"Storage optimization in distributed environments is a major concern when talking about reliability in this kind of schemes. Although replication is the most used option, erasure coding is a more optimized one.
However, erasure coding uses a lot of bandwidth to replace one node. In a dynamic scheme, where nodes enter and leave the system frequently, bandwidth use could be an important drawback.
Regenerating Codes introduced by Dimakis et al. minimize the code repair problem by applying Network Coding to the distributed storage scheme. However finding the coefficients for the linear combinations used to replace a node is not easy, specially for the systematic case, and must be calculated for each new node fail.…"

"Performance tuning of Database Management Systems(DBMS) is both complex and challenging as it involves identifying and altering several key performance tuning parameters. The quality of tuning and the extent of performance enhancement achieved greatly depends on the skill and experience of the Database Administrator (DBA). As neural networks have the ability to adapt to dynamically changing inputs and also their ability to learn makes them ideal candidates for employing them for tuning purpose. In this paper, a novel tuning algorithm based on neural network estimated tuning parameters is presented. The key performance indicators are proactively monitored….The tuner alters these tuning parameters using the estimated values using a rate change computing algorithm. The preliminary results show that the proposed method is effective in improving the query response time for a variety of workload types."