The execution pool size acts as a threshold. When the number of job requests exceeds the threshold, the Data Integration Service stops accepting new jobs.

When the number of concurrent jobs exceeds the native and Hadoop execution pool threshold, the Data Integration Service stops taking jobs from the queue until the number of concurrently running jobs falls below the threshold you configured.

When the number of concurrent jobs exceeds the on-demand execution pool threshold, the Data Integration Service rejects on-demand job requests. When the number of concurrently running jobs falls below the threshold you configured, you can resume running on-demand jobs and re-run rejected jobs.

Configure execution pool sizes in Data Integration Service Execution Options settings.