The Survey on Content Addressable Storage

1. Introduction Content addressable storage (CAS) is an emerging mechanism that can reduce the costs associated with this volume of data by eliminating such redundancy. Essentially, CAS uses cryptographic hashing techniques to identify data by its content rather than by name. Consequently, a CAS-based system will identify sets of identical objects and only store or transmit a single copy even if higher-level logic maintains multiple copies with different names. CAS is also a data management approach. CAS uses cryptographic hashing to reduce storage requirements by exploiting commonality across multiple data objects. For example, to apply CAS to a system, we would represent each memory and disk image as a sequence of fixed-sized chunk files, where the filename of each chunk is computed using a collision-resistant cryptographic hash function. Since chunks with identical names are assumed to have identical contents, a single chunk on disk can be included in the representations of multiple memory and disk images. The simplest example of this phenomenon is that many memory and disk images contain long strings of zeros, most of which can be represented by a single disk chunk consisting of all zeros.