This article is based on a product announcement originally reported by Help Net Security in May 2020.
Overview
New Relic introduced New Relic Edge with Infinite Tracing, a fully managed, cloud-native distributed tracing solution built to help software engineering teams rapidly diagnose issues and latency across complex, microservice-based environments. The offering complements New Relic’s existing head-based distributed tracing capabilities, giving teams the flexibility to choose the tracing model that best fits their workloads.
The Problem: Observability in Microservice Architectures
The shift from monolithic application architectures to microservices has brought significant benefits — including greater team autonomy, faster deployment cycles, and better scalability. However, it has also introduced substantial operational complexity.
Modern systems can involve hundreds or even thousands of individual microservices, each with its own dependencies, frameworks, and deployment mechanisms. When a latency spike or error occurs, tracing the root cause across so many interconnected services becomes extremely difficult without the right tooling.
Many organizations rely on tail-based distributed tracing to address this challenge. Unlike head-based tracing — where a sampling decision is made at the start of a request — tail-based tracing evaluates the complete trace after it finishes, allowing teams to retain traces that are actually interesting (slow, erroneous, or anomalous) rather than sampling blindly up front.
The catch with existing tail-based solutions is that they traditionally require teams to deploy, manage, and scale their own on-premises tracing infrastructure. This creates operational overhead, increases costs for data transmission and storage, and pulls engineering effort away from product development.
What Is New Relic Edge with Infinite Tracing?
New Relic Edge with Infinite Tracing is designed to remove that operational burden by delivering tail-based distributed tracing as a fully managed service. Key characteristics include:
- Tail-based sampling: Observes 100% of a workload’s trace data before deciding which traces to forward, ensuring that high-value traces (errors, slow requests) are never dropped.
- Fully managed infrastructure: New Relic operates the tracing edge nodes — teams do not need to deploy, patch, or scale any additional software themselves.
- Cloud-native co-location: The edge nodes are located in the same cloud provider and region as the customer’s workloads, reducing data egress latency and costs associated with sending large volumes of trace data across regions or to on-premises collectors.
- Intelligent forwarding: After evaluating all trace data at the edge, only the most relevant traces are forwarded to New Relic’s core platform for storage and visualization.
- Visualization in New Relic One: Forwarded traces are surfaced in New Relic’s standard tracing UI, enabling teams to correlate distributed traces with metrics, logs, and other telemetry.
Head-Based vs. Tail-Based Tracing: Choosing the Right Model
New Relic Edge with Infinite Tracing does not replace the existing head-based distributed tracing option — it complements it. The two models serve different use cases:
| Attribute | Head-Based Tracing | Tail-Based Tracing (Infinite Tracing) |
|---|---|---|
| Sampling decision | Made at request start | Made after trace completes |
| Trace coverage | Sampled subset | Evaluates 100% of traces |
| Infrastructure overhead | Lower | Managed by New Relic Edge |
| Best for | High-volume, lower-complexity services | Complex microservice environments where errors must not be missed |
Teams can select the model that aligns with their environment’s complexity, data volume, and observability requirements.
Key Benefits for DevOps and SRE Teams
- Faster root cause analysis: Because tail-based sampling retains traces for slow and erroneous requests, engineers can investigate the exact transactions that caused incidents rather than hoping a sampled trace captured the problem.
- Reduced toil: No on-premises tracing collectors to deploy, monitor, or scale. The managed service handles capacity automatically.
- Lower data costs: By filtering at the edge before forwarding to New Relic’s core platform, teams avoid paying to transmit and store every raw trace span.
- Incident resolution before customer impact: Faster diagnosis means teams can detect and resolve latency issues and errors earlier in the incident lifecycle.
Availability
At the time of announcement (May 2020), New Relic Edge with Infinite Tracing was available for customers hosted in AWS US-East-1 (N. Virginia), with additional region support planned. Customers should consult the New Relic documentation for current regional availability, as this will have expanded since the initial release.
How Infinite Tracing Fits Into a Modern Observability Stack
Distributed tracing is one pillar of the broader observability model alongside metrics and logs. Infinite Tracing specifically addresses the tracing pillar’s hardest challenge at scale: retaining the signal (error and latency traces) while discarding the noise (successful, fast, representative traces), without requiring engineering teams to build and maintain the infrastructure to do so.
For organizations running complex microservice platforms on AWS — or planning to expand to additional cloud regions — New Relic Edge with Infinite Tracing offers a managed path to full-fidelity distributed tracing without the operational overhead of self-hosted alternatives.
Original announcement sourced from Help Net Security, May 2020. Quote attributed to Andrew Tunall, Senior Director of Product Management at New Relic.