Currently, server_status_metrics.js expects the repl.apply.ops to always be the same as repl.apply.batchSize. This is not always true as we will skip applying operations where the optime is <= the beginApplyingOpTime.