Send-side matching of data communications messages

摘要

Send-side matching of data communications messages in a distributed computing system comprising a plurality of compute nodes, including: issuing by a receiving node to source nodes a receive message that specifies receipt of a single message to be sent from any source node, the receive message including message matching information, a specification of a hardware-level mutual exclusion device, and an identification of a receive buffer; matching by two or more of the source nodes the receive message with pending send messages in the two or more source nodes; operating by one of the source nodes having a matching send message the mutual exclusion device, excluding messages from other source nodes with matching send messages and identifying to the receiving node the source node operating the mutual exclusion device; and sending to the receiving node from the source node operating the mutual exclusion device a matched pending message.

主权项

1. A method of send-side matching of data communications messages in a distributed computing system comprising a plurality of compute nodes organized for collective operations, the method comprising:
issuing, by a receiving node to a plurality of source nodes, a receive message that specifies receipt of a single message to be sent from any source node, the receive message also including message matching information, a specification of a hardware-level mutual exclusion device, and an identification of a receive buffer; matching by two or more of the source nodes the receive message with pending send messages in the two or more source nodes by comparing the tags of each pending send messages with the message matching information in the receive message; operating by one of the source nodes having a matching send message the hardware-level mutual exclusion device, excluding messages from other source nodes with matching send messages and identifying to the receiving node the source node operating the hardware-level mutual exclusion device; in response to identifying to the receiving node the source node operating the hardware-level mutual exclusion device, issuing, by the receiving node, a cancellation message to the other source nodes that cancels the matching send messages; and sending to the receiving node from the source node operating the hardware-level mutual exclusion device a matched pending send message.