Interval scheduling

Interval scheduling runs missions at fixed time intervals.

Syntax

schedule: every N units

Time units

Unit	Examples
seconds	every 30 seconds
minutes	every 15 minutes
hours	every 6 hours
days	every 1 days
weeks	every 1 weeks

Examples

Seconds

// Every 30 seconds (for real-time sync)
schedule: every 30 seconds

Minutes

// Every 5 minutes
schedule: every 5 minutes

// Every 15 minutes
schedule: every 15 minutes

// Every 30 minutes
schedule: every 30 minutes

Hours

// Hourly
schedule: every 1 hours

// Every 2 hours
schedule: every 2 hours

// Every 6 hours
schedule: every 6 hours

// Every 12 hours
schedule: every 12 hours

Days

// Daily
schedule: every 1 days

// Every 2 days
schedule: every 2 days

Weeks

// Weekly
schedule: every 1 weeks

// Bi-weekly
schedule: every 2 weeks

Interval vs cron

Interval

• Runs at fixed intervals from start time
• Simpler to configure
• Less precise timing

schedule: every 6 hours
// Runs at: start, start+6h, start+12h, ...

cron

• Runs at specific times
• More control
• Predictable times

schedule: cron "0 */6 * * *"
// Runs at: 0:00, 6:00, 12:00, 18:00

Start time

Intervals start from:

1. Daemon start time (for new missions)
2. Last run time (for existing missions)

To run once immediately rather than waiting for the daemon to reach the next interval, start the daemon with --once, which runs every scheduled mission a single time and exits.

Combining with options

Schedule options go inside the optional { } block on the schedule, not as mission-level fields.

With retry

mission RobustSync {
  schedule: every 15 minutes {
    retry: {
      maxRetries: 3,
      delaySeconds: 60
    }
  }
}

With concurrency control

mission ControlledSync {
  schedule: every 5 minutes {
    skipIfRunning: true  // Don't overlap
  }
}

Use cases

Real-time sync

mission RealtimeSync {
  schedule: every 30 seconds

  action Sync {
    get "/events" { since: lastSync }
    store response -> events { key: .id }
  }
}

Hourly updates

mission HourlySync {
  schedule: every 1 hours

  action Sync {
    get "/data"
    store response -> data { key: .id }
  }
}

Daily reports

mission DailyReport {
  schedule: every 1 days

  action Generate {
    get "/stats/daily"
    store response -> reports { key: formatDate(now(), "YYYY-MM-DD") }
  }
}

Weekly archive

mission WeeklyArchive {
  schedule: every 1 weeks

  action Archive {
    get "/records/stale"
    store response -> archive { key: .id, upsert: true }
  }
}

Best practices

Choose appropriate intervals

Data Type	Recommended Interval
Real-time events	30 seconds - 5 minutes
Transactional data	5-15 minutes
Reference data	1-6 hours
Reports	Daily
Cleanup jobs	Weekly

Account for execution time

// If sync takes 10 minutes
schedule: every 15 minutes  // Good: 5 minute buffer

// Not:
schedule: every 5 minutes   // Risk: overlapping runs

Use skipIfRunning

mission SafeSync {
  schedule: every 5 minutes {
    skipIfRunning: true
  }

  action Sync {
    // Long-running sync
  }
}

Troubleshooting

Runs too frequently

Check interval unit:

// This runs every 30 SECONDS
schedule: every 30 seconds

// This runs every 30 MINUTES
schedule: every 30 minutes

Runs overlapping

Add skipIfRunning to the schedule block:

schedule: every 5 minutes {
  skipIfRunning: true
}

Missed runs

Intervals don't backfill. If the daemon was down for 2 hours with a 30-minute interval, you won't get 4 runs.

Consider:

• Adding catch-up logic
• Using incremental sync with since: lastSync