Can we limit the size of OSQuery log written locally so that once fleet manager is back online after downtime will not be dumped with too much logs