Today’s networks are too complex to manage domain by domain. It’s time for a unified view with unified assurance.
Imagine you’re the head of a crack team running video surveillance at a big department store. You have all the latest tools and security systems at your fingertips. Unfortunately, each department has its own separate system. Want to see what’s happening in Housewares? Just open this tool. How about Men’s Clothing? Swivel over to a different screen and open this other tool. Now Women’s Shoes? Swivel yet again—new screen, new tool. And by the way, each system is totally different, using different UIs, terminology and processes. So, if a problem crops up that spans multiple departments, hope you’re not in a big hurry to fix it.
Of course, this is ridiculous—nobody would design a surveillance system in such a fragmented way. It would take forever to investigate problems (not to mention dealing with the constant whiplash caused by all that swiveling). Unfortunately, it’s not that far from how communications service providers (CSPs) monitor and manage their networks today. One system for managing legacy physical devices. Another for virtualized networks. Another for SD-WAN. Another for NFV.
It’s not that operators set out to make their lives more complicated. It’s just that, as CSP networks have evolved, service assurance hasn’t really kept pace. The result is an increasingly fractured view of the network and the services running on top of it. How does this lack of visibility impact CSPs and their customers? And what kind of capabilities do they need to stay on top of their networks as they are today, instead of a decade ago? Let’s take a closer look.
CSP Networks Keep Getting More Complicated
Not long ago, service provider networks were (from the perspective of assurance, at least), fairly uniform. Every IP service was delivered over physical devices that operators could monitor and manage, usually with just a few tools. Today, CSP networks are a complex mix of virtualized network functions (NFV), legacy physical devices, software-defined wide-area network (SD-WAN) overlays, often provided on a customer-by-customer basis, and more. What customers view as a “service” may now traverse multiple domains and both physical and virtual devices. But within CSP organizations, each area typically has its own set of monitoring tools, with specialized experts supporting them.
As Anil Rao, principal analyst for Analysys Mason notes in Reimagining service assurance for NFV, SDN and 5G, “Existing physical networks will coexist with new NFV networks for the foreseeable future, creating a complex network environment and introducing a new dimension of assurance and operations complexity. New-age automated assurance systems must provide monitoring and operations automation capability for hybrid physical, virtual and cloud native networks and services.”
The Costs of Fractured Monitoring
Why is it imperative to gain a more unified view of the network? Because escalating complexity, paired with diminished visibility, leads to a host of negative business outcomes. As Appledore Research details in Rapid Automated Service Assurance in the NFV and SDN Network, running multiple assurance systems in “silos” leads to:
- High operating costs as networks require more and more specialized tooling and expertise
- Inability to quickly translate device- or domain-specific alarms into real-world business impact to customers
- Slow, inefficient workflows to detect, diagnose and repair issues
- Need to maintain multiple duplicate data sets and models, further increasing costs and complexity
- Higher license fees for tools and software dedicated to each separate domain
These issues can carry a significant cost. According to some analysts, service providers lose $11,000 for every minute of downtime while they try to piece together what’s happening, where, and which customers it’s affecting. And that doesn’t include the hit CSPs take to customer satisfaction, which can be even more significant. As Analysys Mason analyst Terry van Staden notes, “We know from our related business services research that customer satisfaction has a substantial impact on churn and cross-selling potential. Our data supports the common-sense assumption that satisfied customers are far less likely to churn and more likely to purchase additional services than those that are unsatisfied.”
Addressing this fractured network view is important for current networks. But in the near future,as you roll out 5G services, it will become absolutely essential. 5G places demands on the network unlike anything CSPs have had to contend with before: Exponentially higher traffic density. Up to 77,000% higher throughput. Latency requirements up to 6,000x lower than previous-generation services. In a 5G world, network issues that introduce just a few milliseconds of delays can render critical applications (autonomous vehicles, remote telemedicine, industrial automation) completely unusable.
Anatomy of a Modern Approach to Service Assurance
These are serious issues, but they’re not intractable. To fix them, you need a more holistic way to monitor and manage services across your entire environment—physical, virtual, SD-WAN and more—in one place. A modern approach to service assurance should enable you to:
- Bridge physical and virtual worlds: If you’re going to deliver better service experiences and meet more stringent SLAs, future service assurance tools should allow you to monitor and manage services traversing both physical and virtual domains through a single solution. It should provide a unified view centered on your customers and their services—not on the various infrastructure domains.
- Visualize everything: In today’s sprawling, dynamic operator networks, just knowing what’s out there is a huge challenge. A next-generation service assurance solution should automatically discover the topology of the entire multivendor network—including the transport, physical, virtual and services layers. It should automatically recognize when something in the network changes and update its relational map, so you’re never working from out-of-date information.
- Manage all networks through a single pane of glass: Next-generation assurance solutions should integrate service monitoring and network management end to end. That means correlating devices such as hosts, switches and routers with VMs, NFV, SDN and SD-WAN environments. It’s this correlation that empowers operations teams to identify faults and performance issues quickly. It’s also a necessary prerequisite for systems to respond to issues automatically and remediate the actual service rather than just a portion of the network.
- Support multi-tenant, multivendor environments: Just as you don’t want to have to rely on different tools for different network domains, you don’t want to have to use different tools for different customers. Instead, you should be able to monitor and proactively manage multiple customers, even with diverse environments, in one place. Your operations team should be able to visualize, analyze and optimize your environment to accelerate resolution times, assure high availability and meet stringent SLAs.
Fortunately, capabilities like these are no longer science fiction. You don’t have to rely on network monitoring and management tools built for an earlier time. It’s time to step up to holistic, multi-layer service assurance.
To learn how VMware can help, visit us at VMworld 2019 or: VMware Smart Assurance
To learn how Dell EMC can help, visit: Dell EMC Service Provider Solutions
Blog by Karina Dahlke, Telco and Edge Cloud Business Unit, VMware