Algorithms for increasing performance in distributed file systems

Pavel Bžoch, Jiří Šafařík
2012 Acta Electrotechnica et Informatica  
Need of storing a huge amount of data has grown over the past years. Whether data are of multimedia types (e.g. images, audio, or video) or are produced by scientific computation, they should be stored for future reuse or for sharing among users. Users also need their data as quick as possible. Data files can be stored on a local file system or on a distributed file system. Local file system provides the data quickly but does not have enough capacity for storing a huge amount of the data. On
more » ... other hand, a distributed file systems (DFS) provide many advantages such as reliability, scalability, security, capacity, etc. In the paper, traditional DFS like AFS, NFS and SMB will be explored. These DFS were chosen because of their frequent usage. Next, new trends in these systems with a focus on increasing performance will be discussed. These include the organization of data and metadata storage, usage of caching, and design of replication algorithms. This paper provides overview of existing algorithms which are used in DFS. Described algorithms can be used as a basis for any future work.
doi:10.2478/v10198-012-0005-7 fatcat:52om3lw63fhx3hdnygjhomfhsq