Fast resource recovery after thread crash

• US 8,627,142 B2
• Filed: 01/11/2013
• Issued: 01/07/2014
• Est. Priority Date: 01/14/2010
• Status: Expired due to Fees

• Current Assignee: Juniper Networks Incorporated
• Original Assignee: Juniper Networks Incorporated
• Inventors: Lynn, Michael
• Primary Examiner(s): Patel, Kamini
• Application Number:
  US13/739,285
• Publication Number:
  US 20130132773A1
• Time in Patent Office: 361 Days
• Field of Search: 714/15
• US Class Current: 714/15
• CPC Class Codes: G06F 11/0715 in a system implementing multitasking multitasking per se G06F9/46
• CPC Class Codes: G06F 11/0793 Remedial or corrective actions recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40
• CPC Class Codes: G06F 11/1402 Saving, restoring, recovering or retrying
• CPC Class Codes: G06F 9/5016 the resource being the memory

Abstract

A resource recovery system may maintain a counter in memory that indicates a number of times one or more threads of execution, which use shared resources, have crashed. The system may associate a first value of the counter with a resource allocated to a thread of the one or more threads, and may set an indicator associated with the thread to indicate whether the thread has crashed. The system may determine whether to re-allocate the resource to the thread based on the first value of the counter associated with the resource and based on the indicator associated with the thread.

Claims

1. A method comprising:

receiving, by a device, a resource request for a thread, of a plurality of threads, associated with the device;

storing, by the device and in a queue for allocating resources to the thread, metadata associated with a resource corresponding to the resource request and information identifying the thread;

retrieving, by the device, the metadata and the information from the queue to allocate the resource to the thread;

associating, by the device, a first value with the metadata,

the first value corresponding to a value of a counter at a time that the resource is allocated to the thread, and

the value of the counter being incremented each time any of the plurality of threads crashes; and

re-allocating, by the device and when the thread crashes, the resource to the thread based on associating the first value with the metadata.

2. The method of claim 1, where the resource includes a data unit received by the device from another device via a network, and

where storing the metadata associated with the resource and the information identifying the thread includes:

storing metadata associated with a header of the data unit and the information identifying the thread in the queue for allocating resources to the thread.

3. The method of claim 1, further comprising:

incrementing, when the thread crashes, the value of the counter from the first value to a second value; and

where re-allocating the resource to the thread includes;

determining, when the thread crashes, that;

the resource is allocated to the thread,

the resource is associated with the first value, and

the first value is less than the second value, and

re-allocating the resource to the thread based on the first value being less than the second value.

4. The method of claim 1, further comprising:

determining that one of the plurality of threads has crashed;

incrementing the value of the counter to a second value based on determining that one of the plurality of threads has crashed;

determining whether the second value is greater than the first value;

determining whether the thread comprises the one of the plurality of threads that has crashed when the value of the incremented counter is greater than the first value,

the resource being re-allocated to the thread when the thread comprises the one of the plurality of threads that has crashed.

5. The method of claim 4, further comprising:

determining not to re-allocate the resource to the thread when the second value is not greater than the first value.

6. The method of claim 1, further comprising:

determining that one of the plurality of threads has crashed;

associating the one of the plurality of threads with an indicator,

the indicator indicating that the one of the plurality of threads has crashed; and

determining that the resource is allocated to the thread based on the information identifying the thread being associated with the metadata; and

where re-allocating the resource to the thread includes;

determining that the thread is associated with the indicator based on determining that the resource is allocated to the thread,

determining that the thread comprises the one of the plurality of threads that has crashed based on the thread being associated with the indicator, and

re-allocating the resource to the thread based on the thread comprising the one of the plurality of threads that has crashed and the first value.

7. The method of claim 1, further comprising:

incrementing the value of the counter to a second value based on determining that the one of the plurality of threads has crashed; and

where re-allocating the resource to the thread further includes;

determining that the first value is less than the second value, and

re-allocating the resource to the thread further based on the first value being less than the second value.

8. A device comprising:

a processor to;

receive a resource request for a thread, of a plurality of threads, associated with the device,

store, in a memory associated with the device;

metadata associated with a resource corresponding to the resource request, and

information identifying the thread,

retrieve the metadata and the information identifying the thread from the memory to allocate the resource to the thread,

associate, based on retrieving the metadata and the information identifying the thread, the metadata with;

a first value, and

the information identifying the thread,

the first value corresponding to a value of a counter at a time that the metadata is stored in the memory, and

the value of the counter being incremented each time any of the plurality of threads crashes, and

re-allocate, when the thread crashes, the resource to the thread based on associating metadata with the first value and the information identifying the thread.

9. The device of claim 8, where, when re-allocating the resource to the thread, the processor is to:

re-allocate, when the thread crashes, the resource to the thread based on;

associating the metadata with the first value and the information identifying the thread, and

a value of the counter at a time subsequent to the thread crashing.

10. The device of claim 8, where the resource includes a data unit received by the device from another device via a network, and

where, when storing the metadata associated with the resource and the information identifying the thread, the processor is to:

store metadata associated with a header of the data unit and the information identifying the thread in the memory.

11. The device of claim 8, where the processor is further to:

increment, when the thread crashes, the value of the counter from the first value to a second value, and

where, when re-allocating the resource to the thread, the processor is to;

determine, when the thread crashes, that the thread has been allocated the resource and that the resource is associated with the first value,

determine that the first value is less than the second value, and

re-allocate the resource to the thread based on the first value being less than the second value.

12. The device of claim 8, where the processor is further to:

determine that one of the plurality of threads has crashed,

increment the value of the counter to a second value based on determining that one of the plurality of threads has crashed,

determine whether the second value is greater than the first value,

determine whether the thread comprises the one of the plurality of threads that has crashed when the second value is greater than the first value,

the resource being re-allocated to the thread when the thread comprises the one of the plurality of threads that has crashed.

13. The device of claim 12, where the processor is further to:

determine not to re-allocate the resource to the thread when the second value is not greater than the first value.

14. The device of claim 8, where the processor is further to:

determine that one of the plurality of threads has crashed,

associate the one of the plurality of threads with an indicator,

the indicator indicating that the one of the plurality of threads has crashed, and

determine that the resource is allocated to the thread based on the information identifying the thread being associated with the metadata; and

where, when re-allocating the resource to the thread, the processor is to;

determine that the thread is associated with the indicator based on determining that the resource is allocated to the thread,

determine that the thread comprises the one of the plurality of threads that has crashed based on the thread being associated with the indicator, and

re-allocate the resource to the thread based on the thread comprising the one of the plurality of threads that has crashed and the first value.

15. A non-transitory computer-readable medium storing instructions, the instructions comprising:

one or more instructions which, when executed by a processor of a device, cause the processor to;

allocate a resource to a thread, of a plurality of threads associated with the device,

associate, based on allocating the resource, metadata associated with the resource with a first value and information identifying the thread,

the first value corresponding to a value of a counter at a time that the resource is allocated to the thread, and

the value of the counter being incremented each time any of the plurality of threads crashes, and

re-allocate, when the thread crashes, the resource to the thread based on associating the first value with the metadata.

16. The non-transitory computer-readable medium of claim 15, where the one or more instructions to re-allocate the resource to the thread include:

one or more instructions to re-allocate, when the thread crashes, the resource to the thread based on;

associating the metadata with the first value and the information identifying the thread, and

a value of the counter at a time subsequent to the thread crashing.

17. The non-transitory computer-readable medium of claim 15, where the resource includes a data unit received by the device from another device via a network, and

where the metadata includes metadata associated with a header of the data unit.

18. The non-transitory computer-readable medium of claim 15, where the instructions further comprise:

one or more instructions to increment, when the thread crashes, the counter from the first value to a second value, and

where the one or more instructions to re-allocate the resource to the thread include;

one or more instructions to determine, when the thread crashes, that the thread has been allocated the resource and that the resource is associated with the first value,

one or more instructions to determine that the first value is less than the second value, and

one or more instructions to re-allocate the resource to the thread based on the first value being less than the second value.

19. The non-transitory computer-readable medium of claim 15, where the instructions further comprise:

one or more instructions to determine that one of the plurality of threads has crashed,

one or more instructions to increment the value of the counter to a second value based on determining that one of the plurality of threads has crashed,

one or more instructions to determine whether the second value is greater than the first value,

one or more instructions to determine whether the thread comprises the one of the plurality of threads that has crashed when the second value is greater than the first value,

the resource being re-allocated to the thread when the thread comprises the one of the plurality of threads that has crashed.

20. The non-transitory computer-readable medium of claim 19, where the instructions further comprise:

one or more instructions to determine not to re-allocate the resource to the thread when the second value is not greater than the first value.