Configuring large-scale storage using a middleware with machine learning

ions utilized over configuration parameters Each deployment of the SAN configuration middleware will see a different underlying SAN infrastructure (i.e. the myriad different data centres of different cloud providers), each with its own custom set of configuration parameters. Transporting these data to and from the best practice repository will require the use of an abstract data format for SAN configuration snapshots. We translate the heterogeneous SAN configuration parameters into a common representation. One important side effect of the mapping functions defined to do this translation is the removal of parameter values that might disclose confidential information about the environment of particular clients. The best practice repository stores data in an object-oriented, hierarchical format that extends the CIM/SMI-S profiles. For further details, see the DMTF Common Information Model specification [22], and the SNIA Storage Management Initiative Specification [23]. These profiles are defined by standard bodies that include most of the significant industry participants. All snapshots are tagged with timestamps and collect information about the open or unresolved problem tickets that have been raised in that particular data centre. The problem tickets also require translation into a standard format. In this case, particular classifications are required to be applied by the user raising Copyright 2011 John Wiley & Sons, Ltd. Concurrency Computat.: Pract. Exper. (2011)