Operations
This section covers the runtime side of Husky after workflow authoring is complete.
Common operator tasks
- validate config before deployment
- start or stop the daemon
- inspect status, history, and logs
- reload config safely
- secure the API
- recover after crashes
- package Husky under system managers
Normal lifecycle
husky validate --strict
husky start
husky status
husky reload
husky stop
Foreground debugging
Use:
husky daemon run
This is the simplest way to observe scheduler behavior, retry timing, and log output directly in a terminal.
Runtime state
The daemon writes generated state under .husky/ by default. That directory typically contains:
- PID file
- socket
- SQLite database
- API address file
- daemon log file
Production-minded concerns
Husky supports runtime configuration for:
- bind address and base path
- TLS
- bearer or basic auth
- RBAC
- logging output and format
- SQLite path and retention
- executor pool size and shell
- metrics and tracing settings in daemon config