. The days of service providers and large enterprises buying rack after rack of specialized network computing storage to launch a new service or application are coming to an end.

Our customers in the market as a whole are on a journey to software defined everything. Where network compute and storage along with automation and orchestration will allow them to innovate and deliver applications and services much quicker through a virtualized infrastructure.

At SevOne we're committed to making that journey with our customers and our technology partners and provide complete performance visibility through the technology transition that is about to happen.

In this whiteboard session I'd like to introduce to you our alliance partnership with Nuage Networks recently required by Acatel-Lucent and describe the benefits of integrating the Nuage software define network solution and SevOne performance monitoring. Let's go to the whiteboard.

On the whiteboard I've drawn a fairly generic SDN deployment where we have the Nuage controllers at the top. We have our physical underlay network at the bottom. We have our hyper visors that run the controller and other network function virtualization components and actually even included an edge gateway. We go to an access to the outside world through non virtualized services.

Why have I drawn this? What is the benefit of SevOne and Nuage ? Well the benefit is this, we have our user up here who is thinking about building a new website. As we talked about over time it takes time to build these sorts of websites in a traditional network. But, with a virtualization using the Nuage software and other virtualize components we can build this network really quickly.

So our network probably consists of upfront a firewall. That firewall might be connected to let's say an IPS for packet inspection because we want to be secure. That's connected to a virtualized load balancer because maybe our websites is going to get popular and have lots of different components to it. Off of the web load balancer we'll have our websites. We'll probably have a couple data base servers as well. Then we have our users who are actually going to get to these particular websites.

Each of these is a virtualized component off of my hypervisor along with the Nuage software, running a VM here that allows us to very quickly configure things like these virtualized services and or a virtualized path through the network because these may all sit on different logical segments. The value of SevOne is collecting the performance statistics of this entire environment. That means first we probably want to have some API's and we do that collect things like inventory. How many things have been spun up in this virtualized network?

We're also going to want to have an API call that collects topology. So we were going to learn in real time the topology of the network. Legacy performance systems probably discover this network through methods that probably wouldn't work in a virtualized but certainty they do discovery maybe once every 24 hours. In the world of SevOne and API's connected to these controllers. We can understand the topology and service chain, a combination of network functional virtualization devices (NFV's). We can understand those in real time and display that to our end users.

We'll also probably use traditional monitoring protocol such as SNMP to get stats out of these guys. So we're going to be polling maybe the load balancer for connection per second or connection per virtual IP. We're going to be polling the firewall to get stats there. SevOne also has the ability to take in logging data so we may have IPS sending our logs there. We may have these virtual servers sending logs there. We might be getting information such as IPMI from the physical hardware that the hypervisor runs on.

We're also going to be taking measurements from the infrastructure, where the networking device is so the physical devices can be polled with SNMP as well. They also might be sending us NetFlow. We'll take in the NetFlow to understand who's talking to who on what protocols at what volume the traffic through the network.

Lastly, we have API's that go to the cloud as well. I may extend my virtual server infrastructure not only in my private cloud that I built here but also, I might share that with a service provider like AWS or Azure and have the ability to bring those metrics together. We have a single paint of glass in SevOne. I'll have graphs that may represent my firewall statistics and that may be polled by SNMP. I may have another performance graph that's actually coming from my IPS and this may be derived from performance logs.

I might also have the ability to collect information from my load balancer. As I mentioned we might have a number of users or a number of connections to a particular VIP. Lastly, I might record flow data on who's talking to who, on which protocols and how much volume of traffic that is passing through the network.

SevOne has the capability of baselining everyone of these metrics and alerting when something is not normal. Whether we're gathering stats from a traditional manager and protocol like SNMP or we're using API calls to go to Azure, AWS. We're connecting to the API controllers of Nuage to understand the topology and the service change that have been built in the network and bringing that all together in a performance report and HTML single painting glass in real time.

That's the SevOne Nuage interaction. Thank you for watching this whiteboard session.