End-to-End Request Tracing
When something in a business platform is slow or behaves unexpectedly, the question is always: what actually happened? A user clicks save. Things happen. Something somewhere takes longer than expected or produces an unexpected result. Without visibility into those middle steps, diagnosing the problem is guesswork.
End-to-end request tracing provides that visibility.
What a Trace Contains
A trace captures the full lifecycle of a request: from the moment the browser sends it to the moment the response is returned. Inside that lifecycle, every significant operation is timestamped:
- Which database queries ran, how long each took, how many rows they returned
- Which cache lookups hit and which missed
- Which definition resolutions ran and what they produced
- Which workflow steps were evaluated (for actions that trigger workflows)
- What validation ran and what it returned
- Where time was spent in each processing stage
The trace shows you not just "it took 800ms" but "it took 800ms because these three database queries took 650ms combined, and this query here ran without using an index."
Who Uses Traces
Workspace administrators investigating slow pages or unexpected behavior in their specific workspace, using the production profiling feature to capture traces for specific operations.
During setup and testing, when a complex configuration produces unexpected results and you need to understand exactly how the platform resolved it.
For performance optimization, identifying which operations are genuinely slow and targeting optimization effort at the actual bottleneck rather than guessing.
Trace Access
Traces are accessible from workspace settings for users with administrator access. They're enabled per-request (not for all traffic), and the captured data is visible only to workspace administrators.
The Diagnostic Value
The alternative to tracing is hypothesis-driven debugging: guess what might be slow, test that hypothesis, repeat. Tracing is evidence-driven: look at what actually happened, find the slow part, fix the specific thing.
Evidence-driven debugging is faster and more reliable. That's the value.