Issue Summary

Similar to JRASERVER-62072, scheduled Jobs will be lost track of when a database operation fails at the moment the job is claimed by the Scheduler. When connectivity resumes, these jobs will not run again. This seems to be a regression as the Caesium Recovery Job does not bring jobs back to their functioning state.

This seems to affect both Jira Server and Data Center but a DC instance can recover from this after time.

Steps to Reproduce

1. Schedule a Job such as a mail handler.
2. Interrupt database connectivity to the application.

Expected Results

Jira will run the job as soon as connectivity is established back between the database and application.

Actual Results

Jira will not run the job until a restart is performed.

Workaround

Restarting Jira will allow jobs to resume their functions.