System and method for distributed processing of file volume

摘要

A file processing system includes a system targeting unit to target a distributed file system, a chunk generator to generate a chunk including file blocks and to generate copies of the chunk to be stored in the distributed file system, a data storage medium to store the chunk, and a chunk operation requesting unit to request storage of the generated chunk copies in the distributed file system. A distributed file system includes a first data server and a second data server. Each data server includes a chunk storage unit to store a chunk copy received from a file processing system, and a chunk operation performing unit to generate an identical copy of the stored chunk copy that is undamaged to replace a damaged chunk copy.

主权项

1. A file processing system, comprising:
a processor configured to,
select a distributed file system, the distributed file system includes a plurality of data servers in which chunk copies are distributed and stored, the distributed file system forming a large-scale virtual storage device such that the plurality of data servers appear as a single storage device to a client system connected thereto;generate a chunk by dividing a file volume to be stored in the distributed file system; andrequest storage of the chunk in the distributed file system, the chunk being stored based on an ownership of the chunk, and wherein the distributed file system is configured to,
store, on a first data server of the plurality of data servers, the chunk in response to the request to store the chunk from file processing system,determine if the chunk is damaged,perform a replica migration operation to generate an undamaged chunk copy of the chunk if the chunk is damaged, the replica mitigation operation generating the undamaged chunk copy by,
selecting, one of the plurality of data servers not having a copy of the chunk as a new data server, the new data server being a different data server from the first data server, andcopying an undamaged chunk copy from a second data server of the plurality of data servers storing an undamaged chunk copy to the new data server, the undamaged chunk copy that is copied from the second data server to the new data server being an undamaged identical version of the chunk stored in the first data server, the first data server and the second data server storing chunk copies having a different owner.