A checkpoint is a job restart point created by a Teradata PT job. Should the job be interrupted for any reason, it can be restarted from the checkpoint instead of from the beginning of the job. Checkpoints help guarantee that the work performed by an interrupted job up to the checkpoint will not have to be redone.
When a Teradata PT job takes a checkpoint, the producer operator in the currently-executing job step stops putting rows in the output data stream and the consumer operator processes all the rows in the input data stream, committing Teradata Database updates or flushing output file buffers. All executing operators write records to the job checkpoint files with the information that will allow them to resume their processing, with no loss or duplication of data, at the point the checkpoint was completed.
Teradata PT offers the following options related to checkpointing:
- Specify an alternate location for checkpoint directory.
- Specify a user-defined checkpoint interval.
- Specify a limit to the number of times a job will automatically restart.