Dear all, We were writing a code to reduce the overhead metadata management according to increasing number of files for a given directory. This work includes works for a wrapper for read API (provides accessing multiple records with 1 read command), directory cache problem, and reverse hash table for managing record info. Now, we can get 25-30% performance improvement against previous work, but this result is only for wrapper and directory cache issue. In this week, we will be continuing this work. Sincerely, Jongmin -- Jongmin Gim Post-Doctoral Research Associate Dept. of Elecrical and Computer Enginerring Texas A&M University 333D Wisenbaker College Station, TX 77843-3259 Phone: 979-587-8311