Scheduled Pipelines

Run pipelines on a schedule with cron expressions

Overview

Schedule pipelines to run automatically at specific times using cron expressions. Perfect for recurring tasks like daily model updates, nightly deployments, or weekly cleanups.

Creating a Schedule

In the UI

1. Open your pipeline
2. Click the Triggers tab
3. Click Add Schedule
4. Configure:
   • Name: “Daily Model Sync”
   • Cron Expression: 0 2 * * *
   • Enabled: Toggle on
5. Click Save

Using the SDK

from bud import BudClient

client = BudClient()

# Create a scheduled trigger
schedule = client.schedules.create(
    pipeline_id="pipe_abc123",
    name="Daily Model Sync",
    cron_expression="0 2 * * *",
    enabled=True
)

print(f"Schedule created: {schedule.id}")

Common Cron Patterns

Pattern	Description	Example Use Case
0 * * * *	Every hour	Hourly health checks
0 0 * * *	Daily at midnight	Daily backups
0 2 * * *	Daily at 2 AM	Nightly model updates
0 0 * * 0	Weekly on Sunday	Weekly cleanup
0 0 1 * *	Monthly on 1st	Monthly reports
*/15 * * * *	Every 15 minutes	Frequent monitoring

Managing Schedules

Pause a Schedule

# Pause temporarily
client.schedules.pause(schedule_id="sched_xyz789")

In the UI: Click Pause button on the schedule.

Resume a Schedule

# Resume execution
client.schedules.resume(schedule_id="sched_xyz789")

Trigger Manually

Run a scheduled pipeline immediately without waiting:

# Trigger right now
client.schedules.trigger(schedule_id="sched_xyz789")

Example: Nightly Model Deployment

Create a pipeline that deploys updated models every night at 2 AM:

from bud import BudClient

client = BudClient()

# Create pipeline
pipeline = client.pipelines.create(
    name="Nightly Model Deployment",
    definition={
        "steps": [
            {
                "id": "add_model",
                "action": "model_add",
                "params": {
                    "model_uri": "org/model-name",
                    "model_source": "hugging_face"
                }
            },
            {
                "id": "deploy",
                "action": "deployment_create",
                "params": {
                    "model_id": "{{steps.add_model.output.model_id}}",
                    "cluster_id": "cluster_prod"
                },
                "depends_on": ["add_model"]
            }
        ]
    }
)

# Schedule it
schedule = client.schedules.create(
    pipeline_id=pipeline.id,
    cron_expression="0 2 * * *",  # 2 AM daily
    enabled=True
)

Best Practices

Use Off-Peak Hours: Schedule resource-intensive tasks during low-traffic periods

Add Notifications: Configure alerts for schedule failures

Test First: Run manually before enabling the schedule

Monitor Execution: Check run history regularly

Set Timeouts: Prevent long-running schedules from overlapping

Troubleshooting

Schedule not running

Cause: Schedule is paused or cron expression is incorrect

Solution: Verify schedule is enabled and test cron pattern at crontab.guru

Executions overlapping

Cause: Previous execution hasn’t finished when next one starts

Solution: Increase schedule interval or optimize pipeline performance

Failed at specific time

Cause: Resource contention or dependency unavailable

Solution: Check execution logs, adjust schedule timing

Next Steps

Pipeline Concepts

Learn about DAGs and actions

Troubleshooting

Common issues and solutions