Dataphin:Runtime configuration

Scope

These settings apply only to runtime instances with the attribute set to system configuration.

Permissions

Only super administrators and system administrators can modify runtime configurations by default.

Procedure

1. On the Dataphin homepage, choose Development > Task maintenance.

2. Access the Global Configuration page.

   Choose Global Configuration > Runtime Configuration.

   

3. Click Edit to modify Timeout Configuration and Rerun Configuration.

   • Timeout configuration

     Tasks that exceed the runtime threshold are automatically stopped and marked as failed. You can set thresholds for these task types:

     Task type	Description
     One-time Run	Includes ad hoc queries, trial runs of computing tasks, smoke testing of logical tables, and trial runs of integration tasks. Default: 2.00 hours.
     Recurring Task	All tasks on a recurring schedule. Default: 48.00 hours.
     One-time Task	Covers one-time instances for computing, integration, synchronization, and tag test executions. Default: 48.00 hours.

     Note

     Valid range: 0.01 to 168 hours (up to 2 decimal places).

   • Rerun configuration

     Rerun settings control whether tasks automatically retry after failure:

     • Automatic rerun on failure

       • Rerun count: Number of automatic retries after failure. Range: 1–10. Default: 1 time.

       • Rerun interval: Interval between retries. Range: 1–60 minutes. Default: 5 minutes.

     • Automatic rerun on timeout failure

       When enabled, tasks terminated for exceeding the runtime threshold automatically retry using the automatic rerun on failure settings. Default: disabled.

4. Click Confirm.

5. In the confirmation dialog box that appears, click Confirm System Configuration Modification, then click Confirm.

   Important

   After you save the changes, all instances set to system configuration (including new instances and those not yet started) use the updated settings. Changes take up to 5 minutes to take effect.

Recommendations

Use these guidelines to set values based on your workload:

• One-time run: Maximum average runtime of ad hoc queries, computing task trial runs, logical table smoke tests, and integration task trial runs for the tenant, plus one hour.

• Recurring Task: Maximum runtime of all recurring tasks for the tenant, plus one hour.

• One-time task: Maximum runtime of all one-time instances (computing, integration, sync, and tag test runs) for the tenant, plus one hour.

• Rerun count: Number of successful automatic reruns across all recurring, one-time, and data backfill instances for the tenant, plus one.

• Rerun interval: Longest successful rerun interval across all recurring, one-time, and data backfill instances for the tenant, plus three minutes.