Every organization looking at using network function virtualization (NFV) needs to consider monitoring. If you’re going to deploy a service, it’s really important that you know if it’s working — and what’s wrong if it’s not. That’s why monitoring — often called assurance — is a key part of every request for information, every request for proposal, every proof of concept, every NFV deal.
Here’s the challenge: the economics of on-premises monitoring are all wrong for the NFV business cycle.
You want to start small. But you’re supposed to pay — day one — for a pile of servers big enough to handle the five-year business plan. Don’t forget about a professional services team to install it all — and there’s no way that either of you know if what they’re building today is what you’ll need next year. It’s scary to think how much money you could be wasting.
Your team has to learn NFV OpenStack, MANO and virtual network functions (VNFs). Now you’re supposed to layer on learning how to keep an assurance solution running, too? Who wants to build disaster recovery, or stop what you’re doing and plan for a version upgrade? A 24/7 specialist operations team?
And you know that if you start with Ericsson today, you’re going to be expanding with Cisco tomorrow. Or vice versa. Is the initial investment in people, processes and equipment going to be wasted?
Running NFV is bad enough without having to keep an assurance system running, too.
Pay Today. Use Never. Who Wants That?
You’ve dreamed of how it should work. You’re starting with one site, or maybe two. They’re just big enough to test the VNFs in your service chains and get all the orchestration processes going. Assurance? Yep, but fast and easy and low effort. Really, all you want to do is drop in a virtual machine (VM) and start up your browser.
Now it’s time for production. That single VM is a single point of failure. How simple is high availability? You shouldn’t have to think about that, or even buy any extra software.
The business starts selling managed internet solutions, and it’s really taking off. Wouldn’t it be great if your costs grew at exactly the same rate as the sales? No “this upgrade will support another 1,000 customers” when you need to add the next five.
A new business line comes in, and, instead of OpenStack, it looks like VMware is the best solution. There shouldn’t be any unexpected expenses. Grow a customer at a time, and don’t worry about the underlying technology.
And get that single butt to kick to keep all the assurance running.
How does that sound?
Zenoss Delivers the NFV Dream Today
Zenoss is perfect for NFV deployments.
Starting out is easy. Zenoss runs and operates the central system. It’s a big data database, containerized architecture ready to scale to continent-sized needs, but all you need to get started is one VM. The Zenoss team continuously ensures your application is running well, provides backup/restore and disaster recovery services, and even installs new versions. You just need a browser.
That single VM understands all the components of the NFV infrastructure, will talk to individual VNFs, and track traffic on private tenant networks. It securely communicates to the Zenoss cloud where your data is private to you. By the way, there’s no data stored at the collector, and it’s completely configured remotely, so there’s not even a need to maintain the Zenoss VMs.
The Zenoss central database builds and maintains a model of the connections between your tenant service chains and the underlying infrastructure. If an OpenStack server has a problem, you’ll know exactly which tenant VMs need to be brought up elsewhere before you take the server down for repair.
Multiple sites? Put a VM in each site. High availability? Install a pair of VMs in a site, and the active-active design and automated recovery ensures you won’t drop a beat in a failure situation.
It comes with standard support for OpenStack, for VMware, for any VNF with SNMP. And if you do buy something new, or have to use something old, it’s easy to extend. Do it yourself or hire a specialist — most new devices take only a few weeks of work.
It’s practically trivial to integrate with any MANO for orchestration with the simple Zenoss JSON API. A few script lines, and when you provision a tenant or a service chain, their service is assured.
And the pricing model is straightforward and subscription-based. No big, lumpy surprises — just up and down to match your business.
When you’re ready to roll out NFV, Zenoss should be your first choice. If you want to try it out in your pod, contact us for a free demonstration.
