Mechanism for saving crash dump files of a virtual machine on a designated disk

• US 8,365,020 B2
• Filed: 03/18/2010
• Issued: 01/29/2013
• Est. Priority Date: 03/18/2010
• Status: Active Grant

• Current Assignee: Red Hat Israel Ltd. (International Business Machines Corporation)
• Original Assignee: Red Hat Israel Ltd. (International Business Machines Corporation)
• Inventors: Laor, Dor
• Primary Examiner(s): MCCARTHY, CHRISTOPHER S
• Application Number:
  US12/726,825
• Publication Number:
  US 20110231710A1
• Time in Patent Office: 1,048 Days
• Field of Search: None
• US Class Current: 714/38.11
• CPC Class Codes: G06F 11/0712 in a virtual computing platform, e.g. logically partitioned systems
• CPC Class Codes: G06F 11/0778 Dumping, i.e. gathering error/state information after a fault for later diagnosis
• CPC Class Codes: G06F 11/366 using diagnostics G06F11/0703 takes precedence
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects

Abstract

A mechanism for saving crash dump files of a virtual machine (VM) on a designated disk is disclosed. A method of embodiments of the invention includes configuring an operating system (OS) of a VM managed by a hypervisor of a host machine to store any crash dump files created by the VM to a designated crash dump virtual disk associated with the VM but accessible outside of operations of the VM, determining that the VM experienced a crash event, stopping operations of the VM, and obtaining a crash dump file created by the VM that details the crash event from the designated crash dump virtual disk.

Claims

1. A computer-implemented method, comprising:

associating, by a hypervisor that virtualizes a plurality of virtual machines (VMs) on a host machine, each VM of the plurality of VMs with a crash dump disk that is solely dedicated to the VM, wherein each crash dump disk is located separate from its associated VM and separate from operating system (OS) files of its associated VM;

configuring, by the hypervisor, an OS of each VM with a crash file path to the crash dump disk associated with the VM;

configuring, by the hypervisor, each VM of the plurality of VMs to;

generate crash dump files for the VM upon a crash event of the VM; and

store, via the crash file path, the generated crash dump files to the crash dump disk associated with the VM;

identifying, by the hypervisor, a crashed VM of the plurality of VMs that experienced a crash event;

stopping, by the hypervisor, operations of the crashed VM; and

obtaining, by the hypervisor from the crash dump disk associated with the crashed VM, a crash dump file created by the crashed VM that details the crash event.

2. The method of claim 1, further comprising providing the crash dump file of the crashed VM to a controller that manages the host machine.

3. The method of claim 2, wherein the providing the crash dump file is in response to a request for the crash dump file from the controller.

4. The method of claim 1, further comprising generating a non-maskable interrupt (NMI) upon identifying the crashed VM, and sending the NMI to the crashed VM, wherein the NMI causes the crashed VM to create the crash dump file and store the crash dump file in the crash dump disk associated with the crashed VM.

5. The method of claim 4, wherein the NMI is used to execute an interrupt handler routine at the VM that causes the VM to create the crash dump file and send the crash dump file to the designated crash dump disk.

6. The method of claim 1, wherein identifying the crashed VM comprises monitoring each VM of the plurality of VMs for one or more predetermined events that indicate a crash event of one of the plurality of VMs.

7. The method of claim 1, wherein identifying the crashed VM comprises receiving a notification from one of the plurality of VMs that the VM is encountering a situation requiring creation of a crash dump file.

8. A system, comprising:

a processing device;

a memory communicably coupled to the processing device; and

a hypervisor to virtualize the processing device and the memory for use by a plurality of virtual machines (VMs) hosted by the hypervisor, the hypervisor configured to;

associate each VM of the plurality of VMs with a crash dump disk that is solely dedicated to the VM, wherein each crash dump disk is located separate from its associated VM and separate from operating system (OS) files of its associated VM;

configure an OS of each VM of the plurality of VMs with a crash file path to the crash dump disk;

configure each VM of the plurality of VMs to;

generate crash dump files for the VM upon a crash event of the VM; and

store, via the crash file path, the generated crash dump files to the crash dump disk associated with the VM;

identify a crashed VM of the plurality of VMs that experienced a crash event;

stop operations of the crashed VM; and

obtain, from the crash dump disk associated with the crashed VM, a crash dump file created by the crashed VM that details the crash event.

9. The system of claim 8, wherein the hypervisor further configured to provide the crash dump file of the crashed VM to a controller that manages the host machine.

10. The system of claim 8, wherein the hypervisor further configured to generate a non-maskable interrupt (NMI) upon identifying the crashed VM, and send the NMI to the crashed VM, wherein the NMI causes the crashed VM to create the crash dump file and store the crash dump file in the crash dump disk associated with the crashed VM.

11. The system of claim 10, wherein the NMI is used to execute an interrupt handler routine at the crashed VM that causes the crashed VM to create the crash dump file and send the crash dump file to the crash dump disk associated with the crashed VM.

12. The system of claim 8, wherein the hypervisor configured to identify the crashed VM further comprises the hypervisor configured to monitor each VM of the plurality of VMs for one or more predetermined events that indicated a crash of one of the plurality of VMs.

13. The system of claim 8, wherein the hypervisor configured to identify the crashed VM further comprises the hypervisor configured to receive a notification from the crashed VM that the crashed VM is encountering a situation requiring creation of a crash dump file.

14. A non-transitory machine-readable storage medium including data that, when accessed by a machine, cause the machine to perform operations comprising:

associating, by a hypervisor that virtualizes a plurality of virtual machines (VMs) on a host machine, each VM of the plurality of VMs with a crash dump disk that is solely dedicated to the VM, wherein each crash dump disk is located separate from its associated VM and separate from operating system (OS) files of its associated VM;

configuring, by the hypervisor, an OS of each VM with a crash file path to the crash dump disk associated with the VM;

configuring, by the hypervisor, each VM of the plurality of VMs to;

generate crash dump files for the VM upon a crash event of the VM; and

store, via the crash file path, the generated crash dump files to the crash dump disk associated with the VM;

identifying, by the hypervisor, a crashed VM of the plurality of VMs that experienced a crash event;

stopping, by the hypervisor, operations of the crashed VM; and

obtaining, by the hypervisor from the crash dump disk associated with the crashed VM, a crash dump file created by the crashed VM that details the crash event.

15. The non-transitory machine-readable storage medium of claim 14, wherein the data, when accessed by the machine, causes the machine to perform further operations comprising providing the crash dump file stored to a controller that manages the host machine.

16. The non-transitory machine-readable storage medium of claim 14, wherein the data, when accessed by the machine, causes the machine to perform further operations comprising:

generating a non-maskable interrupt (NMI) upon identifying the crashed VM; and

sending the NMI to the crashed VM, wherein the NMI causes the VM to create the crash dump file and store the crash dump file in the crash dump disk associated with the crashed VM.

17. The non-transitory machine-readable storage medium of claim 14, wherein identifying the crashed VM comprises monitoring each VM of the plurality of VMs for one or more predetermined events that indicate a crash event of one of the plurality of VMs.

18. The non-transitory machine-readable storage medium of claim 14, wherein identifying the crashed VM comprises receiving a notification from one of the plurality of VMs that the VM is encountering a situation requiring creation of a crash dump file.