This is the maximum number of fault records UCS Manager will retain under this policy's scope before it starts discarding the oldest ones. An admin running a large domain with lots of transient hardware noise sets this limit to control how much fault history accumulates, since without a cap a domain with many chassis could build up an unmanageably large fault log that slows down querying. A size limit of 500 means once 500 fault records exist under this policy's scope, the oldest cleared faults start getting purged to make room for new ones, even if their configured retention interval hasn't fully elapsed yet.