Method and infrastructure for detecting and/or servicing a failing/failed operating system instance

• US 7,979,749 B2
• Filed: 11/15/2006
• Issued: 07/12/2011
• Est. Priority Date: 11/15/2006
• Status: Expired due to Fees

• Current Assignee: GlobalFoundries, Inc.
• Original Assignee: International Business Machines Corporation
• Inventors: Burugula, Ramanjaneya Sarma, Jann, Joefon, Pattnaik, Pratap Chandra
• Primary Examiner(s): Maskulinski; Michael C
• Assistant Examiner(s): Arcos; Jeison C
• Application Number:
  US11/599,272
• Publication Number:
  US 20080115012A1
• Time in Patent Office: 1,700 Days
• Field of Search: 714/37, 714/47, 714/48, 709/1, 709/223, 710/8, 713/164
• US Class Current: 714/38.1
• CPC Class Codes: G06F 11/0712 in a virtual computing platform, e.g. logically partitioned systems
• CPC Class Codes: G06F 11/079 Root cause analysis, i.e. error or fault diagnosis in a hardware test environment G06F11/22; in a software test environment G06F11/36

Abstract

A method and infrastructure for a diagnosis and/or repair mechanism in a computer system, that includes an auxiliary service system running on the computer system.

Claims

1. A diagnosis mechanism for a computer system, comprising:

an auxiliary service system capable of gaining access to resources of a failed or failing operating system running on a computer system capable of supporting a plurality of concurrently-running operating system instances, each said operating system instance owning or sharing one or more processing elements, a certain amount of memory, and one or more input/output (I/O) devices, and automatically diagnosing the failed or failing operating without affecting a functioning of other instances of operating system concurrently operating on said computer system,

wherein said computer system comprises a Symmetric Multi-Processor (SMP), wherein said concurrently-running operating system instances and said auxiliary service system each occupies a respective Logical PARtition (LPAR), said auxiliary service system comprises an ambulance service LPAR, and said LPARs are controlled by a hypervisor,

wherein each said operating system instance, during an initial booting procedure, prepares a memory region for storing metadata of its kernel and non-kernel data structures and other information that can potentially assist said auxiliary service system during a failure of said operating system instance, said operating system instance providing a location of said memory region to said hypervisor before beginning a normal course of operation after said initial booting procedure,

machine-readable instructions permitting said auxiliary service system to identify locations of at least one of kernel structures and other data structures in a distressed operating system,

wherein said auxiliary system further comprises at least one data structure validator that validates values of said at least one of kernel structures and data structures.

2. The diagnosis mechanism of claim 1, wherein said auxiliary service system is further capable of at least one of:

repairing said operating system;

annunciating that said operating system requests servicing; and

providing a report of said diagnosing.

3. The diagnosis mechanism of claim 1, wherein said auxiliary service system comprises an operating system instance of said plurality of concurrently-running operating system instances, an operating system of said auxiliary service system being substantially dormant until invoked to perform said diagnosing on a distressed operating system instance in said computer system.

4. The diagnosis mechanism of claim 1, wherein said computer system comprises a computer system in a lights-out data service center.

5. The diagnosis mechanism of claim 1, wherein said hypervisor, upon becoming aware of a failed or failing operating system instance, informs the auxiliary service system of the location of the prepared memory region storing the metadata of the failed or failing operating system instance.

6. The diagnosis mechanism of claim 1, further comprising:

an access granting mechanism permitting said auxiliary service system access to resources of said operating system instances, including said one or more processing elements, memory, and I/O devices of each said operating system instances.

7. The diagnosis mechanism of claim 6, wherein said access granting mechanism further comprises:

machine-readable instructions to identify physical memory address ranges and I/O devices owned by a distressed operating system instance.

8. The diagnosis mechanism of claim 6, wherein said access granting mechanism further comprises:

machine-readable instructions to map a memory from a distressed operating system instance and to add the memory to the auxiliary service system without modifying contents of the memory.

9. The diagnosis mechanism of claim 6, wherein said access granting mechanism further comprises:

machine-readable instructions to map or mount filesystems in the I/O devices of a distressed operating system instance to the auxiliary service system.

10. The diagnosis mechanism of claim 6, further comprising a set of messages comprising one or more of:

a message from said access granting mechanism to said auxiliary service system notifying the auxiliary service system of an operating system instance in distress;

a message from said access granting mechanism to said auxiliary service system, informing said auxiliary service system of physical memory locations used by the distressed operating system instance; and

a message from said access granting mechanism to said auxiliary service system informing said auxiliary service system about the I/O devices accessed by the distressed operating system instance.

11. The diagnosis mechanism of claim 6, further comprising:

a set of messages exchanged between said operating system instances and said access granting mechanism.

12. The diagnosis mechanism of claim 11, said auxiliary service system comprising:

an auxiliary service system operating system; and

a memory management subsystem that maps a physical memory of a distressed operating system instance into a virtual address space of said auxiliary service system operating system.

13. The diagnosis mechanism of claim 12, wherein said auxiliary service system further comprises:

an I/O management system capable of mapping storage devices belonging to the distressed operating system instance accessible from the auxiliary service system.

14. An automated method of diagnosis in a computer system capable of supporting a plurality of operating system instances concurrently operating, each said operating system instance owning or sharing one or more processing elements, a certain amount of memory, and one or more input/output (I/O) devices, said method comprising:

running an auxiliary service system on said computer system, said auxiliary service system capable of automatically diagnosing an operating system instance running on said computer system without affecting an operation of remaining operating system instances,

wherein said computer system comprises a Symmetric Multi-Processor (SMP), wherein said plurality of operating system instances and said auxiliary service system each occupies a respective Logical PARtition (LPAR), said auxiliary service system comprises an ambulance service LPAR, and said LPARs are controlled by a hypervisor,

wherein each said operating system instance, during an initial booting procedure, prepares a memory region for storing metadata of its kernel and non-kernel data structures and other information that can potentially assist said auxiliary service system during a failure of said operating system instance, said operating system instance providing a location of said memory region to said hypervisor before beginning a normal course of operation after said initial booting procedure; and

executing machine-readable instructions, upon detecting that an operating system instance has failed or potentially will fail, permitting said auxiliary service system to identify locations of at least one of kernel structures and other data structures in said failed or potentially failing operating system instance,

wherein said auxiliary service system further comprises at least one data structure validator, said method further comprising using said data structures to validate values of said at least one of kernel structures and data structures.

15. The automated method of claim 14, said auxiliary service system further capable of at least one of:

repairing an operating system instance running on said computer system;

annunciating that an operating system instance running on said computer system is requesting servicing; and

providing a report of said diagnosing.

16. The automated method of claim 14, further comprising:

permitting said auxiliary service system access to resources of said operating system instance, including said one or more processing elements, memory, and I/O devices of said operating system instance.

17. A tangible signal-bearing storage medium tangibly embodying a program of machine-readable instructions executable by a digital processing apparatus to perform an automated method of diagnosis in a computer system, said method comprising:

running an auxiliary service system on said computer system capable of concurrently operating a plurality of operating system instances, said auxiliary service system capable of diagnosing a failed or failing operating system instance running on said computer system without affecting an operation of remaining instances of said plurality of operating system instances,

wherein said computer system comprises a Symmetric Multi-Processor (SMP), wherein said plurality of operating system instances and said auxiliary service system each occupies a respective Logical PARtition (LPAR), said auxiliary service system comprises an ambulance service LPAR, and said LPARs are controlled by a hypervisor,

wherein each said operating system instance, during an initial booting procedure, prepares a memory region for storing metadata of its kernel and non-kernel data structures and other information that can potentially assist said auxiliary service system during a failure of said operating system instance, said operating system instance providing a location of said memory region to said hypervisor before beginning a normal course of operation after said initial booting procedure; and

permitting said auxiliary service system to identify locations of at least one of kernel structures and other data structures in a distressed operating system,

wherein said auxiliary system further comprises at least one data structure validator that validates values of said at least one of kernel structures and data structures.

18. The storage medium of claim 17, said auxiliary service system further capable of at least one of:

repairing an operating system instance running on said computer system;

annunciating that an operating system instance running on said computer system is requesting servicing; and

providing a report of diagnosing said operating system instance.

19. The storage medium of claim 17, wherein said computer system is capable of supporting a plurality of operating system instances, each said operating system instance owning or sharing one or more processing elements, a certain amount of memory, and one or more input/output (I/O) devices, said method further comprising:

permitting said auxiliary service system access to resources of said operating system instance, including said one or more processing elements, memory, and I/O devices of said operating system instance.

20. The storage medium of claim 17, said method further comprising:

mapping a physical memory of a distressed operating system instance into a virtual address space of an operating system in said auxiliary service system.

21. A computer system, comprising:

means for setting up and maintaining a plurality of concurrently-running operating system instances, each said operating system instance owning or sharing one or more processing elements, a certain amount of memory, and one or more input/output (I/O) devices; and

an automatic diagnosis mechanism capable of at least one of diagnosing an operating instance running on said computer system, repairing an operating system instance running on said computer system, annunciating that an operating system instance running on said computer system is requesting servicing, and providing a report of diagnosing said operating system instance, without affecting an operation of any other instances of said concurrently-running operating system instances,

wherein said computer system comprises a Symmetric Multi-Processor (SMP), wherein each of said plurality of concurrently-running system instances and said automatic diagnosis mechanism occupies a respective Logical PARtition (LPAR), said auxiliary service system comprises an ambulance service LPAR, and said LPARs are controlled by a hypervisor, wherein each said operating system instance, during an initial booting procedure, prepares a memory region for storing metadata of its kernel and non-kernel data structures and other information that can potentially assist said auxiliary service system during a failure of said operating system instance, said operating system instance providing a location of said memory region to said hypervisor before beginning a normal course of operation after said initial booting procedure,

said automatic diagnosis mechanism comprising machine-readable instructions permitting said auxiliary service system to identify locations of at least one of kernel structures and other data structures in a distressed operating system, wherein said auxiliary system further comprises at least one data structure validator that validates values of said at least one of kernel structures and data structures.

22. The computer system of claim 21, as comprising a component in a lights-out data center.

23. A method of at least one of operating a data center and using a data center, said method comprising:

at least one of transmitting data to a computer system in said data center and receiving data from said computer system,

wherein said computer system in said data center has provisions to set up and maintain a plurality of concurrently-running operating system instances, each said operating system instance owning or sharing one or more processing elements, a certain amount of memory, and one or more input/output (I/O) devices, said computer system further comprising an automatic diagnosis mechanism capable of at least one of diagnosing an operating instance running on said computer system, repairing an operating system instance running on said computer system, annunciating that an operating system instance running on said computer system is requesting servicing, and providing a report of diagnosing said operating system instance, without affecting an operation of remaining instances of said plurality of concurrently-running operating system instances, said automatic diagnosis mechanism comprising machine-readable instructions permitting said auxiliary service system to identify locations of at least one of kernel structures and other data structures in a distressed operating system,

wherein said auxiliary system further comprises at least one data structure validator that validates values of said at least one of kernel structures and data structures;

wherein said computer system comprises a Symmetric Multi-Processor (SMP),

wherein said plurality of concurrently-running operating system instances and said auxiliary service system each occupies a respective Logical PARtition (LPAR), said auxiliary service system comprises an ambulance service LPAR, and said LPARs are controlled by a hypervisor,

wherein each said operating system instance, during an initial booting procedure, prepares a memory region for storing metadata of its kernel and non-kernel data structures and other information that can potentially assist said auxiliary service system during a failure of said operating system instance, said operating system instance providing a location of said memory region to said hypervisor before beginning a normal course of operation after said initial booting procedure.

24. The method of claim 23, wherein said data received from said computer system comprises said report of said diagnosing said operating system instance.

25. The method of claim 23, wherein said data transmitted to said computer system or received from said computer system comprises data that is not related to a maintenance of said computer system.