A run does not load what it cannot run. When a step or a table sync is left out, DataMind records the reason instead of loading anyway. This page lists the reasons you read in run logs and node lists, and what a restart of the same run repeats.
A production run executes a step only when the transformation is validated and its schedule is on. A test run must match the test that started it. When a step is left out, the run shows why:
Restarting a run does not begin from zero:
A restart continues the same run. To run every step again, including the ones that finished, start a new run.
A table sync that does not load shows as Skipped in the Sync Log, and its Note column gives the reason. The reasons you can read:
When several reasons apply, the Note joins them with a semicolon. See Run, stop and troubleshoot loads for the sync log and its statuses.
When a table has columns flagged as personal data, a production run also tokenises their values after the load finishes. A test run does not tokenise: it stops after the staging copies.
The step or table already finished in that run, so DataMind keeps its result and does not run it again. To make it run again, start a new run. Runs and their results are watched on Deploy, run, truncate and monitor tables.
A production run executes only a transformation that has passed validation. Test the transformation, finalize it, then run again, as Build, test and finalize a transformation describes.
The transformation is not on its schedule, so the run leaves it out. Turn the schedule on before the next run.
The same run continues. Steps that finished keep their result and are not repeated, and steps that never finished run again, including failed and skipped ones. Every restart raises the run's restart count.
Its Note in the Sync Log starts with Skip Reason: and lists the cause, such as Table Is Not Deployed or Node Already Run. Follow that reason, then sync again.
More than one cause can hold at once, and the Note joins them with a semicolon. Read the whole line, fix the causes it names, then sync again.
A scheduled run does not: it skips the table with Skip Reason: Synchronize Option Is Disabled. A sync you start by hand for a single table can still start. See Run, stop and troubleshoot loads.
No. A test run stops after the staging copies, so no PII tokenisation runs. A production run tokenises flagged values after the load finishes.
A step that depends on a failed or skipped step is skipped with Skipped Due To Previous Error. Fix the earlier step, then restart the run. Workflow runs report skipped nodes in their own words; see Workflows.
A test run was started against a different job, so it is refused. Start the test again from the transformation.