@Naveen Kumar Kakarla (Quadrant Resource) Thanks for posting your question in the Microsoft Q&A forum.
pg_cron leverages PostgreSQL's Background Worker Processes. And this extension configures its bgw_restart_time to 1 second. That means if that process crashes, postmaster will wait for 1 second and spin up another instance of the process, which will go through the same initialization code as when it is first run after a PosgtreSQL restart. Meaning that it will execute the entry-point function PgCronLauncherMain, and that one will call MarkPendingRunsAsFailed, which will update all rows in the cron.job_run_details whose status is STARTING or RUNNING to have their status set as FAILED and return_message a 'server restarted'.
Bottom line being that you can see rows marked like that when PostgreSQL itself has been restarted, or when the background worker for pg_cron has crashed.
An alternate approach to START/STOP Flexible Server Instance on schedule is to leverage built-in Azure Automation Tasks based on Azure Logic Apps. This method offers easy scheduling via the Azure portal and provides reporting in the form of task history.
Manage Azure Database for PostgreSQL - Flexible Server using automation tasks
Regards
Geetha