Search docs.molo17.com

    Schedules and events

    Use Schedules in the Control Plane to automate pipeline operations or SQL. The screen has two views: Schedules for time-based jobs and Events for trigger flows started by a webhook or Core Hub event.

    How-to guide

    For a step-by-step walkthrough tested on a running Gluesync, see the how-to guide Automate pipeline actions with schedules, chained events, and webhooks.

    Chronos is the service behind this feature. You do not need to operate Chronos directly to create or run a job.

    Choose the workflow

    Need Use

    Run one action at a date, time, or recurring interval

    A Schedule

    Run several ordered actions from a schedule

    A schedule with Chained events

    Let an external system start a workflow

    An Event with its generated trigger URL

    React to a Core Hub lifecycle event

    An Event with a platform-event trigger

    Create a schedule

    1. Open Schedules from the left rail.

    2. In the Schedules view, set the timezone before creating the job. Existing schedules keep the timezone they were created with.

    3. Select New.

    4. Give the job a name.

    5. Choose the simple day-and-time picker or enter a cron expression.

    6. Select a pipeline. Select groups or entities only when the action should not apply to the whole pipeline.

    7. Select an action.

    8. Save the job.

    9. Run it once on demand and verify the result before enabling unattended execution.

    For recurrence syntax and the complete form reference, see Schedule jobs.

    Choose an action

    Action Effect

    Pause / stop

    Stops replication for the selected pipeline, groups, or entities.

    CDC

    Starts continuous change capture, with the selected initial-load behavior.

    Snapshot with truncate

    Clears the target selection, then loads the current source data with insert semantics.

    Snapshot without truncate

    Loads current source data with upsert semantics and preserves rows not replaced by the snapshot.

    Snapshot + CDC

    Runs the selected snapshot behavior, then continues with CDC.

    Put under maintenance

    Disconnects the pipeline for maintenance or configuration work.

    Resume from maintenance

    Reconnects a pipeline after maintenance.

    Query Studio

    Runs a saved query or custom SQL against one selected SQL-capable connector.

    Read Snapshot tasks before scheduling a destructive or large snapshot.

    Schedule SQL

    Query Studio actions are read-only by default. A write requires the explicit write acknowledgment in the job.

    1. Select the pipeline and SQL-capable connector.

    2. Select a saved query, or enable Use a custom query and enter SQL.

    3. Keep the read-only default unless the job is intentionally allowed to modify data.

    4. Save the schedule. Saving does not execute the query.

    5. Test the job and inspect its result.

    See Schedule Query Studio actions for saved-query resolution, write safeguards, and event use.

    Chain actions

    Add chained events when one scheduled job must perform several steps.

    Async

    Starts an action and immediately continues.

    Sync

    Waits for completion before starting the next action.

    Use sync only when a later step depends on the earlier result. Set the webhook timeout long enough for the slowest snapshot or pipeline action. See Chain actions.

    Create an event-triggered flow

    1. Open Schedules > Events.

    2. Create an event and add its actions in execution order.

    3. Optionally select a Core Hub platform event.

    4. Save the flow.

    5. Copy the generated secret token and trigger URL immediately. The token is shown once.

    6. Test with Fire now before enabling the external caller.

    Regenerating the token invalidates the old one. Store it in a secret manager, not in source code. See Build trigger flows and Webhook and platform-event behavior.

    Operate jobs safely

    • Use names that identify the target and action.

    • Test new jobs manually.

    • Check next-run time after timezone or recurrence changes.

    • Review execution status after platform updates.

    • Disable a job before editing its target or a dependent pipeline.

    • Prefer an event over frequent polling when an external system can signal readiness.