Scheduling & Job Management

BI Publisher's Scheduler (built on Quartz internally in most versions) lets a report be scheduled with recurrence patterns (once, hourly, daily, weekly, monthly, or a custom calendar), each execution capturing a specific parameter set, output format(s), and destination(s) — the same report definition can have multiple distinct schedules with different parameters (e.g., one schedule per region, each parameterized differently, run at different times to spread load).

Every scheduled execution is tracked in Job History, recording status (success, failure, warning), start/end time, and output location — critical for auditing whether a regulatory report actually ran and delivered on time. Failed jobs can typically be configured with retry logic (attempt count, delay between retries) and failure notifications (email an admin distribution list), and the Scheduler itself relies on an underlying Job Diagnostics/Scheduler database schema for persistence, meaning scheduler health is itself a monitored production dependency, not a fire-and-forget setting.

A senior interview angle: load distribution and scheduler capacity. Firing dozens of heavy reports at the exact same time (e.g., everything scheduled for "6:00 AM") can overwhelm both the Scheduler's processing capacity and the underlying database connection pool, so experienced BIP admins stagger schedule times and separate the Scheduler's own job-processing engine onto adequately sized infrastructure, sometimes even a dedicated scheduler cluster node, for high-volume enterprise deployments.