Fault tolerant network server having multiple slideably-mounted processing elements sharing a redundant array of independent disks through SCSI isolators

• US 6,618,821 B1
• Filed: 01/12/2000
• Issued: 09/09/2003
• Est. Priority Date: 01/12/2000
• Status: Expired due to Fees

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Duncan, James R., Hance, Edgar, Raymond, Patrick A., Merits, Gerald J., Pelonero, Mike L., Salinas, Everett R., McAfee, David A.
• Primary Examiner(s): Iqbal, Nadeem
• Assistant Examiner(s): Bonura, Timothy M.
• Application Number:
  US09/481,954
• Time in Patent Office: 1,336 Days
• Field of Search: 214/6, 214/10, 214/11, 214/14
• US Class Current: 714/14
• CPC Class Codes: G06F 11/0796 Safety measures, i.e. ensuring safe condition in the event of error, e.g. for controlling element
• CPC Class Codes: G06F 11/2089 Redundant storage control functionality

Abstract

A fault tolerant network server is described. This server has a pair of processing units, each processing unit has at least one CPU, system memory, an interface for a RAID system, at least one disk drive, a network interface, a cluster network interface, a power supply, and a case. The cases of both processing units are slideably mounted in a rack-mountable server case. The server also has a RAID system mounted in the server case and powered by two, redundant, power supplies and coupled to the processing units through isolators for blocking transactions when a processing unit has insufficient power or is being serviced.

Claims

1. A fault-tolerant network server comprising:

a first processing unit coupled to receive power through a first power cord, the processing unit comprising at least one CPU,

system memory coupled to the at least one CPU, an interface for a RAID system and addressable by the at least one CPU,

at least one disk drive addressable by the at least one CPU, a network interface addressable by the at least one CPU, and

a case containing said at least one CPU, system memory, interface for a RAID system, network interface, and disk drive;

a second processing unit coupled to receive power through a second power cord, the processing unit comprising a case,

at least one CPU disposed within the case,

system memory coupled to the at least one CPU and disposed within the case,

an interface for a RAID system addressable by the at least one CPU and disposed within the case,

a network interface addressable by the at least one CPU and disposed within the case,

at least one disk drive addressable by the at least one CPU and disposed within the case, and

a power supply disposed within the case, coupled to energize the CPU, system memory, interface for a RAID system, network interface, and disk drive;

a RAID system coupled to receive power both through the first power cord and the second power cord, said RAID system coupled to the first processing unit and to the second processing unit including a shelf for mounting hot-pluggable disk drives, a first RAID controller coupled to said shelf and to the first processing unit and a second RAID controller coupled to said shelf and to the second processing unit, wherein the first RAID controller is coupled to said shelf for mounting hot-pluggable disk drives through a first isolation circuit, and wherein the second RAID controller is coupled to said shelf for mounting hot-pluggable disk drives through a second isolation circuit, and wherein the first and the second isolation circuits each include a manually activated disable switch that, when in a disable setting, causes the isolation circuit to block all transactions between a RAID controller and a processing unit, a connection present detection circuit that, when a processing unit connection cable is detected as being disconnected from the isolation circuit, causes the isolation circuit to block all transactions between the RAID controller and the processing unit connection cable, and an activity monitor that creates a visual indication of transactions between the processing unit and the RAID controller; and

a network server case, the case of each of the first and second processing units being slideably mounted within the network server case, and the RAID system being disposed within the network server case.

2. The failure-resistant network server of claim 1, wherein the first processing unit and the second processing unit run software capable of performing a fail-over should either of the first processing unit and the second processing unit fail.

3. A fault-tolerant network comprising:

a first CPU;

a first isolation circuit coupled to the first CPU;

a first RAID controller coupled to the first isolation circuit;

a second RAID controller coupled to the first RAID controller;

a disk drive coupled to the first and second RAID controllers;

a second isolation circuit coupled to the second RAID controller; and

a second CPU coupled to the second isolation circuit,

wherein the first and second isolation circuits each include a manually activated disable switch.

4. The fault-tolerant network of claim 3 further comprising system memory coupled to the first and second CPUs.

5. The fault-tolerant network of claim 3 in which the first and second RAID controllers each comprise two SCSI ports.

6. The fault-tolerant network of claim 3 in which the first and second isolation circuits each comprise a SCSI isolation circuit.

7. The fault-tolerant network of claim 3 in which the manually activated disable switch comprises a momentary-contact switch.

8. The fault-tolerant network of claim 3 in which the manually activated disable switch comprises a toggle flip-flop.

9. The fault-tolerant network of claim 3 in which the manually activated disable switch further comprises a current setting indicator.

10. The fault-tolerant network of claim 9 in which the current setting indicator comprises a light emitting diode.

11. The fault-tolerant network of claim 3 further comprising a heartbeat interconnection between the first and second RAID controllers.

12. A fault-tolerant network comprising:

at least one processing unit;

at least one isolation circuit;

at least one RAID controller, all in communication with each other; and

a RAID shelf including at least one disk drive in communication with the at least one RAID controller,

wherein the at least one isolation circuit includes a manually activated disable switch.

13. The fault-tolerant network of claim 12 further comprising system memory coupled to the at least one processing unit.

14. The fault-tolerant network of claim 12 in which the at least one RAID controller comprises two SCSI ports.

15. The fault-tolerant network of claim 12 in which the at least one isolation circuit comprises a SCSI isolation circuit.

16. The fault-tolerant network of claim 12 in which the manually activated disable switch comprises a momentary-contact switch.

17. The fault-tolerant network of claim 12 in which the manually activated disable switch comprises a toggle flip-flop.

18. The fault-tolerant network of claim 12 in which the manually activated disable switch further comprises a current setting indicator.

19. The fault-tolerant network of claim 18 in which the current setting indicator comprises a light emitting diode.

20. The fault-tolerant network of claim 12 further comprising a heartbeat interconnection between a first and a second RAID controller.