We should use the long running clusters to run schedules with compactions! Jira issue: CRDB-55423 Jira issue: CRDB-55529