An efficient scheme for fault tolerance in cloud environment

Abstract EN

A DCell network topology is designed to support infrastructure data center, and operations routing, and decision-making are a challenge for network designers routing the data packet from the source node to the target node, and selecting the shortest path.

In this paper propose a new algorithm to address failures (Node, Link, Rack) The algorithm has three parts, part one: using the routing resolution to address link failure, selecting the average number of random paths for the server and selecting the shortest alternative path (time), delay, and throughput.

To find out the adjacent nodes we use point to point.

Part two: Handling Node Failure by finding the shortest two paths (disjoint).

After reverting to the previous node, routing, checking and jumping to the alternate path protocol.

Part Three: Addressing Rack Failure an alternative path is chosen for the same node level from the failed Rack1 to another Rack2 shortest path.

When comparing the proposed algorithm with the algorithms used, it has excelled (transmission efficiency, high throughput, less time, less delay, bandwidth 100Mbps, fault tolerance).