A computer accessible medium may be encoded with instructions which, when executed: replicate a checkpoint segment from a first local storage of a first node to at least one other node; and load a copy of the checkpoint segment from the other node to a second local storage of a second node. The checkpoint segment is stored into the first local storage by an application, and comprises a state of the application. The copy of the checkpoint segment is loaded into a second local storage responsive to a request from the second node to load the copy. The second node is to execute the application. In some embodiments, the copy of the checkpoint segment may also be loaded into a global storage.