Testing and service assurance must be active and automated to enable real agility, efficiency and great customer experience in emerging networks
The industry’s move towards virtualisation requires new automated approaches to active testing and service assurance. Here Marcus Friman, the chief product officer and co-founder of Netrounds, a specialist provider of automated active test and service assurance platforms with more than 270 customers, explains why the correlation between active and passive test data added to a DevOps style approach to testing processes can enable communications service providers (CSPs) to achieve true agility
VanillaPlus: How does Netrounds’ active test and service assurance platform enable CSPs to decrease operating costs and improve operating margins?
Marcus Friman: Active testing and assurance are used to improve the experience that customers receive and are a critical means by which CSPs can prevent churn, differentiate themselves and achieve higher rates of customer satisfaction. What we do is allow CSPs to automate the processes of service activation testing, quality monitoring, and troubleshooting, which provides a vital advantage in terms of also improving operational efficiency.
The key point is the automation because manual processes are too lengthy and costly to be conducted at CSP scale. Automation means CSPs can perform test and service assurance activities to a greater extent and they can do so in a much more efficient way.
VP: Can you explain more about the importance of active and automated assurance, especially in a virtualised environment?
MF: First of all, virtualised environments will be a lot more complex because there will be no standalone boxes and services will instead be delivered on top of shared resources. Secondly, one of the biggest causes of service problems is change. The ultimate goal is for CSPs to become more agile and in the future changes will occur much more frequently, making it even more imperative to use automated, active testing to ensure that services are delivered right. Without testing after a change has been made many customers could be affected before the problems are detected.
Active testing will therefore no longer be a nice to have – it will be a must have. The transition from physical to hybrid networks will be reliant on active testing and service assurance to ensure customer quality of experience is maintained.
VP: How will test and assurance fit together with orchestration in virtualised and hybrid – physical and virtual networks – network operations?
MF: This is obviously a shift in the industry and I think most current test and assurance systems in general aren’t ready for managing these kinds of environments. Test and assurance vendors need to develop suitable solutions that are application programming interface (API)-driven so their operations can be automated through orchestration in an OSS workflow.
In our case, the Netrounds active test and assurance platform was originally built as software with no hardware legacy, but other virtualised variants of traditional, hardware-based testing solutions will require manual configuration steps. Manual configuration is fine for a proof of concept but for commercial roll-out all the steps must be automated, everything from deploying test agents in the NFVI (NFV Infrastructure), making them available for testing, triggering the tests using an API (application programming interface), retrieving the results, and even un-deploying them when done.
To introduce automation in general – not just for virtualisation projects – brings important benefits for CSPs. Automation in legacy networks is highly valuable and CSPs should already today consider investing in automated and software-based active testing and service assurance to see cost and time savings benefits immediately while also positioning themselves effectively for transition to tomorrow’s virtualised environments.
VP: What do you see as the greatest issues test and assurance needs to overcome?
MF: There are three main issues. The first is automation, which I’ve already spoken about. The second is the capability to address both legacy and virtual environments – this isn’t an either/or situation and both should be addressed with a single solution. The third is the test and service assurance processes within CSPs.
CSPs need a process that is efficient for both of these environments in combination with orchestration – and this is a challenge. CSPs also need to move towards a DevOps type of workflow in order to cope with frequent change.
VP: What impacts are the industry’s move to real-time, active measurements having on traditional passive testing methods?
MF: Active and passive methods are complimentary so even though there’s a move now to fast-changing networks, both methods are still needed. The connection between active and passive methods will increase in the next couple of years. Making sense of data from passive solutions is hard because it is difficult to correlate it with the end user experience and a vast volume of passive data is generated.
Improving the analytical correlation between passive test and active test data is helpful in enabling CSPs to make greater sense out of the passive data. For example, if you are passively monitoring virtual network functions (VNFs) and devices such as the central processing unit (CPU) load and other metrics as part of an end-to-end service, just making sense of the data and how it affects the end user experience is next to impossible. However, if you complement it with active monitoring you can start correlating the active with the passive and see, for example, that service quality issues started at the same time as when the CPU load increased. This enables you to scale up CPU capacity and then make sure, with an active solution, that has addressed the issue. If it does solve the problem, you can learn from it and become more proactive in future similar situations.
The combination of passive and active test data is the future but if you leave out the active part it will be impossible to achieve service assurance in a satisfactory way.
VP: How is testing changing to move away from function specific hardware to new models such as cloud-based deployment?
MF: Active testing plays a supporting role to CSPs achieving real agility, operational efficiency and customer experience in new, emerging networks. The automation of manual processes needs to happen and testing needs to change to reflect this.
I think testing will be closer to an OSS automation workflow in the future. It will be tied into the orchestration loop and therefore completely different from current, manual testing which is tied to service creation. Testing will be a part of orchestration and very closely linked with the turning up and modifying of services.
Activation testing has been lower on the list of CSPs’ priorities because today it involves expensive truck rolls and it takes time. In addition, the business case hasn’t been clear until now but the need for agility and automation makes it mandatory. With all technologies there must be a problem to really drive uptake. For automated active testing and service assurance, the virtualisation transformation makes physical and manual processes unsustainable and creates a problem for CSPs. The time for a new approach is now and Netrounds is extremely well positioned as part of the NFV ecosystem to enable our CSP customers to achieve their transformation goals and provide excellent quality of experience to their customers.