Dynamic multiple cluster system reconfiguration

• US 7,577,727 B2
• Filed: 06/27/2003
• Issued: 08/18/2009
• Est. Priority Date: 06/27/2003
• Status: Expired due to Fees

• Current Assignee: Sanmina-SCI Corporation
• Original Assignee: Newisys, Inc. (Sanmina-SCI Corporation)
• Inventors: Glasco, David B., Zeitler, Carl, Prasadh, Guru, Newawarker, Shashank, Kota, Rajesh
• Primary Examiner(s): Etienne, Ario
• Assistant Examiner(s): NASH, LASHANYA RENEE
• Application Number:
  US10/607,819
• Publication Number:
  US 20050021699A1
• Time in Patent Office: 2,244 Days
• Field of Search: 709/203, 709220-222, 709/227, 712/15, 712/1, 712/10
• US Class Current: 709/221
• CPC Class Codes: G06F 15/17337 Direct connection machines, e.g. completely connected computers, point to point communication networks coupling between buses G06F13/4004

Abstract

According to the present invention, methods and apparatus are provided to allow dynamic multiple cluster system configuration changes. In one example, processors in the multiple cluster system share a virtual address space. Mechanisms for dynamically introducing and removing processors, I/O resources, and clusters are provided. The mechanisms can be implemented during reset or while a system is operating. Links can be dynamically enabled or disabled.

Claims

1. A method comprising:

terminating a plurality of applications running in an operating system environment that supports dynamic removal of a first processor cluster from a plurality of processor clusters, the first processor cluster including a first plurality of processors and a first interconnection controller interconnected using a point-to-point architecture;

identifying the first processor cluster for removal;

flushing a plurality of caches associated with the first processor cluster;

modifying a plurality of routing tables associated with each processor cluster to reflect removal of the first processor cluster;

disabling link layer communications associated with the first processor cluster, wherein the first processor cluster is disconnected after disabling link layer communications associated with the first processor cluster;

maintaining physical layer communications associated with the first processor cluster to allow connection of a replacement processor cluster.

2. The method of claim 1, wherein a fence bit is written by a service processor.

3. The method of claim 1, wherein a fence bit is written by a JTAG interface associated with a processor.

4. The method of claim 1, wherein the first cluster of processors and a second cluster of processors share a single virtual address space.

5. The method of claim 4, wherein the second cluster of processors includes a second interconnection controller.

6. The method of claim 5, wherein the second interconnection controller includes a physical layer enable indicator.

7. The method of claim 5, wherein the second interconnection controller includes a fence indicator configurable to prevent the transmission of logical packets between the first interconnection controller and the second interconnection controller.

8. The method of claim 5, wherein the second interconnection controller includes a reinitialization indicator configurable to direct the second interconnection controller to reinitialize the link.

9. An apparatus comprising:

means for terminating a plurality of applications running in an operating system environment that supports dynamic removal of a first processor cluster from a plurality of processor clusters, the first processor cluster including a first plurality of processors and a first interconnection controller interconnected using a point-to-point architecture;

means for identifying a first processor cluster for removal;

means for flushing a plurality of caches associated with the first processor cluster;

means for modifying a plurality of routing tables associated with each processor cluster to reflect removal of the first processor cluster;

means for disabling link layer communications associated with the first processor cluster, wherein the first processor cluster is disconnected after disabling link layer communications associated with the first processor cluster;

means for maintaining physical layer communications associated with the first processor cluster to allow connection of a replacement processor cluster.

10. The apparatus of claim 9, wherein a fence bit is written by a service processor.

11. The apparatus of claim 9, wherein a fence bit is written by a JTAG interface associated with a processor.

12. The apparatus of claim 9, wherein the first cluster of processors and the second cluster of processors share a single virtual address space.