Skip to main content

Overview

Stalwart Mail Server provides detailed tracing and logging information to help users monitor and understand its behavior. The logging mechanisms can be configured to output to a file, standard output, or send tracing information to OpenTelemetry. It is possible to configure multiple loggers to output to different destinations, and the level of detail can be adjusted per logger. This makes it easier for users to diagnose issues and understand the inner workings of the software.

The available tracing mechanisms include:

  • OpenTelemetry: A vendor-neutral standard that provides a set of tools and APIs for collecting distributed traces and metrics. It enables seamless integration with various observability platforms, allowing for consistent monitoring across different systems and services.
  • Log Files: Traditional text-based files where Stalwart records detailed logs of system events and activities. These logs can be stored locally and accessed for troubleshooting, auditing, or analysis.
  • Journald: A system service for collecting and managing log data, particularly in Linux environments. Journald provides structured and centralized logging, making it easier to search, filter, and manage logs.
  • Console: The standard output and error streams where Stalwart can display real-time logs and messages. This is useful for debugging and monitoring the system during active sessions.

Tracing method​

In the configuration file, the method each tracer is defined with the the tracer.<id>.type attribute, where <id> is a unique identifier for the tracer. The supported tracers are:

  • open-telemetry: Sends tracing information to an OpenTelemetry collector.
  • log: Writes tracing information to a log file.
  • console: Prints tracing information to the console.
  • journal: Sends tracing information to the systemd journal. Only available on Linux/Unix systems.

The level of tracing information recorded is determined by the tracer.<id>.level attribute. The available levels are:

  • error: Log only the most critical errors that could prevent the system from functioning properly.
  • warn: Log warnings, which indicate potential problems that could impact the system's stability.
  • info: Log informational messages, which provide a general overview of the system's behavior.
  • debug: Used for debugging purposes and provides detailed information about the system's internal operations.
  • trace: This level is the most verbose logging level and provides an extremely detailed log of all events that occur in the system.
  • disable: Disables the tracer.

Tracers can be enabled or disabled by setting the tracer.<id>.enable attribute to true or false, respectively. When a tracer is disabled, no tracing information will be recorded by that tracer.