Detailed text description: Log browser architecture
This page describes the log browser architecture graphic.
Overview
The graphic shows how application and pipeline logs are collected, processed, and displayed in the Log Browser.
Main stages
- Log sources:
- Service runtime outputs (stdout and stderr).
- Pipeline and execution-step log streams.
- Collection and ingestion:
- Logs are collected by the platform observability pipeline.
- Metadata is attached, such as service, replica, and execution context.
- Storage and retention:
- Logs are stored centrally with a retention window.
- Older entries are purged or overwritten according to retention and quota rules.
- Visualization:
- Users query and filter logs in the Log Browser UI.
- Filtering can use fields such as service, severity, and time range.
Key takeaway
The architecture centralizes multiple log streams into one searchable interface for operational debugging.