Our paper presents a data recovery system which rectifies a disaster in cloud. Cloud computing provides a wide range of online storage where security issues play a major role. In existing system, Cloud computing not providing the security on data and only provide data recovery process. In existing system, data recovery process implement on finding the failure cloud server and then recover the failure server data but recover data was not checking this is actual data or not. In proposed system using new algorithm(SHA1) to improve the more security to user file. The user file is splitted into small data pieces and stored with multiple servers connected to a cloud. A holidaymaker or blind folks will face drawback in finding safe place if any natural disaster happens. For this reason, we tend to have projected a location based mostly early disaster warning and evacuation system for each traditional and blind people exploitation Google Map. as a result of of the use of sensible phones in developing countries will increase, the system is mplemented on robot mobile phone. So, our system consists of a third-party server named Disaster Management Server (DMS), robot device on that our application put in and user. Updates of the disaster (tsunami, cyclone or flood) area unit place on DMS by the native weather workplace. To get automatic notification of future disaster device user registers on Disaster Management Server (DMS) else user will gets manual notification. The user keep in bit with DMS to get most renew knowledge obtained by GSM. The projected application notifies the user situated in potential disaster zone with visual and audio disaster warning and evacuation guideline mix with nearest location of shelter or safe zone on the map of the application. DMS conjointly tracked Evacuation progress and national identification of user. The experimental result proves the effectiveness of our system