System and Operations

System and operations keep the service stable, configurable, observable, and recoverable. This area spans startup configuration, database-backed runtime settings, background tasks, health checks, logs, mail, audit, and CLI operations.

Capability Boundaries

Capability	Notes	Related docs
Startup configuration	Listen address, database, logging, WebDAV prefix, node mode, rate limiting, cache	Configuration Overview, Server, Database
Runtime system settings	Admin-console hot settings for site, registration, mail, WOPI, trash, task retention	System Settings
Background tasks	Thumbnails, archive jobs, migration, offline download, cleanup, periodic jobs, retries	System Settings, Operations CLI
Mail	SMTP, templates, outbox, test mail	Mail
Monitoring	Health, readiness, Prometheus metrics, Grafana dashboard	Monitoring and Grafana
Audit	Admin operations, team audit, filterable audit logs	Admin Console
CLI	doctor, offline config, node enroll, cross-database migration	Operations CLI
Backup and upgrades	Database, config, local upload directories, version upgrades, rollback	Backup and Restore, Upgrade and Version Migration

Backend Modules

Module	Owns
config::loader, ops::config	Static configuration loading and runtime system settings
runtime::startup, runtime::tasks	Primary/follower startup and periodic background tasks
task	User-visible background tasks, scheduling, retries, cleanup
mail::sender, mail::outbox, mail::template	Mail delivery and templates
ops::health, api::routes::health, metrics	Health checks, readiness, metrics
ops::audit	Audit log recording, querying, presentation
cli::*	Offline operations commands

Configuration Boundaries

• Anything required before startup belongs in config.toml or ASTER__... environment variables.
• Administrator-adjustable runtime behavior belongs in the database-backed system_config table.
• System default definitions are centralized in src/config/definitions.rs.
• New user-visible background tasks should use task::create_task_record() and wake the dispatcher.

Troubleshooting Direction

• Service does not start: check config path, database connection, occupied ports, directory permissions.
• Readiness fails: check migrations, default policies, node mode, follower binding state.
• Background tasks pile up: check dispatcher state, retention settings, failure reasons, retry count.
• Admin console is unavailable but config must change: use the config subcommand in Operations CLI.