Operational logging and support evidence
Overview
Section titled “Overview”The application records safe operational telemetry and durable business audit history for different purposes. HTTP telemetry tracks generated request identifiers, matched route templates, status and duration. Worker errors use bounded categories. Business histories record actor and operation context. These sources help diagnose incidents without routinely recording raw secrets or full request bodies.
Purpose
Section titled “Purpose”Use this article to prepare a useful support case and preserve operational evidence without exposing authentication or customer data.
Prerequisites
Section titled “Prerequisites”-
Authorized access to the relevant application history or service logs.
-
An incident time, timezone, affected operation and product release.
Procedure
Section titled “Procedure”-
Reproduce the issue once if safe, then record the visible safe error and the time. Avoid repeated mutations when the first outcome is unknown.
-
Inspect the record, document, import, communication or workflow history appropriate to the issue. Note the last successful state and failed attempt.
-
Use System Health to record worker and queue context around the same time.
-
Collect sanitized service logs through the installation’s documented log command. Retain request identifiers and error categories; remove customer content, tokens, reset links and configuration values.
-
Escalate with a concise reproduction, expected and actual result, release identity, affected module and redacted evidence. Keep internal identifiers in the private support channel only when needed.
Expected result
Section titled “Expected result”Support receives enough context to correlate a failure while sensitive content remains protected.
Important notes
Section titled “Important notes”-
The application server uses app.server to install safe Uvicorn logging. Package service definitions should retain that entry point.
-
Proxy access logs deliberately avoid raw request targets, and raw proxy error logs may be suppressed. A failure before the API can therefore require gateway-level diagnosis.
-
Telemetry buffers are bounded during database outages; prolonged outages may lose older detailed counters.
-
Audit history is application evidence, not a claim of an externally certified or database-administrator-proof audit system.
Troubleshooting
Section titled “Troubleshooting”-
No application request appears for a failed page load: check the gateway/network layer and correlate the time.
-
A generic server error provides little detail: retain its request context and inspect safe service categories rather than enabling uncontrolled payload logging.
Related articles
Section titled “Related articles”Search the guides or talk to the ZBIX team.
Software products, application engineering and IT services from Zbix Technologies.
Visit our company website ↗