Tracing is the de facto approach to understanding application behavior for many engineers. Yet beneath the surface, tracing is another form of an age-old approach: logging. The symbiotic relationship between tracing and logging offers the possibility to deliver otherwise unseen insights into our applications’ performance and health.
Both tracing and logging capture events that detail your application’s journey. Rather than treat traces as a standalone entity, adopt the mindset that traces are an extension of the familiar logging paradigm. Integrating traces into your existing logging infrastructure gives you the power to query, filter and transform trace events just as you do logs.
Treating traces as logs lets you apply the power of our APL (Axiom Processing Language) to trace data. It also lets you integrate traces into your traditional logging practices, making trace data as actionable as log data.
Moreover, this integration reveals the potentially sophisticated interplay between traces and waterfalls. By extracting embedded data from span attributes or span events, you can generate results just a click away from the waterfall visualization.
A simple example
Let’s look at a scenario where live streaming of spans, operations, and requests can be dynamically filtered down to a single service or a specific service version. Notice how Axiom’s APL query language can extract the information we want from span attributes at query time, rather than requiring us to reformat how that data was sent to Axiom before storing it. Axiom keeps all our raw data, and lets us perform transformations as needed as part of whatever queries we later want to run on it.
Treating traces and logs as unified information gives you much greater agility to pinpoint bottlenecks, identify latency issues, and respond proactively to potential hiccups – all while your application is in motion.
Use Axiom for all your event data
Logs, traces and other event data are no problem for Axion to capture, transform and analyze. It captures OpenTelemetry data at full fidelity. It can transform data captured in its raw form at query time, so you needn’t lose your original event formats to interact with emerging standards.
Our new pricing starts as low as $25 per month, free for personal projects. No surprise bills, ever. Contact us today to get started: email@example.com