Object-based checkpoints in distributed systems

In distributed applications, multiple objects are cooperated to achieve some objective. The objects may suffer from certain kinds of faults. If some object o is faulty, it is rolled back to the checkpoint and objects which have received messages from o are also required to be rolled back. In this paper, we define influential messages whose receivers are required to be rolled back if the senders are rolled back in the object-based computation model. By using these influential messages, object-based checkpoints are defined to denote semantically consistent global states of the system which are nevertheless inconsistent with the traditional message-based definition. We show how the number of checkpoints can be reduced by taking only the object-based checkpoints