公开(公告)号：US09311379B2

公开(公告)日：2016-04-12

申请号：US13710075

申请日：2012-12-10

Applicant: International Business Machines Corporation

Inventor： Daniel M. Hursh ,  Paul E. Movall ,  Eric J. Nelson ,  Brett J. Reese

IPC: G06F17/30 ,  G06F11/20 ,  G06F9/50

CPC classification number: G06F17/30578 ,  G06F9/5027 ,  G06F9/5072 ,  G06F11/2094

Abstract: A method may receive, at a first node, a first message with a first message data structure from a second node. The first message data structure may include a counter for each resource of a plurality of resources that is indicative of a most current age of a master copy of the resource known by the second node. The first node may include a current data structure with a counter for each resource that is indicative of a most current age of the master copy of the resource known by the first node. The method may determine whether at least one counter in the current data structure is older than the corresponding counter in the message data structure. The method may update the counter that is older, e.g., to learn the most current age of the master copy of the resource known by the second node, based on the determination.

Abstract translation: 方法可以在第一节点处从第二节点接收具有第一消息数据结构的第一消息。 第一消息数据结构可以包括指示由第二节点已知的资源的主副本的最新时代的多个资源中的每个资源的计数器。 第一节点可以包括当前数据结构，其中具有指示由第一节点已知的资源的主副本的最新时代的每个资源的计数器。 该方法可以确定当前数据结构中的至少一个计数器是否比消息数据结构中的相应计数器更早。 该方法可以基于该确定来更新较旧的计数器，例如，以了解第二节点已知的资源的主副本的最新时代。

公开(公告)号：US09197497B2

公开(公告)日：2015-11-24

申请号：US13743110

申请日：2013-01-16

Applicant: International Business Machines Corporation

Inventor： David R. Engebretsen ,  Henry J. May ,  Paul E. Movall

IPC: G06F15/177 ,  H04L12/24 ,  H04L12/931

CPC classification number: H04L41/0803 ,  H04L49/357

Abstract: A method includes receiving firmware at a network entity and determining whether the network entity is to operate as a controlling Fiber Channel forwarder based on the firmware.

Abstract translation: 一种方法包括在网络实体处接收固件，并且基于固件来确定网络实体是否作为控制光纤通道转发器进行操作。

公开(公告)号：US09077659B2

公开(公告)日：2015-07-07

申请号：US13710530

申请日：2012-12-11

Applicant: International Business Machines Corporation

Inventor： William J. Armstrong ,  Josep Cors ,  David R. Engebretsen ,  Paul E. Movall ,  Eric G. Thiemann

IPC: H04L12/28 ,  H04L12/721 ,  H04L12/931

CPC classification number: H04L45/44 ,  H04L49/355

Abstract: Techniques are provided for packet routing in a distributed network switch. The distributed network switch includes multiple switch modules operatively connected to one another, and each switch module includes multiple bridge elements and a management controller. In one embodiment, a shared interface routing (SIR) framework is provided that includes an analysis and bifurcation layer and at least one packet interface. A packet is received over a first logical network and via a physical port, the packet being destined for at least a first application executing on the management controller. The analysis and bifurcation layer analyzes the packet and sends the packet to the packet interface, which then routes the packet to the first application.

Abstract translation: 为分布式网络交换机中的数据包路由提供了技术。 分布式网络交换机包括可操作地彼此连接的多个交换机模块，并且每个交换机模块包括多个桥接元件和管理控制器。 在一个实施例中，提供了包括分析和分叉层以及至少一个分组接口的共享接口路由（SIR）框架。 通过第一逻辑网络并经由物理端口接收分组，该分组目的地是至少在管理控制器上执行的第一应用。 分析和分支层分析数据包，并将数据包发送到数据包接口，然后将数据包路由到第一个应用程序。

公开(公告)号：US20130208722A1

公开(公告)日：2013-08-15

申请号：US13709967

申请日：2012-12-10

Applicant: International Business Machines Corporation

Inventor： William J. Armstrong ,  Claude Basso ,  Josep Cors ,  David R. Engebretsen ,  Paul E. Movall ,  Eric Thiemann ,  Colin B. Verrilli ,  Bruce M. Walk

IPC: H04L12/56

CPC classification number: H04L45/44 ,  H04L49/355

Abstract: Techniques are provided for packet routing in a distributed network switch. The distributed network switch includes multiple switch modules operatively connected to one another, and each switch module includes multiple bridge elements and a management controller. In one embodiment, a shared interface routing (SIR) framework is provided that includes an analysis and bifurcation layer, at least one packet interface, and an analysis assist layer. A packet is received over a first logical network and via a physical port, the packet being destined for at least a first application executing on the management controller. The analysis assist layer analyzes the packet to determine a reason code to assign to the packet. The analysis and bifurcation layer then analyzes the packet based at least in part on the reason code.

Abstract translation: 为分布式网络交换机中的数据包路由提供了技术。 分布式网络交换机包括可操作地彼此连接的多个交换机模块，并且每个交换机模块包括多个桥接元件和管理控制器。 在一个实施例中，提供了共享接口路由（SIR）框架，其包括分析和分岔层，至少一个分组接口和分析辅助层。 通过第一逻辑网络并经由物理端口接收分组，该分组目的地是至少在管理控制器上执行的第一应用。 分析辅助层分析分组以确定分配给分组的原因代码。 分析和分岔层至少部分基于原因代码分析数据包。

公开(公告)号：US20140245085A1

公开(公告)日：2014-08-28

申请号：US13791289

申请日：2013-03-08

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION

Inventor： Paul S. HALVERSON ,  Grady W. Jensen ,  Paul E. Movall

IPC: G06F11/07

CPC classification number: G06F11/0766 ,  G06F11/0745 ,  G06F11/0778 ,  G06F11/0787

Abstract: An approach is described for managing error logs and corresponding error descriptors among a plurality of controller nodes of a distributed network fabric. According to such approach, an error log may include output from a runtime code trace. The runtime code trace may include storing trace statements from a currently running process in one or more of a plurality of trace buffers. The trace buffers may be memory-mapped circular buffers. Additionally, an error descriptor may be created from the error log. A globally unique identifier may be assigned to the error descriptor, and both the error descriptor and the globally unique identifier may be broadcast to each other controller node among the plurality of controller nodes.

公开(公告)号：US09148369B2

公开(公告)日：2015-09-29

申请号：US13709967

申请日：2012-12-10

Applicant: International Business Machines Corporation

Inventor： William J. Armstrong ,  Claude Basso ,  Josep Cors ,  David R. Engebretsen ,  Paul E. Movall ,  Eric Thiemann ,  Colin B. Verrilli ,  Bruce M. Walk

IPC: H04L12/28 ,  H04L12/721 ,  H04L12/931

CPC classification number: H04L45/44 ,  H04L49/355

Abstract: Techniques are provided for packet routing in a distributed network switch. The distributed network switch includes multiple switch modules operatively connected to one another, and each switch module includes multiple bridge elements and a management controller. In one embodiment, a shared interface routing (SIR) framework is provided that includes an analysis and bifurcation layer, at least one packet interface, and an analysis assist layer. A packet is received over a first logical network and via a physical port, the packet being destined for at least a first application executing on the management controller. The analysis assist layer analyzes the packet to determine a reason code to assign to the packet. The analysis and bifurcation layer then analyzes the packet based at least in part on the reason code.

Abstract translation: 为分布式网络交换机中的数据包路由提供了技术。 分布式网络交换机包括可操作地彼此连接的多个交换机模块，并且每个交换机模块包括多个桥接元件和管理控制器。 在一个实施例中，提供了共享接口路由（SIR）框架，其包括分析和分岔层，至少一个分组接口和分析辅助层。 通过第一逻辑网络并经由物理端口接收分组，该分组目的地是至少在管理控制器上执行的第一应用。 分析辅助层分析分组以确定分配给分组的原因代码。 分析和分岔层至少部分基于原因代码分析数据包。

公开(公告)号：US09087085B2

公开(公告)日：2015-07-21

申请号：US13710132

申请日：2012-12-10

Applicant: International Business Machines Corporation

Inventor： Daniel M. Hursh ,  Paul E. Movall ,  Eric J. Nelson ,  Brett J. Reese

IPC: G06F17/30

CPC classification number: G06F17/30289 ,  G06F17/30194 ,  G06F17/30424 ,  G06F17/30575

Abstract: A method of managing a database may include, at a node, assigning a pre-assimilation value to a hardware instance identifier of a hardware component before the node is assimilated into a multi-node computing system. The method may also include receiving a post-assimilation value after the node has been assimilated into the multi-node computing system. The method may also include updating the pre-assimilation value with the post-assimilation value.

Abstract translation: 管理数据库的方法可以包括在节点将节点同化到多节点计算系统之前，将硬件实例标识符的预同化值分配给硬件组件的节点。 该方法还可以包括在节点被同化到多节点计算系统中之后接收后同化值。 该方法还可以包括利用后同化值更新预同化值。

公开(公告)号：US08996925B2

公开(公告)日：2015-03-31

申请号：US13791289

申请日：2013-03-08

Applicant: International Business Machines Corporation

Inventor： Paul S. Halverson ,  Grady W. Jensen ,  Paul E. Movall

IPC: G06F11/00 ,  G06F11/14 ,  G06F11/07

CPC classification number: G06F11/0766 ,  G06F11/0745 ,  G06F11/0778 ,  G06F11/0787

Abstract: An approach is described for managing error logs and corresponding error descriptors among a plurality of controller nodes of a distributed network fabric. According to such approach, an error log may include output from a runtime code trace. The runtime code trace may include storing trace statements from a currently running process in one or more of a plurality of trace buffers. The trace buffers may be memory-mapped circular buffers. Additionally, an error descriptor may be created from the error log. A globally unique identifier may be assigned to the error descriptor, and both the error descriptor and the globally unique identifier may be broadcast to each other controller node among the plurality of controller nodes.

Abstract translation: 描述了用于在分布式网络结构的多个控制器节点之间管理错误日志和相应的错误描述符的方法。 根据这种方法，错误日志可以包括来自运行时代码跟踪的输出。 运行时代码跟踪可以包括在多个跟踪缓冲器中的一个或多个中存储来自当前运行的进程的跟踪语句。 跟踪缓冲区可以是内存映射循环缓冲区。 另外，可以从错误日志创建错误描述符。 可以将全局唯一标识符分配给错误描述符，并且错误描述符和全局唯一标识符都可以广播到多个控制器节点中的彼此控制器节点。

公开(公告)号：US08990642B2

公开(公告)日：2015-03-24

申请号：US13773987

申请日：2013-02-22

Applicant: International Business Machines Corporation

Inventor： Paul S. Halverson ,  Grady W. Jensen ,  Paul E. Movall

IPC: G06F11/00 ,  G06F11/14 ,  G06F11/07

CPC classification number: G06F11/0766 ,  G06F11/0745 ,  G06F11/0778 ,  G06F11/0787

Abstract: An approach is described for managing error logs and corresponding error descriptors among a plurality of controller nodes of a distributed network fabric. According to such approach, an error log may include output from a runtime code trace. The runtime code trace may include storing trace statements from a currently running process in one or more of a plurality of trace buffers. The trace buffers may be memory-mapped circular buffers. Additionally, an error descriptor may be created from the error log. A globally unique identifier may be assigned to the error descriptor, and both the error descriptor and the globally unique identifier may be broadcast to each other controller node among the plurality of controller nodes.

公开(公告)号：US20140201338A1

公开(公告)日：2014-07-17

申请号：US13743110

申请日：2013-01-16

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION

Inventor： David R. Engebretsen ,  Henry J. May ,  Paul E. Movall

IPC: H04L12/24

CPC classification number: H04L41/0803 ,  H04L49/357

Abstract: A method includes receiving firmware at a network entity and determining whether the network entity is to operate as a controlling Fiber Channel forwarder based on the firmware.

Abstract translation: 一种方法包括在网络实体处接收固件，并且基于固件来确定网络实体是否作为控制光纤通道转发器进行操作。