“Sticks and stones will break my bones, but words will never hurt me.”
It was always the comeback line delivered to a schoolyard bully as a child, but it may not be accurate for adults in the networking world.
The term "network observability" won't break any bones, but it could hurt your chances of buying the right network tools. Without a unified consensus about its meaning, NetOps pros might not get the best solution to tackle problems like failed applications and end-user frustration.
The term's meaning remains fuzzy, and the consensus is still elusive, but a solid definition is closer than ever. A new research report from EMA titled "Network Observability: Delivering Actionable Insights to Network Operations" combines quantitative market research with qualitative interviews to establish an authoritative definition.
As more vendors use the term, confusion about its exact meaning has grown. The latest research from EMA surveyed 402 network professionals and spoke one-on-one with several of them to develop a definition for a product capability dubbed network observability.
“Tell me what to do to fix my network problems.” That's what IT pros want from a network observability solution.
Every IT organization monitors its networks and analyzes various data to help troubleshoot and understand where and why network problems occur. At some point, though, monitoring is an accumulation of data. Observability is more. Network observability represents an evolution of network monitoring.
Network observability is a network monitoring system that collects a complete and diverse set of network data to provide deep visibility and actionable insights into a network's current and future state. Those actionable insights include network performance, application performance, network security, and end-user experience.
On the frontlines of network operation, network observability could involve using a single tool. It could also include several tools linked together via integration and a data lake. The latest research provides IT pros and vendors with a defined blueprint for total network observability.
Those two simple words grew out of the DevOps revolution, in which observability had a broad meaning to include the state of applications and infrastructure.
NetOps teams are worried about much more. User problems often manifest as a "network problem," so the ability to pinpoint the source of the problem or say, "it's not me" is critical. A mound of monitoring data doesn't make it easier to achieve a quick MTTR if there's no way to understand the location of the problem and what is causing it. A pile of data about packets and flows, combined with device logs and configuration changes, is not enough.
That is where a network observability system enters the conversation. It must monitor and analyze the network's state using a diverse and ever-growing data set. The term "network observability" shows that vendors are using the word to offer a new wave of innovation that moves beyond network monitoring and network performance management (NPM).
Network observability needs to offer actionable insights. It can come in various ways (and comparing the ways multiple vendors do this can be interesting and fun), but it will increasingly use AI and machine learning.
Before now, network observability could mean almost anything without a definition that both sides stipulate its meaning. With a unifying definition, vendors can move on to the core features and functionality where their solution excels and is different from competitors.
A fair question is whether any vendor – even one – today offers network observability. IT pros should look for the ability to provide a deep dive into their networks and collect more diverse and voluminous data to get end-to-end visibility into operations.
It is not just NPM. Network observability solutions should be broader in scope. They should not just look at network performance but also include information about the end-user experience, business impacts, and security.
Robert Gates is Senior Analyst, Network Management, at Enterprise Management Associates, which is hosting a Nov. 9 webinar about the firm’s network-observability research.
Related articles:
Subsea cable alternatives can provide a level of comfort for those concerned about disruptions. Realistically, they will continue to play a small, but critical role in the short term.
Maintaining existing network infrastructures often incurs huge technical debt before newer technologies are adopted over time.
Amid ongoing regional uncertainties, telecom infrastructure in the Middle East has expand to include terrestrial low-latency network infrastructure, which offers resilience and agility in navigating political complexities without relying solely on undersea networks.
