Task scheduler

Right-click a database and choose Task Scheduler to run a query or a backup on an interval.

It only runs while NabuSQL is open

The scheduler is part of the application: it checks every 30 seconds whether a task is due, and runs the ones that are. Close NabuSQL and nothing fires. It is a convenience for a machine you keep the app running on — not a replacement for cron, systemd timers or the Windows Task Scheduler for anything production depends on.

Creating a task

Field	Meaning
Name	How the task is listed
Connection and Database	What it runs against
Action	Query or Backup
SQL	The statement, for a query task
Output path	The target file, for a backup task
Interval (minutes)	How often it runs — shown as hours when it is 60 or more
Enabled	Off pauses the task without deleting it

The task list

Each task shows its interval, when it last ran, and the outcome of that run — OK: with the result (rows affected, or where the backup was written) or ERR: with the server's error message. That is where you look when a task quietly stopped working.

Run now executes a task immediately, independently of its schedule. Use it to confirm a new task works rather than waiting for the interval to come round.

How due tasks are decided

A task is due when the time since its last run is at least its interval. There is no catch-up: if NabuSQL was closed for two days, the task runs once at the next check, not forty times.

Tasks are stored in scheduled_tasks.json in the application-data directory, so they survive restarts.

Sensible uses

• Refreshing a materialised summary table on a development machine.
• A periodic backup of a local database you are actively changing.
• A cleanup DELETE on a scratch database.

For a scheduled production backup, use your operating system's scheduler with mysqldump or pg_dump. A backup that depends on someone leaving a GUI open is not a backup you can rely on.