Quick refresher: A service chain delivers a service to your customers, like SMS messaging or video transmission or email, and it is composed of several virtual network functions (VNFs). VNFs typically run in OpenStack virtual machines, which, in turn, run on the network functions virtualization infrastructure (NFVI).
Service providers spin up and down service chains in order to meet their customers’ needs. Your customer needs a distributed video service? Spin up as many video service chains as they need and add in another chain for the content delivery network. With potentially hundreds of service chains running at the same time in dozens of geographically distributed locations, ensuring that each customer is happy is a challenging task!
Service Chains and Infrastructure Both Need to Work
Actually, service chain success depends on two complementary factors. We need to know whether each service chain is working and delivering the expected quality, and we need to be able to identify the root cause as infrastructure, the VNF software, or some other element.
Netrounds and Zenoss have partnered to deliver this combination. The Netrounds active testing and monitoring solution uses software-based, small-footprint Test Agents, frequently delivered as VNFs, that provide both service activation validation and ongoing service chain monitoring.
Their solution focuses on end-to-end service chain performance and network service quality, all from the end-user perspective. That’s what your customers care about.
The Netrounds platform can run as a SaaS or on-premises solution in fixed and mobile networks. It uses active, traffic-generating Test Agents strategically placed throughout your network to monitor and measure key performance indicators (KPIs) in a multilayer and multidomain fashion. Virtual Test Agent (vTA) VNFs for NFV testing and monitoring are just one of the types of Test Agents available for use in your varied network topologies. Other Test Agent types include Software Test Agents and Embedded Test Agents, as well as the ability to use reflection-based technology to bounce traffic off built-in reflectors to other networking devices and equipment. Netrounds is also a Cisco SolutionsPlus partner.
The Netrounds team delivers on four key NFV needs:
- Validation of proper functioning of new and changed service chains after they are provisioned and go into production and before you have customers start using a potentially misconfigured service
- Distinguish whether infrastructure issues are actually affecting customers or if the alarms are for a failed backup or standby device that are not affecting the quality of your customers’ networks — this allows technicians and operations staff to prioritize and fix problems that are important rather than rushing to fix unused standby hardware not impacting customers
- Identify data plane problems like misconfigured routers, firewalls or load balancers
- Validate the continuous operation of a service chain from inside or outside the service chain in hybrid environments
These use cases complement Zenoss NFV monitoring perfectly.
Here’s an Example
Pictures are worth a thousand words, so let me show you how Netrounds and Zenoss work together.
When an orchestrator sets up a service chain, it can include Netrounds Test Agents. The agents validate initial operation of the entire chain and identify any failures that arise later. It performs the test when the service chain is first provisioned and in production, and continues to monitor with actively generated test traffic sent at regular intervals thereafter.
With Netrounds, you’ll know that the end-to-end service, and the service chain itself, was working when first put into production — and if something ever goes wrong, you’ll know exactly when the service fails to meet SLAs through alarms and notifications received in real time. This allows technicians and service staff to fix the problem, potentially even before customers experience any adverse service quality. Here at Zenoss, we call this “east-west monitoring.”
In the last article, we saw how Cisco Network Service Orchestrator will add new VNFs to Zenoss Resource Manager and create service organizers for each chain. From this, Zenoss maintains an infrastructure dependency model for each chain to provide a continuous root-cause analysis across the compute, storage, network and NFVI control infrastructure components. We call this “north-south monitoring.”
The north-south monitoring identifies the specific NFV infrastructure components supporting each element of the service chain, including any on-premises Netrounds agents. When we bring in alarms and notification events from Netrounds to the Zenoss event management system via APIs, we get an automatically combined east-west and north-south analysis.
Look at the picture above for a simplified example. Netrounds lets us know immediately when your service chain has failed. And looking at the detailed monitoring report, we have an idea of where the failure occurred — it is most likely the Parental Control VNF. Now, coupling that information with the Zenoss view, we see that one of the underlying pieces of infrastructure supporting the chain has an issue that is affecting the Parental Control VNF, and that could potentially also affect the Antivirus VNF. This knowledge lets us choose the proper recovery action.
- If there is no infrastructure issue, we can often recover by restarting the VNF, and Netrounds will verify that our recovery action worked.
- If there is an underlying infrastructure issue, we can ask the orchestrator to recreate the chain on working infrastructure, and Netrounds will again validate the recovery. Only then do we kick off the infrastructure recovery task.
Time to Get Started!
NFV has been an exciting challenge for Zenoss, and one we are thrilled to work on. We would love to address your gnarliest infrastructure monitoring problems, too. Talk to one of our experts, and learn how Zenoss can help you!