Asynchronous checkpointing and recovery algorithm. - Distributed systems- video 19

This algorithm is based on Asynchronous checkpointing The main concept of this algorithm is to find a consistent set of checkpoints from which a system can be recovered in case of failure. Each process maintains a set of messages it has sent to all other processes and also maintains the number of messages it has received from other processes. Triplet: The triplet should be recorded, Whenever an event is occurred ( send event or receive event). The format of triplet is as below Triplet( m, s, msgs_sent) Take the process p. it is at the state s, sends the message m or receives the message m, processes it moves state to s ‘ , and sends messages out.