DESIGN TO SURVIVE FAILURES AND FAULT HANDLING Failure is nothing but fault in performance or defect in performance. Whenever we design a application we need to make sure the application works properly without any disturbances with 100% performance, especially when we run the application in cloud where a lot of applications run we need to make sure the application run productively overcoming all the faults or hindrance. There are two basic categories of failures they are: 1) Transient, self – healing failures 2) Enduring failures that require intervention. Transient failures which can be self – healing we can get the result if we try again. Such as network errors etc., For enduring failures, we can implement monitoring & logging functionality which would notify when an error occurs and also offers a method named root-cause analysis. Machine Failures: A failed server is automatically replaced by new one, a perfectly designed will recover automatically and quickly. Service Failures: When an App deployed in the cloud fails how we would be taking an alternate step in order to make the user comfortable and make the app operational. Region Failures: We would be running our apps in different locations and when there is a region failure we would be able to get another data center working so we would be deploying the app in azure so that we could use the app from different regions when one of the region is down.