Data Sync Job Operations

Data Import Job Overview

Once the job is configured, manage it on the platform with the following operations: start, pause, resume, restart, and delete.

• Start the job by changing the status to "Initializing."
• After a successful startup, the status changes to "Running."
• Pause the job in both "Initializing" and "Running" states, changing its status to "Paused."
• If paused successfully, the job can be restarted or resumed. Note: For synchronization jobs on the Postgres source side, if they remain paused without resuming, the disk space used by the WAL LOG on the source PG will continue to increase.
• "Resume" retains the previous job status and continues the synchronization process.
• "Restart" resets the previous job status and begins synchronization anew.
• Only full synchronization jobs will "succeed" or "fail," while jobs involving incremental synchronization will continue to run.
• If the job is not active (never "started," "paused," "successful," or "failed"), it can be deleted. Upon deletion, all status and history information will be wiped clean.

Only fully synchronized jobs will display a success or failure status. Incrementally synchronized jobs will keep running, allowing you to monitor metrics like latency, RPS, etc.

Job History

Review the job instance detailed below. A new instance is created every time it is initiated.

Job Indicators

The Dashboard presents six indicators, categorized into the following tables:

• Idle Time: The millisecond gap between the current system time and the last data entry point for synchronization. This value will persistently rise without data coming from the source.
• Emit Event Time: The millisecond gap between the last data processing time and the most recent data entry into synchronization. This value will not increase if there is no data from the source.
• Instance Records: The total count of records the job has processed.
• Instance Bytes: The total number of bytes handled by the job.
• Instance Rps: The job's records processed per second (RPS).
• Instance BPS: The job's bytes processed per second (BPS); to determine if incremental jobs are delayed, refer to the Idle Time and Emit Event Time indicators.

Notices:

• Idle Time is minimal, suggesting no delay.
• Idle Time is extensive, while the emission event time is also tiny and remains unchanged, indicating a lack of data at the source.
• Idle Time and Emission Event Time are extensive (essentially equivalent), suggesting a significant job delay.