perlogger

Perlogger is a term used in software development to describe a logging component or pattern that produces logs on a per-unit basis, such as per HTTP request, per user session, or per operation. It is not a formal standard and may refer to different implementations in different projects.

The main purpose of a perlogger is to provide granular observability by capturing context and events tied to a specific unit of work. This facilitates debugging, performance monitoring, and auditing, and it often uses correlation identifiers to link logs across services. Perloggers typically accompany traces and contextual metadata to enable end-to-end analysis.

Typical features include per-unit log contexts, structured or JSON logging, tagging, support for different log levels, and pluggable backends (files, consoles, cloud logging). Some designs emit logs at the end of the unit or stream them to centralized systems while maintaining a unique scope or namespace for each unit. The approach aims to balance detail with manageability, often employing configuration and filtering to control volume.

Implementations vary. A perlogger may be realized as a library that creates a new per-unit logger for each unit of work, or as middleware that initializes a per-unit context in web frameworks. In batch or asynchronous systems, a per-task logger might be created for the duration of a job, ensuring logs stay associated with the correct operation.

Relations and variants. Variants include per-request loggers, per-user loggers, or per-task loggers. The concept is related to structured logging, log correlation, and sampling strategies to manage data volume and privacy concerns, as well as to how logs are collected and analyzed in a broader observability stack.

Considerations. Perloggers introduce overhead and data growth, so careful design is needed to avoid exposing sensitive information and to ensure proper integration with central logging and monitoring pipelines.