Classification of access to distributed file systems
Classification of access to distributed file systems
Background
There are many distributed file systems, which can be roughly divided into two categories: The first category includes distributed file systems such as NFS and CIFS, which provide universal functions and are generalised. The second category includes special distributed file systems that are optimised for special applications and provide extended functionalities, such as increased availability or high performance in cluster environments. Over time, however, requirements change so that a single file system is often inadequate or has to be reconfigured. In order to optimise a distributed file system, logs and statistics on the access behaviour of the applications used are required.
Job description
In this thesis, a component is to be developed with which the access behaviour of applications to (distributed) file systems can be logged. The collected data will then be analysed to determine whether recurring patterns can be identified that allow access to be classified. The knowledge gained will be used to optimise distributed file systems.