IBM solidDB数据库文档中.ppt

Training Material: solidDB Architecture Page * The second failover scenario is when the Secondary fails You can tell when this happens because A watchdog or SysAdmin Polling the secondary fails. The Primary state is either PRIMARY ALONE or PRIMARY UNCERTAIN. Training Material: solidDB Architecture Page * Step 1 - The first thing to do If the Primary is in the PRIMARY UNCERTAIN state, is to set it to PRIMARY ALONE using ADMIN COMMAND HSB set primary alone Training Material: solidDB Architecture Page * Step 2 - Fix the old Secondary and restart Training Material: solidDB Architecture Page * Step 3 - Reconnect the secondary to the primary If the secondary has been down for a long period of time the primary may have switched to STANDALONE In this case you have to Set primary back to PRIMARY ALONE and perform a netcopy before you can reconnect. Step 4 – Check the connection status Training Material: solidDB Architecture Page * The third failover scenario is when there is no communication between The Primary and Secondary In this case The Primary server State is either PRIMARY UNCERTAIN or PRIMARY ALONE. Training Material: solidDB Architecture Page * How to Recover Step 1 – If the Primary is in the PRIMARY UNCERTAIN state, the first thing to do is to set it to PRIMARY ALONE so that the application can continue to execute commands Training Material: solidDB Architecture Page * Step 2 - Fix the network connection between the Primary and Secondary servers. Training Material: solidDB Architecture Page * Step 3 – If the Primary is PRIMARY ALONE, you can just Reconnect If the primary has switched to STANDALONE You will have to Set primary back to PRIMARY ALONE and perform a netcopy before you can reconnect. Step 4 – Check the connection status 关于netcopy的代替 原因：FTP/scp比Netcopy快 步骤： 关闭Secondary，如果它是启动的。 删除Secondary 的所有数据文件和日志文件 在Primary连续执行下面两条命令，先做一次checkpoint，然后马上关闭checkpoint. admin command mcp -s; admin command cp off; 用操作系统的scp或FTP命令，直接将Primary的数据文