Checkpoint debugging using mirrored virtual machines

• US 8,639,984 B2
• Filed: 08/09/2011
• Issued: 01/28/2014
• Est. Priority Date: 08/09/2011
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: McNeeney, Adam J.
• Primary Examiner(s): PATEL, KAMINI B
• Application Number:
  US13/205,739
• Publication Number:
  US 20130042150A1
• Time in Patent Office: 903 Days
• Field of Search: 714/38.1
• US Class Current: 714/38.1
• CPC Class Codes: G06F 11/1438 Restarting or rejuvenating
• CPC Class Codes: G06F 11/1484 involving virtual machines
• CPC Class Codes: G06F 11/366 using diagnostics G06F11/0703 takes precedence

Abstract

A system of debugging computer code includes a processor: obtaining state information corresponding to a first machine at a checkpoint initiated during execution of the computer code on the first machine; and configuring the second machine to a same operating state as the first machine at the checkpoint to create a mirrored version of the first machine. The system also includes receiving a notification that execution of the program on a first machine has failed, and in response to receiving the notification: triggering a processor of the second machine to initiate execution of a copy of the code from a specific code execution point at which the checkpoint was; activating a debugger module to run concurrently with the execution of the program on the second machine and collect and store the debug data as corresponding to execution failure of the computer code at the first machine.

Claims

1. A computer readable storage medium comprising computer readable code for debugging a computer program executing in a mirrored virtualized data processing system environment, which code, when executed by a processor causes the processor to:

obtain state information corresponding to a previous operating state of the first machine at a checkpoint performed during the execution of the computer code on the first machine;

configure, with the state information obtained, a second machine having a same physical configuration to a same operating state as the previous operating state of the first machine at the checkpoint, wherein the second machine is a mirrored version of the first machine relative to execution of the computer code;

receive a notification indicating that execution of the computer code on the first machine has failed; and

in response to receiving the notification;

trigger a processor of the second machine to initiate execution of a copy of the computer code on the second machine from a specific code execution point at which the checkpoint was initiated on the first machine;

activate a debugger module to run concurrently with the execution of the copy of the computer code on the second machine and collect debug data corresponding to execution of the computer code on the second machine from the checkpoint up to the failure of the computer code execution on the second machine; and

store the debug data as debug data corresponding to execution failure of the computer code on the first machine.

2. The computer readable storage medium of claim 1, wherein:

the first machine and the second machine are a first virtual machine and a second virtual machine, each respectively configured and maintained by a hypervisor and each having an identical provisioning of an amount and type of processor, memory and other resources to enable the first virtual machine and the second virtual machine to respectively perform an identical execution of the computer code and of the copy of the computer code;

the computer code executes in the first virtual machine with no concurrent execution of a debugger to enable more efficient execution of the computer code and to reduce an amount of processor and other resources utilized while running the computer code on the first machine; and

the copy of the computer code is executed concurrently with the debugger module on the second machine from only the specific code execution point at which the checkpoint was initiated and only in response to detection of the execution failure of the computer code at the first machine.

3. The computer readable storage medium of claim 1, wherein obtaining the state information corresponding to a previously initiated checkpoint comprises:

accessing a pre-established storage location at which the state information is stored during the checkpoint on the first machine, wherein the state information is stored at the pre-established storage location by a checkpoint module performing the checkpoint on the first machine, wherein the checkpoint module completes the processes of;

suspending execution of the computer code in the first machine;

capturing current state information of the first machine, including the specific point of execution in the computer code;

storing the state information at the pre-established storage location that is accessible to the second machine; and

resuming execution of the computer code on the first machine in response to successful completion of the storing of the state information.

4. The computer readable storage medium of claim 1, wherein:

the state information comprises information and data corresponding to at least a processor state and a memory state existing within the first machine at the checkpoint and is linked to the specific code execution point corresponding to the checkpoint; and

obtaining the state information comprises retrieving the processor state, memory state and associated specific code execution point from the accessible storage.

6. The computer readable storage medium of claim 1, wherein the computer readable code causes the processor to:

select the checkpoint and corresponding state information and specific code execution point;

in response to receiving the notification, identify which second machine is physically pre-configured as a mirrored second machine of the first machine;

access the pre-established storage location; and

select a correct state information from the pre-established storage to utilize in configuring the mirrored second machine to operate similarly to the first machine from the identified checkpoint in the computer code to yield a similar failure as occurred on the first machine.

7. The computer readable storage medium of claim 2, wherein the first virtual machine and the second virtual machine are one of:

collocated on a same physical device with a single hypervisor; and

located on separate host devices connected via a communication interconnect from among a direct interconnect and a network.

5. A computer readable storage medium comprising computer readable code for debugging a computer program executing in a mirrored virtualized data processing system environment, which code, when executed by a processor causes the processor to:

obtain state information corresponding to a previous operating state of the first machine at a checkpoint performed during the execution of the computer code on the first machine;

configure, with the state information obtained, a second machine having a same physical configuration to a same operating state as the previous operating state of the first machine at the checkpoint, wherein the second machine is a mirrored version of the first machine relative to execution of the computer code;

receive a notification indicating that execution of the computer code on the first machine has failed; and

in response to receiving the notification;

parse the notification for an indication of a failure type from among one or more pre-established failure types having associated failure characteristics and debug requirements;

select, based on the failure type indicated within the notification, the state information corresponding to a selected checkpoint from among a plurality of previous checkpoints performed at different times on the first machine, with each of the plurality of previous checkpoints having a different set of state information and a different specific code execution point, wherein the selected checkpoint enables complete debugging of the failure type;

trigger a processor of the second machine to initiate execution of a copy of the computer code on the second machine from a specific code execution point at which the checkpoint was initiated on the first machine;

activate a debugger module to run concurrently with the execution of the copy of the computer code on the second machine and collect debug data corresponding to execution of the computer code on the second machine from the checkpoint up to the failure of the computer code execution on the second machine;

run the debugger module from the selected checkpoint with one or more available debugger processes that enables debugging of the failure type; and

store the debug data as debug data corresponding to execution failure of the computer code on the first machine.

8. A system for debugging a failure detected in a computer program, the system comprising:

a computer processor; and

a checkpoint debugger module which, when executed by the computer processor, causes the computer processor to;

receive, after completion of a previous checkpoint, a notification indicating that execution of the computer code on the first machine has failed;

in response to receiving the notification after the previous checkpoint;

trigger a processor of a second machine that is a mirrored version of the first machine relative to the previous checkpoint to initiate execution of a copy of the computer code on the second machine from a specific code execution point at which the checkpoint was initiated on the first machine;

activate a debugger module to run concurrently with execution of a copy of the computer code on the second machine and collect debug data corresponding to execution of the computer code on the second machine from the previous checkpoint up to a failure of the computer code execution on the second machine; and

store the debug data as debug data corresponding to execution failure of the computer code on the first machine.

9. The system of claim 8, wherein:

the first machine and the second machine are a first virtual machine and a second virtual machine each respectively configured and maintained by a hypervisor and each having an identical provisioning of an amount and type of processor, memory and other resources to enable the first virtual machine and the second virtual machine to respectively perform an identical execution of the computer code and of the copy of the computer code;

the computer code executes in the first virtual machine without a concurrent execution of the debugger module to enable more efficient execution of the computer code and to reduce an amount of processor and other resources utilized while running the computer code on the first machine; and

the copy of the computer code is executed concurrently with the debugger module on the second machine from only the specific code execution point at which the checkpoint was initiated and only in response to detection of the execution failure of the computer code at the first machine.

10. The system of claim 8, further comprising a hypervisor associated with the second machine and which:

obtains state information corresponding to a previous operating state of the first machine at a checkpoint performed during the execution of the computer code on the first machine prior to a failure of the computer code execution; and

configures, with the state information obtained, a second machine having a same physical configuration to a same operating state as the previous operating state of the first machine at the checkpoint, wherein the second machine is a mirrored version of the first machine relative to execution of the computer code.

11. The system of claim 8, wherein:

the hypervisor obtains the state information corresponding to a previously initiated checkpoint by accessing a pre-established storage location at which the state information is stored during the checkpoint on the first machine, wherein the state information is stored at the pre-established storage location by a checkpoint module performing the checkpoint on the first machine, wherein the checkpoint module completes the processes of;

suspending execution of the computer code in the first machine;

capturing current state information of the first machine, including the specific point of execution in the computer code;

storing the state information at the pre-established storage location that is accessible to the second machine;

resuming execution of the computer code on the first machine in response to successful completion of the storing of the state information;

the state information comprises information and data corresponding to at least a processor state and a memory state existing within the first machine at the checkpoint and is linked to the specific code execution point corresponding to the checkpoint; and

obtaining the state information comprises retrieving the processor state, memory state and associated specific code execution point from the accessible storage.

12. The system of claim 11, wherein the checkpoint debugger module further causes the processor to:

parse the notification for an indication of a failure type from among one or more pre-established failure types having associated failure characteristics and debug requirements;

select, based on the failure type indicated within the notification, the state information corresponding to a selected checkpoint from among a plurality of previous checkpoints performed at different times on the first machine, with each of the plurality of previous checkpoints having a different set of state information and a different specific code execution point, wherein the selected checkpoint enables complete debugging of the failure type; and

run the debugger module from the selected checkpoint with one or more available debugger processes that enables debugging of the failure type.

13. The system of claim 8, wherein the checkpoint debugger module further causes the processor to:

select checkpoint and corresponding state information and specific code execution point;

in response to receiving the notification, identifying which second machine is physically pre-configured as a mirrored second machine of the first machine;

accessing the pre-established storage location; and

selecting a correct state information from the pre-established storage to utilize in configuring the mirrored second machine to operate similarly to the first machine from the identified checkpoint in the computer code to yield a similar failure as occurred on the first machine.

14. The system of claim 13, wherein the first virtual machine and the second virtual machine are one of:

collocated on a same physical device with a single hypervisor; and

located on separate host devices connected via a communication interconnect from among a direct interconnect and a network.