Backup switching control system and method

A backup switching control method in a computer system, comprising:

• connecting a plurality of combinations of operating machines and backup machines to a network, wherein said backup machines take over said data processing from said operating machines when a failure of a corresponding one of said operating machines occurs for each said combination, respectively;
• providing at least one system resource for each said combination connected to corresponding ones of said operating and backup machines either directly or through said network, wherein said operating machines exclusively occupy said system resources for processing data within said system including at least one of inputting, outputting and storing data with said system resources;
• for each said combination, said operating machine and said backup machine performing a switching control function including sending a system disconnection command and a dump acquisition command from said backup machine to said operating machine, and further for sending a disconnection complete notice from said operating machine to said backup machine;
• transmitting said dump acquisition command and said disconnection command to said operating machine from said backup machine when a failure occurs in said operating machine, said operating machine executing a dump process in response to receiving said dump acquisition command that dumps data externally of said operating machine and executing a disconnection process in response to receiving said disconnection command independently of the completion of said dump process;
• wherein said operating machine outputs to said backup machine the disconnection complete notice once said operating machine is disconnected from said system and whereby said backup machine, after receiving said disconnection complete notice, takes over the data processing from said operating machine including occupying said system resources for performing ongoing processing within said system.

A backup switching system for a combination of first and second data processors, wherein a first processor is an operating processor and a second processor is a backup processor, said processors being connected to at least one system resource, wherein said operating processor exclusively occupies said system resource during a normal operation of said operating processor, comprising:

• each of said operating and backup processors respectively having a main processor, an auxiliary processor, a shared memory connected for data exchange with each of said main and auxiliary, and a communication port;
• means connecting said operating and backup processors through said communication ports;
• said backup processor having means for detecting when a failure occurs in said operating processor;
• said auxiliary processor having means operative in response to said detecting means detecting a failure in said operating processor for sending to said operating processor a command for releasing said system resource and for dumping data;
• said operating processor having means responding to said sending means for dumping data being processed to a memory external of said operating processor and means for releasing said system resource independently of completion of dumping of data by said data dumping means, and means for transmitting a disconnection complete notice to said backup processor indicating that said system resource has been released by said operating processor;
• and said backup processor having means for receiving said notice and for exclusively occupying said system resource in response to receiving said notice.

A backup switching system, comprising:

• a plurality of combinations of first and second data processing machines connected to a network, wherein said first machines are operating machines and said second machines are backup machines, said machines of each said combination being connected to at least one system resource either directly or thorough said network, wherein said operating machines exclusively occupy said system resources during normal operation of said operating machines;
• each of said operating and backup machines within each said combination respectively having a main processor, an auxiliary processor, a shared memory connected for data exchange with each of said main and auxiliary processors, and a communication port connected to said network;
• each said backup machine having means for detecting when a failure occurs in a corresponding one of said operating machines in each said combination;
• in each said backup machine, said auxiliary processor having means operative in response to said detecting means detecting a failure in said corresponding operating machine for sending to said corresponding operating machine a command for releasing said system resource and for dumping data;
• each said operating machine having means responding to said sending means for dumping data being processed to a memory external of said operating machine and for releasing said system resource independently of completion of dumping of the data by said data dumping means, and means for transmitting a disconnection complete notice to said corresponding backup machine indicating that said system resource has been released by said operating machine;
• and each said backup machine having means for receiving said notice and for exclusively occupying said system resource after receiving said notice.