On inserting bulk data for linear hash files

Abstract

In this investigation, we propose a new approach for bulk insert under linear hash organization, which is one of the known problems of dynamic hash techniques. Generally Dynamic Hash allows us to adjust the size of hash space dynamically according to the amount of data so that we obtain the nice time/space efficiency of the hash space. One of the typical techniques is linear hash where we can keep the hash space size linearly in terms of the amount of data. However, the technique doesn't always provide us with suitable properties, especially we face to severe deficiencies at bulk insert/delete operations, because the successive operations cause heavy manipulation (one extension at each operation) so that huge amount of I/O access happen. Here in this work, we propose a new approach for bulk insert (delete) to relieve the thrashing situation and to reduce total I/O access. We discuss some experimental results and show how well the approach works. © 2011 Springer-Verlag.