After parsing and aggregating the size and count deltas in the oplog entries from the last metadata checkpoint timestamp, read the existing size and count info from fast_count_metadat_store and calculate the new total size and count for all dirty collections. These values will be used in the writes by the new metadata checkpoint.