Distributed acceleration devices management for streams processing

• US 9,009,723 B2
• Filed: 03/19/2013
• Issued: 04/14/2015
• Est. Priority Date: 08/29/2008
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Degenaro, Louis Ralph, Giles, James Ryan, Da Silva, Gabriela Jacques
• Primary Examiner(s): Eskandarnia, Arvin
• Assistant Examiner(s): DESAI, MARGISHI V
• Application Number:
  US13/847,478
• Publication Number:
  US 20150058614A1
• Time in Patent Office: 756 Days
• Field of Search: 709/228, 718/104
• US Class Current: 718/104
• CPC Class Codes: G06F 2209/509 Offload
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: H04L 49/70 Virtual switches

Abstract

A method for managing distributed computer data stream acceleration devices is provided that utilizes distributed acceleration devices on nodes within the computing system to process inquiries by programs executing on the computing system. The available nodes and available acceleration devices in the computing system are identified. In addition, a plurality of virtual acceleration device definitions is created. Each virtual acceleration device definition includes attributes used to configure at least one of the plurality of identified acceleration devices. When an inquiry containing an identification of computing system resources to be used in processing the inquiry is received, at least one virtual acceleration device definition that is capable of configuring an acceleration device in accordance with the computing system resources identified by the inquiry is identified. That acceleration device is configured in accordance with the identified virtual acceleration device definition and is used to process the inquiry.

Claims

1. A distributed computing system comprising:

a plurality of nodes;

a plurality of acceleration devices configurable in a plurality of distinct functions and disposed on one or more of the plurality of nodes;

a database comprising a plurality of virtual acceleration device definitions, each virtual acceleration device definition comprising attributes comprising an acceleration device function from the plurality of distinct functions and acceleration device hardware requirements and a compiled program and external configuration data that when loaded into an acceleration device having the hardware requirements configures that acceleration device to realize the acceleration device function;

an application programming interface to receive an inquiry comprising an identification of computing system resources to be used in processing the inquiry;

a resource requirements and preferences manager to identify at least one virtual acceleration device definition having a suitable acceleration device function and capable of configuring an acceleration device in accordance with the computing system resources identified in the inquiry; and

a deployment agent to configure at least one of the plurality of identified acceleration devices having the acceleration device hardware requirements in accordance with each identified virtual acceleration device definition.

2. The distributed computing system of claim 1, further comprising:

a plurality of discovery agents to detect the plurality of acceleration devices dynamically; and

a dynamic router controller to route computing system stream data to and from each identified acceleration device dynamically.

3. The distributed computing system of claim 2, wherein the dynamic router controller configures the acceleration device to receive packets across a network originating from one or more nodes within the computing system and to produce packets across the network destined to one or more of the nodes.

4. The distributed computing system of claim 1, further comprising a system scheduler to schedule and to dispatch at least one inquiry process element to process the inquiry, each inquiry process element utilizing at least one of the acceleration devices.

5. The distributed computing system of claim 4, wherein the system scheduler is further configured to identify a program corresponding to the scheduled process element during runtime, and the deployment agent is further configured to load the identified program into the acceleration device coincidental with dispatching the corresponding process element.

6. The distributed computing system of claim 4, wherein the database further comprises at least one policy to govern the use of acceleration devices within the computing system and the system scheduler performs scheduling of process elements utilizing acceleration devices and configuring the acceleration devices in accordance with the at least one policy.

7. The distributed computing system of claim 4, wherein the system scheduler identifies a first set of process elements, dispatches the first set of process elements to nodes utilizing acceleration devices and identifies a second set of process elements separate from the first set, wherein the process elements in the second set are not dispatched to nodes due to a lack of available acceleration devices.

8. The distributed computing system of claim 7, wherein:

the system scheduler identifies a revised first set of process elements comprises process elements utilizing one or more virtual acceleration device definitions that differ from the virtual acceleration device definitions utilized with the previous first set of process elements and identifies a revised second set of process elements; and

the deployment agent reconfigures one or more acceleration devices in accordance with the different virtual acceleration device definitions associated with the revised first set of process elements.

9. The distributed computing system of claim 7, wherein the system scheduler utilizes administrator policy specifications and user rules to rank processing elements and places each process element that utilizes the acceleration devices into either the first set of process elements or the second set of process elements in accordance with the rank.

10. The distributed computing system of claim 1, wherein the acceleration devices comprise a field programmable gate array, an application specific integrated circuit, a complex programmable logic device or combinations thereof.

11. The distributed computing system of claim 1, wherein the identified computing system resources comprise data stream acceleration devices required to process the inquiry and preferred data stream acceleration devices to be used in processing the inquiry.

12. The distributed computing system of claim 1, wherein database comprises a persistent memory storage location.