Skip to main content
Schedules live next to the thing they refresh — there is no separate scheduling surface.

​
Source schedules

Each source’s load schedule lives on the source’s Schedule tab (Integrations → open the source → Schedule). A scheduled load uses everything saved on that source: connection, target schema, refresh mode, table selection, and per-table cursors.

​
dbt schedules

dbt schedules live in Models → Schedule (Add DBT Schedule), scoped to the environment you are viewing. Choose a layer or a specific model scope and a cadence; each scheduled run lands in Monitoring → Runs.

​
Cadence options

Both schedule types support basic cadence controls and cron expressions. Plain-English cron assistance can generate a cron expression before saving.

​
Where runs show up

  • Scheduled dbt runs — the complete ledger is Monitoring → Runs; Models keeps the schedule and manual-run controls.
  • Scheduled source loads — the complete ledger is Monitoring, with source-local context also available on the source.
  • Failures and in-flight work remain visible there, and failures trigger notifications when a channel is configured.
Schedules execute inside Platinur’s worker on an in-process scheduler — there is no separate scheduler service to operate or monitor. Expected occurrences and actual claims are stored durably. If the worker is offline at a due time, Platinur records the occurrence as missed instead of inferring success from a missing run. v1 does not automatically replay missed source or dbt work after recovery; the next configured occurrence runs normally, and you can start an immediate manual run when catching up is urgent. Scheduled work uses the same workload capacity and source-exclusion controls as manual work. A source schedule that overlaps an active sync is recorded without starting a duplicate load.