index=_internal sourcetype=scheduler status=skipped | timechart span="5m" count by savedsearch_name
Skipped typically means a scheduled search did not finish before its next scheduled run should start so that next run is skipped to avoid loading up the queue infinitely. So a good starting place is the searches that are being skipped. Do everything you can to optimize THOSE searches. Consider running them less frequently or over a smaller time range. Consider adding more indexers. Run the Healthchecks on the Monitoring Console and fix KNOWN system problems, etc. Above all else, DO NOT run scheduled real-time searches.