When you initiate a data backfill operation for a periodic node, pending instances and generated instances are created based on the scheduling type. You can manage these instances as needed, including viewing operation logs and rerunning operations.
Terminology
-
Pending instances: Instances scheduled to be generated and submitted for execution at a future time. Periodically scheduled data backfill nodes generate pending instances at 11:00 PM on the day before the scheduled runtime. For more information, see View and manage pending instances.
-
Generated instances: Executable instances created from immediate-execution data backfill nodes and pending instances. For more information, see Overview of data backfill instance operations.
Data backfill instance relationships
Different data backfill operations produce different instance types. The relationships between operations and instances are as follows:
-
One-time immediate execution data backfill node: Instances are immediately generated and scheduled to run.
-
One-time scheduled execution data backfill node: Pending instances are generated first. When the scheduled runtime is reached, data backfill instances are generated and scheduled to run.
-
Periodically scheduled execution data backfill node: Pending instances are generated at 11:00 PM on the day before the scheduled runtime. When the scheduled runtime is reached, data backfill instances are generated and scheduled to run.
-
Manually executed data backfill node: Instances are immediately generated and scheduled to run.
Data backfill instance entry
-
In the top menu bar of the Dataphin homepage, select Develop > O&M.
-
In the left-side navigation pane, select Instance O&M > Data Backfill Instances.