Clicked Save, but no new change-history row?
Suppose you open a job, inspect its schedule, and press Save without changing a single setting. The job has not necessarily run and its configuration has not changed. Should this be recorded as a new configuration change?
The answer depends on what you are measuring.
Configuration history is not execution history
A configuration-history row describes how a job's settings changed over time. A run-history row describes an actual execution of the job. Confusing the two can make routine administration appear to have changed a workflow.
For Databricks Lakeflow Jobs, system.lakeflow.jobs and system.lakeflow.job_tasks describe job and task configuration history. They are not interchangeable with job-run timeline records.
The October 19, 2026 change
Databricks' official October release notes state that saving a Lakeflow Job without changing its settings no longer appends rows to system.lakeflow.jobs or system.lakeflow.job_tasks. The change_time remains at the previous real configuration change. This applies to Save operations through the UI, Jobs API and Declarative Automation Bundles.
This does not mean the Save control never performs work, that a job run cannot occur separately, or that every workspace will receive the change on exactly the release date. The vendor notes that releases are staged.
A simple hypothetical timeline
| Time | Action | What to inspect |
|---|---|---|
| 9:00 AM | A job's schedule is truly modified | Configuration-change history |
| 10:00 AM | Save is pressed with no settings changed | No-op save: no new configuration-change row under the new behavior |
| 11:00 AM | The job runs on schedule | Job execution history, separately |
The actions and times above are invented educational examples, not observations from any Databricks workspace.
Takeaway
Before calling something a "job change," ask whether the saved configuration actually differed. Track configuration and job execution in their respective data sources, especially when evaluating operational trends and audit trails.
Primary sources
Evidence limits: This educational article interprets manufacturer documentation. Examples are hypothetical, not measured tests on production equipment, databases or customer workspaces. It reports no measured detection, recovery or processing outcome. Operational changes require authorization and the documentation for the actual product and version.