| CPC G06F 16/164 (2019.01) [G06F 16/134 (2019.01); G06F 16/182 (2019.01)] | 18 Claims |

|
1. A metadata processing method, comprising:
obtaining, by a management server connected with at least one first node, a second node and at least one third node, metadata to be processed, wherein the metadata to be processed comprises a directory structure and file attributes;
according to a load condition of the at least one first node for storing the file attributes and based on a rule that file attributes corresponding to files of a same directory are stored in a same first node, storing, in a first node by the management server, the file attributes corresponding to files of directories in the directory structure, or storing, in first nodes by the management server, the file attributes corresponding to files of directories in the directory structure in a distributed manner; wherein each of the first node or first nodes in which the file attributes are stored has a remaining storage space greater than a storage space required for storing the file attributes corresponding to files of a corresponding directory, and has a load less than a first preset threshold value;
obtaining, by the management server, distributed storage information of the file attributes corresponding to the files of the directories in the directory structure on the first node or first nodes, and storing, by the management server, the directory structure and the obtained distributed storage information in the second node; and storing, by the management server, the files of the directories in the directory structure in the at least one third node; and
obtaining, by the management server, directories other than a root directory from the directory structure, and performing following operations for each obtained directory: obtaining, by the management server, identification information of a parent directory of the obtained directory; allocating, by the management server, identification information to the obtained directory; obtaining, by the management server, according to the distributed storage information, identification information and location information of a first node storing file attributes corresponding to files of the obtained directory; generating, by the management server, directory entry information according to the obtained identification information of the parent directory, a name of the obtained directory, the identification information of the obtained directory, and the obtained identification information and location information of the first node; and storing, by the management server, the generated directory entry information in the second node.
|