摘要 |
A method, apparatus and system for data disaster tolerance are provided in embodiments of this disclosure, the method comprising: receiving node failure information from a node; detecting along a predecessor direction and a successor direction of a failure node indicated in the node failure information according to a pre-stored node sequence to determine a first effective predecessor node and a first effective successor node, and all failure nodes between the first effective predecessor node and the first effective successor node; instructing those of all effective nodes that have local content registration index stored on the failure nodes and the first effective successor node to perform a primary index recovery process, respectively, so as to recover primary indexes of all of the failure nodes into the primary index of the first effective successor node. |