Diagnostic fault management controller for distributed computing

• US 10,437,754 B1
• Filed: 09/23/2016
• Issued: 10/08/2019
• Est. Priority Date: 09/23/2016
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Ebisuzaki, Gavin Akira, Patel, Vijay, BeSerra, Christopher James
• Primary Examiner(s): Peyton, Tammara R
• Application Number:
  US15/274,524
• Time in Patent Office: 1,110 Days
• CPC Class Codes: G06F 13/1668 Details of memory controller
• CPC Class Codes: G06F 13/24 using interrupt G06F13/32 takes precedence
• CPC Class Codes: G06T 1/60 Memory management
• CPC Class Codes: G06T 2200/28 involving image processing hardware

Abstract

A management controller may request units of diagnostic information from a BIOS of the management controller'"'"'s host computing device. The management controller may trigger an interrupt, in response to which the BIOS, by the execution of a processor of the host, may cause the diagnostic information to be copied to a video memory of the management controller. Upon the completion of the interrupt handler, a graphics controller of the management controller may cause the diagnostic information to be transferred to a non-volatile memory, and transferred out-of-band to a client device.

Claims

1. A system, comprising:

a first one or more processors;

a graphics controller;

a video memory addressable by the graphics controller and by the first one or more processors, the memory further connected to a second one or more processors of a computing device via a first communications bus;

a non-volatile memory, the non-volatile memory addressable by the first one or more processors;

one or more memories having stored thereon executable instructions that, upon execution by the first one or more processors, cause the system at least to;

send data indicative of a request to receive a first diagnostic information from a basic input/output system of the computing device;

receive data indicating that at least a subset of the first diagnostic information has been loaded into the video memory by the execution of basic input/output system instructions by the second one or more processors;

cause the graphics controller to transfer the first diagnostic information from the video memory to the non-volatile memory; and

send the data indicative of the first diagnostic information stored in the volatile memory over a second communications bus.

2. The system of claim 1, wherein the one or more memories have stored thereon further executable instructions which, upon execution by the first one or more processors, cause the system at least to:

trigger a first interrupt comprising the request to receive the first diagnostic information; and

trigger a second interrupt comprising a second request for a second diagnostic information.

3. The system of claim 1, wherein the one or more memories have stored thereon further executable instructions which, upon execution by the first one or more processors, cause the system at least to:

determine that a second subset of the first diagnostic information has not been loaded into the video memory; and

send data indicative of a request to receive the second subset of the first diagnostic information.

4. The system of claim 1, wherein the one or more memories have stored thereon further executable instructions which, upon execution by the first one or more processors, cause the system at least to:

cause the graphics controller to compress the at least a subset of the first diagnostic information prior to transferring the compressed at least a subset of the first diagnostic information to the non-volatile memory.

5. The system of claim 1, wherein the one or more memories have stored thereon further executable instructions which, upon execution by the first one or more processors, cause the system at least to:

determine, in response to data indicating that the computing device has recovered from a system failure, that the at least a subset of the first diagnostic information has been stored in the non-volatile memory;

load the at least a subset of the first diagnostic information into the video memory; and

send information indicating that the at least a subset of the first diagnostic information is available to be read from the video memory by the second one or more processors of the computing device.

6. A method of obtaining diagnostic information from a computing device, the method comprising:

triggering an interrupt of one or more processors of the computing device and providing, by the interrupt, information indicative of a request for a first diagnostic information;

receiving data indicating that at least a subset of the first diagnostic information has been stored in a first memory of the management controller, wherein the at least a subset of the first diagnostic information was received by a first communications bus;

causing the management controller to transfer the first diagnostic information from the first memory to a second memory of the management controller; and

sending data indicative of the first diagnostic information stored in the second memory over a second communications bus.

7. The method of claim 6, further comprising:

triggering a second interrupt in response to determining that the first diagnostic information comprises a second subset of the first diagnostic information in addition to the at least a subset of the first diagnostic information.

8. The method of claim 6, further comprising:

receiving, at the management controller, information indicative of a plurality of diagnostic information available from the computing device, the plurality of diagnostic information comprising the first diagnostic information; and

triggering the interrupt based on the information indicative of the plurality of diagnostic information.

9. The method of claim 6, further comprising:

causing a graphics controller of the management controller to compress the at least a subset of the first diagnostic information.

10. The method of claim 6, wherein the interrupt is processed by a management interrupt handler of an input/output system of the computing device.

11. The method of claim 6, further comprising:

disabling a first frame buffer of a plurality of frame buffers used by a graphics controller of the management controller, each of the plurality of frame buffers corresponding to a non-overlapping region of the first memory; and

providing an address of the first frame buffer in the first memory in the information indicative of the request for the first diagnostic information.

12. The method of claim 6, further comprising:

receiving, at the management controller, information indicating that the computing device has recovered from a system failure;

loading the at least a subset of the first diagnostic information into the first memory from the second memory;

triggering a second interrupt indicating that the at least a subset of the first diagnostic information has been loaded into the first memory; and

receiving, at the management controller, information indicating that the at least a subset of the first diagnostic information may be overwritten in the first memory.

13. The method of claim 6, wherein a basic input/output system causes the at least a subset of the first diagnostic information to be written to the first memory and the basic input/output system sends, to the management controller, the data indicating that at the least a subset of the first diagnostic information has been stored in a video memory.

14. The method of claim 6, wherein the second communication bus has a lower maximum bandwidth than the first communication bus.

15. The method of claim 6, further comprising:

causing a graphics controller of the management controller to encrypt the at least a subset of the first diagnostic information prior to transferring the encrypted at least a subset of the first diagnostic information to the second memory.

16. A system comprising:

means for receiving a first set of diagnostic information over a first communications bus of a management controller and storing the first set of diagnostic information in a video memory of the management controller;

means for receiving data indicating that the first set of diagnostic information has been stored in the video memory;

means for transferring the first set of diagnostic information from the video memory to a second memory;

means for determining to request a second set of diagnostic information subsequent to the transferring of the first set of diagnostic information from the video memory; and

means for sending the first set of diagnostic information stored in the second memory to a client device via a second communications bus.

17. The system of claim 16, wherein the video memory is addressable by a first processor of the management controller and a second processor of a computing device external to the management controller.

18. The system of claim 16, wherein a host computing device causes the first set of diagnostic information to be written to video memory and sends the data indicating that the first set of diagnostic information has been stored in the video memory.

19. The system of claim 16, further comprising:

means for performing at least one of compressing or encrypting the first set of diagnostic information prior to storing the first set of diagnostic information on the second memory.

20. The system of claim 16, further comprising:

means for transferring the first set of diagnostic information from the second memory to the video memory upon recovery from system failure.