Abstract

Information Lifecycle Management (ILM) is a strategic concept for storage of information and documents. ILM is based on the idea that in an enterprise information have different values. Information with different values are stored on different storage hierarchies. ILM offers significant potential cost savings by tiering storage and 90% of decision makers consider implementing ILM (Linden 2006). Nonetheless, there are too few experience reports and experimenting and researching in real systems are too expensive. This paper addresses this issue and contributes to supporting and assisting IT managers in their decision-making process. ILM automation needs migration rules. There are well-known static, heuristic migration rules and we present a new dynamic migration rule for ILM. These migration rules are implemented in an ILM simulator. We compare the performance of the new dynamic rule with the heuristics. The simulative approach has two advantages. On the one hand it offers predictions about the dynamic behaviour of an ILM migration rule and, on the other hand, it dispenses with real storage hardware. Simulation leads to decisions under certainty. When making a decision under certainty, the major problem is to determine which is the trade-off among different objectives. Cost-benefit analysis can be used to this purpose. A decision matrix is laid where rows represent choices and columns represent states of nature. The simulated results support the choice of migration rules and help to avoid mismanagement and poor investments in advance. The results raise the awareness of choosing the best alternative.

Share

COinS