Nec Corporation

Paolo Laffranchini

Maurizio Dusi

Francesco Gringoli

A method for operating a distributed computing system includes transmitting, by a first node in a first level to an intermediate node in a second level, a first message and storing, by the first node, the first message. The method further includes receiving, by the intermediate node, the first message, generating a second message, and transmitting, to a third node, the second message. The method additionally includes receiving, by the intermediate node from the third node, a first acknowledgement message indicating that the second message was successfully received by the third node, and transmitting, by the intermediate node to the first node after receiving the first acknowledgement message, a second acknowledgement message. The method includes deleting, by the first node in response to receiving the second acknowledgment message, the stored first message.

Method for operating a distributed computing system and a distributed computing system

A method for operating a distributed computing system including a plurality of nodes connected in a general network topological sequential manner includes storing a first message on a first node; removing the first message from the first node only if an N-level acknowledgement message is received by the first node indicating that a second message based on the data of the first message was successfully received by a second node on a second level, wherein the difference between the values for the first and second level is a predefined value N; and saving a node state of at least the second node or a third node on a level higher than the first level before sending out an acknowledgement message to a node on a level below if an outgoing message towards the second node is generated based on the node state.

Growing community of inventors

Desenzano del Garda, Italy

Paolo Laffranchini