Fault tolerant computer system

• US 20060150003A1
• Filed: 12/16/2005
• Published: 07/06/2006
• Est. Priority Date: 12/16/2004
• Status: Active Grant

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Abe, Shinji
• Granted Patent: US 7,610,509 B2
• Time in Patent Office: Days
• US Class Current: 714/11
• CPC Class Codes: G06F 11/1629 Error detection by comparing the output of redundant processing systems
• CPC Class Codes: G06F 11/2005 using redundant communication controllers
• CPC Class Codes: G06F 11/2007 using redundant communication media

Abstract

A fault tolerant (FT) computer system includes a first system; and a second system configured to operate in synchronization with the first system. Each of the first and second systems includes a CPU; and a routing controller connected with the CPU. The first system includes a first I/O device as an activist I/O device, and the second system includes a second I/O device as a standby I/O device. The routing controller controls a routing between the CPU and the first I/O device and the second I/O device. When a fault has occurred in the first I/O device, the routing controller in said first system routes a request data a request data received from the CPU and destined to the first I/O device, to the second I/O device.

Claims

1. A fault tolerant (FT) computer system comprising:

a first system as an active system comprising first I/O devices; and

a second system as a standby system comprising second I/O devices, which are same as said first I/O devices, connected with said first system through a link section, and configured to operate in synchronization with said first system,

wherein each of said first and second systems comprises;

a CPU; and

a routing controller connected with said CPU,

said routing controller in said first system controls a routing between said CPU in said first system and said first I/O devices and said second I/O devices, and

when a fault has occurred in one of said first I/O devices, said routing controller in said first system routes a request data received from said CPU in said first system and destined to said first I/O device, to one of said second I/O devices corresponding to said I/O device.

2. The fault tolerant computer system according to claim 1, wherein said routing controller in said first system comprises:

an address converter connected with said CPU;

a response data converter connected with said CPU; and

a router connected with said address converter and said response data converter,

when the fault has occurred in said first I/O device, said address converter generates a converted request data from said request data by replacing a device data indicating said first I/O device by a device data indicating said second I/O device, if receiving said request data from said CPU, and forwards the converted request data to said router, and

said router routes said converted request data to said second I/O device based on said device data contained in said converted request data.

3. The fault tolerant computer system according to claim 2, wherein said device data is a PCI bus number, a device number and a function number in a PCI hierarchy structure.

4. The fault tolerant computer system according to claim 3, wherein said routing controller in said first system further comprises:

a register referred to from said address converter and configured to store said PCI bus number, said device number and said function number.

5. The fault tolerant computer system according to claim 2, wherein said device data is an address in a system memory map space.

6. The fault tolerant computer system according to claim 5, wherein said routing controller in said first system further comprises:

a register referred to from said address converter and configured to store said address.

7. The fault tolerant computer system according to claim 2, wherein said response data converter generates a converted response data from a response data by rewriting a part of said response data, when said second I/O device operates in a normal state and said response data converter receives said response data from said second I/O device through said router, and outputs said converted response data to said CPU in said first system.

8. The fault tolerant computer system according to claim 7, wherein said response data converter generates said converted response data by replacing a class code of said second I/O device which is contained in said response data by another class code.

9. The fault tolerant computer system according to claim 8, wherein said routing controller in said first system further comprises:

a register referred to from said response data converter and configured to store said another class code.

10. The fault tolerant computer system according to claim 2, wherein said first system comprises a first PCI bridge as one of said first I/O devices connected with said routing controller of said first system and said first I/O device, and

said second system comprises a second PCI bridge as one of said second I/O devices connected with said routing controller of said second system and said second I/O device, and

wherein a VGA Enable bit of said first PCI bridge and a VGA Enable bit of said second PCI bridge are both set to Enable when said first I/O device and said second I/O device operate in a normal state.

11. The fault tolerant computer system according to claim 10, wherein when an access to said VGA Enable bit of said second PCI bridge is generated, said response data converter receives said VGA Enable bit from said second PCI bridge through said router, and converts the received VGA Enable bit into Disable.

12. The fault tolerant computer system according to claim 10, wherein said first I/O devices and said second I/O devices are open-system devices.

13. The fault tolerant computer system according to claim 1, wherein said first I/O device and said second I/O device are VGA devices.

14. A fault tolerant (FT) computer system comprising:

a first system comprising a first I/O device as an activist I/O device; and

a second system comprising a second I/O device as a standby I/O device, which is same as said first I/O device, connected with said first system through a link section, and configured to operate in synchronization with said first system,

wherein each of said first and second systems comprises;

a CPU; and

a routing controller connected with said CPU to control a routing between said CPU and said first I/O device and said second I/o device, and

said routing controller comprises a register group and a controller, and

said controller comprises;

an address converter connected with said CPU to selectively convert a request data outputted from said CPU based on a first data stored in said register group;

a response data converter connected with said CPU to selectively convert a response data to be outputted to said CPU, based on a second data stored in said register group; and

a router connected with said address converter and said response data converter to route said request data and said response data,

15. The fault tolerant computer system according to claim 14, wherein when a fault has occurred in said first I/O device, said routing controller in said first system routes said request data received from said CPU in said first system and destined to said first I/O device, to said second I/O device.

16. The fault tolerant computer system according to claim 14, wherein said response data converter in said first system converts said response data based on said second data, if receiving said response data from said second I/O device through said router, when said first and second I/O devices operate in a normal state, and outputs said converted response data to said CPU in said first system.

17. The fault tolerant computer system according to claim 14, wherein said first system comprises a first PCI bridge as one of said first I/O devices connected with said routing controller of said first system and said first I/O device, and

said second system comprises a second PCI bridge as one of said second I/O devices connected with said routing controller of said second system and said second I/O device, and

wherein a VGA Enable bit of said first PCI bridge and a VGA Enable bit of said second PCI bridge are both set to Enable when said first I/O device and said second I/O device operate in a normal state.

18. The fault tolerant computer system according to claim 14, wherein said first I/O device and said second I/O device are VGA devices.