Resource management for peripheral component interconnect-express domains

Patent metadata

Field	Value
Publication number	US-9672167-B2
Application number	US-201414315099-A
Country	US
Kind code	B2
Filing date	Jun 25, 2014
Priority date	Jul 22, 2013
Publication date	Jun 6, 2017
Grant date	Jun 6, 2017

Abstract

Embodiments of the present invention provide a solution for managing inter-domain resource allocation in a Peripheral Component Interconnect-Express (PCIe) network. One processor among a plurality of link processors is elected as a management processor. The management processor obtains information about available resources of PCIe network. When a resource request from a request processor is received, the management processor allocates a resource of the available resources to the requesting processor. The management processor instructs one or more link processors to program one or more inter-domain NTBs through which the traffic between the allocated resource and the requesting processor is going to flow according to the memory address information of the allocated resource, to allow cross-domain resource access between the requesting processor and the allocated resource.

First claim

What is claimed is:

1. An apparatus for managing inter-domain resource allocation in a Peripheral Component Interconnect-Express (PCIe) network which includes a plurality of PCIe domains each managed by a link processor and connected to one or more other domains via a non-transparent bridge (NTB), comprising:
a memory;
a management processor coupled with the memory, the management processor configured to:
obtain information about available resources of the domains reported by each domain's link processor, wherein the information about available resources includes memory address of each available resource reported;
store the obtained information about the available resources in the memory;
receive a resource request from a requesting processor of one of the domains;
in response to the resource request, allocate a resource of the available resources to the requesting processor, wherein the allocated resource resides in a domain different from the domain with the requesting processor;
obtain memory address information of the allocated resource from the memory;
instruct one or more link processors to program one or more inter-domain NTBs through which the traffic between the allocated resource and the requesting processor is going to flow according to the memory address information of the allocated resource, to allow cross-domain resource access between the requesting processor and the allocated resource.

2. The apparatus according to claim 1 , wherein the information about the available resources reported includes information about type of each resource.

3. The apparatus according to claim 2 , wherein the information about type includes information about Make, Model, or Function of an available resource for identifying a driver to load for a processor.

4. The apparatus according to claim 3 , the management processor further configured to send information about the type of the allocated resource obtained from the memory to the requesting processor for identifying a driver to be load to the requesting processor.

5. The apparatus according to claim 1 , wherein the memory address of each available resource includes base address and the size of the memory address.

6. The apparatus according to claim 1 , wherein the information about the available resources reported includes amount information of each available resource.

7. The apparatus according to claim 1 , wherein the management processor allocates the resource by searching the information about the available resources stored the memory according to information about a requested type and a request amount carried in the resource request.

8. The apparatus according to claim 7 , wherein the management processor allocates the resource by further performing an allocation algorithm to determine the resource closest in terms of latency, hop-count, or number of NTBs to be crossed.

9. A method for managing inter-domain resource allocation in a Peripheral Component Interconnect-Express (PCIe) network which includes a plurality of PCIe domains each managed by a link processor and connected to one or more other domains via a non-transparent bridge (NTB), the method comprising:
obtaining information about available resources of the domains reported by each domain's link processor, wherein the information about available resources includes a memory address of each available resource reported;
storing the obtained information about the available resources in a memory;
receiving a resource request from a requesting processor of one of the domains;
in response to the resource request, allocating a resource of the available resources to the requesting processor, wherein the allocated resource resides in another domain different from the domain with the requesting processor;
obtaining memory address information of the allocated resource from the memory;
instructing one or more link processors to program one or more inter-domain NTBs through which the traffic between the allocated resource and the requesting processor is going to flow according to the memory address information of the allocated resource, to allow cross-domain resource access between the requesting processor and the allocated resource.

10. The method according to claim 9 , wherein the information about the available resources reported includes information about type of each resource.

11. The method according to claim 10 , wherein the information about type includes information about Make, Model, or Function of an available resource for identifying a driver to load for a processor.

12. The method according to claim 11 , further comprising sending information about the type of the allocated resource obtained from the memory to the requesting processor for identifying a driver to be load to the requesting processor.

13. The method according to claim 9 , wherein the memory address of each available resource includes base address and the size of the memory address.

14. The method according to claim 9 , wherein the information about the available resources reported includes amount information of each available resource.

15. The method according to claim 9 , wherein the resource is allocated by searching the information about the available resources stored the memory according to information about a requested type and a request amount carried in the resource request.

16. The method according to claim 15 , wherein the resource is allocated by further performing an allocation algorithm to determine the resource closest in terms of latency, hop-count, or number of NTBs to be crossed.

Assignees

• Futurewei Technologies Inc

Inventors

• Egi Norbert
• Lasater Robert
• Boyle Thomas
• Peters John
• Shi Guangyu

Classifications

• H04L47/785

  among multiple network domains, e.g. multilateral agreements · CPC title

• G06F13/4063

  Device-to-bus coupling · CPC title

• G06F13/161 Primary

  with latency improvement · CPC title

• G06F9/5005

  to service a request · CPC title

• G06F13/4036

  and deadlock prevention · CPC title