Stateless processing of replicated state for distributed storage systems

摘要

Embodiments are directed towards storing data in a storage system. A data controller may obtain a write request and write data from a client computer. A write message may be generated and provided to a data coordinator computer. The data coordinator may communicate the write message to a plurality of L-node computers. The data coordinator may obtain write confirmation messages from the L-node computers that indicate that the write data is stored. If enough write confirmation messages are obtained to indicate that a quorum is reached, the data coordinator may communicate a save confirmation message to the data controller. The data controller may generate a write acknowledgement message based on the save confirmation message provided by the data coordinator. The data controller may provide the write acknowledgement message to the client computer that made the original write request.

主权项

1. The method for storing data in a storage system over a network using a network computer that performs actions, comprising:
employing a P-node to perform actions, including:
communicating with a client computer that provides a write request which includes write data;generating one or more journal records that are converted, by the P-Node, from one or more write messages provided by one or more client computer applications, wherein the one or more journal records are provided to a plurality of L-nodes, and wherein content and client computer application context of the one or more journal records is opaque to the plurality of L-nodes;generating a new write message that includes the write data and the one or more journal records;communicating the new write message to one of the plurality of L-nodes that is elected by the plurality of L-nodes to be a data coordinator for the plurality of L-nodes; and employing the data coordinator to perform actions, including:
communicating the new write message to each of the L-nodes, wherein each L-node communicates a write confirmation message to the data coordinator that indicates the write data and the one or more journal records are stored by each L-node; andwhen a predetermined amount of write confirmation messages indicate that the write data and the one or more journal records are stored, communicating a save confirmation message to the P-node; and employing the P-node to generate a write acknowledgement message that is responsive to the write request based on the save confirmation message, wherein the write acknowledgement message is communicated to the client computer.