| 摘要 |
<p>PROBLEM TO BE SOLVED: To reduce a service down time at master server startup or failover time by reducing the consumption of a network bandwidth or master server calculation resource when synchronizing information in the master server and information in a slave server in a distributed system.SOLUTION: At startup of a master server or occurrence of failover, the master server acquires metadata from each of slave servers indicated in slave server list management information, and executes a background synchronization process to be stored in metadata management information. When a request from a client server is received during the background synchronization process, the master server refers to the slaver server list management information and, if metadata to be serviced for the request is not synchronized yet with metadata in the slave servers, acquires the metadata from the slave servers and then returns a response to the request from the client server.</p> |
