摘要

A fault-tolerant improvement solution based on the running-information mechanism was presented to improve the fault tolerance of its running-information (including exception-information, status-information, resource-information and configuration-information), operational stability and user experience. The exception-information for statistics was statisticsed and analyzied to get error events of the status-information and resource-information. Then targeted treatments for the corresponding errors was given. The practical result shows that the proposed fault-tolerant improvement solution can effectively improve OpenStack fault-tolerance, and reduce the various kinds of abnormalities caused by exception-information.

全文