Distributed File System Supporting Parallel Access To Multiple Storage Devices

Our audit of the current Kerberos-based protocol demonstrates that it has various impediments: (i) a metadata server encouraging key trade between the customers and the storage devices has substantial workload that limits the scalability of the protocol; (ii) the protocol does not give forward mystery; (iii) the metadata server produces itself all the session keys that are utilized between the customers and storage devices, and this inalienably prompts key escrow. In this paper, we propose an assortment of confirmed key trade protocols that are intended to address the above issues. We demonstrate that our protocols are fit for lessening up to roughly 54% of the workload of the metadata server and simultaneously supporting forward mystery and escrow-freeness. This requires just a little division of expanded calculation overhead at the customer