Performance of Size-Changing Algorithms in Stackable File Systems

Abstract

Stackable file systems can provide extensible file system functionality with minimal performance overhead and development cost. However, previous approaches are limited in the functionality they provide. In particular, they do not support size-changing algorithms, which are important and useful for many applications, such as compression and security. We propose fast index files, a technique for efficient support of size-changing algorithms in stackable file systems. Fast index files provide a page mapping between file system layers in a way that can be used with any size-changing algorithm. Index files are designed to be recoverable if lost and add less than 0.1% disk space overhead. We have implemented fast indexing using portable stackable templates, and we have used this system to build several example file systems with size-changing algorithms. We demonstrate that fast index files have very low overhead for typical workloads, only 2.3% over other stacked file systems. Our system can deliver much better performance on size-changing algorithms than user-level applications, as much as five times faster.