International Business Machines Corporation

Richard K Errickson

Thomas Anthony Gregg

Peter Kenneth Szwed

Kulwant M Pandey

Leonard W Helmer, Jr

David F Craddock

John Scott Houston

Michael Peter Lyons

Ambrose A Verdibello, Jr

Donald W Schmidt

Charles W Gainey, Jr

Mark S Farrell

David Arlen Elko

Scott M Carlson

Carol B Hernandez

Bruce Marshall Walk

Frank William Brice

Patrick John Sugrue

Ping T Chan

Andrew W Piechowski

Gustav E Sittmann, Iii

Peter Dana Driever

Ugochukwu Charles Njoku

Ambrose Verdibello

Sean Liu

Welela Haileselaissie

Leornard W Helmer, Jr

Dan F Greiner

Damian Leo Osisek

Ronald Morton Smith, Sr

David Harold Surman

Richard Loise Arndt

Ronald Edward Fuhs

Dennis J Dahlen

Jeffrey M Turner

Steven N Goss

Donald Crabtree

Stephen R Burrow

Robert M Whalen, Jr

Scott E Davies

Jinghong Ma

Geoffrey A Crew

R Timothy Tomaselli

An Zhu

Embodiments include method, systems and computer program products for timing synchronization. Aspects include sending to a remote system a first base packet, the first base packet is sent at a first send event time and then recording the first send event time in a memory. Aspects also include sending to the remote system a second base packet, wherein the payload comprises instructions for the remote computing system and receiving a first remote packet, wherein a payload comprises a first received time indicating a time the first base packet was received. Aspects then include recording the first received time and recording a second received time, wherein the second received time indicates a time the first remote packet was received by the base system. Aspects include receiving a second remote packet, wherein a payload comprises a second sent time indicating a time the first remote packet was sent.

Timing synchronization

Disclosed are a method of and system for providing time synchronization among first and second computer systems, where each of the computer systems includes hardware, operating system software and a layer of microcode operating between said hardware and said software. The method comprises the steps of using the microcode of the first computer system to provide a first timestamp, using the microcode of the second computer system to provide a second timestamp and a third timestamp, and using the microcode of the first computer system to provide a fourth timestamp. The method comprises the further steps of using the first, second, third and fourth timestamps to determine a timing difference between the first and second computer systems, and adjusting the timing among said first and second computer systems on the basis of said determined timing difference. Preferably, the first and second computer systems are connected together by a point-to-point link; and the first timestamp is sent from the first computer system to the second computer system, and the second and third timestamps are sent from the second computer system to the first computer system over that point-to-point link.

Method and system for time synchronization among systems using parallel sysplex links

A computer-implemented method for message handling between a receiver and a sender of a throttling system is provided. The computer-implemented method includes receiving a first message that includes a completion queue element by the receiver from the sender. The computer-implemented method includes detecting that a number of outstanding completion queue elements in a completion queue meets or exceeds a high mark and instructing the sender to enter a throttle mode. The computer-implemented method includes receiving a second message without a completion queue element by the receiver from the sender. The computer-implemented method includes detecting that the number of the outstanding completion queue elements in the completion queue meets or is below a low mark and instructing the sender to exit the throttle mode.

Throttling backbone computing regarding completion operations

Embodiments are directed to a computer system for managing data transfer. The computer system includes a memory, a processor communicatively coupled to the memory, a send component and a receive component having a message queue and a controller. A single link interface communicatively couples the send component to the receive component. The single link interface includes a mainline channel and a sideband channel, and the computer system is configured to perform a method. The method includes transmitting mainline channel messages over the mainline channel from the send component to the receive component. The method further includes transmitting sideband channel messages over the sideband channel from the send component to the message queue of the receive component. The method further includes utilizing the controller to control a flow of the sideband channel messages to the message queue without relying on sending feedback to the send component about the flow.

Remotely controlled message queue

According to one embodiment, a method, computer system, and computer program product for computer networking is provided. The present invention may include a method of using media access control (MAC) multicast procedures to discover remote node addressability, and InfiniBand&#xae; (InfiniBand&#xae; and all InfiniBand&#xae; based trademarks and logos are trademarks or registered trademarks of System I/O, Incorporated and/or its affiliates) protocols to connect sets of queue pairs in implementing one or more communication channels, comprising preparing hardware to connect point-to-point links between a systems; discovering remote systems on the other side of the links through a discovery process; creating administrative traffic lanes; verifying the connection between nodes; initiating the auxiliary traffic flow interface; negotiating and connecting buffers; and exchanging node descriptors.

Method for remote node discovery and communication channel validation and connection

A method, computer program product, and system to control event logging and error recovery in a system including adapters, ports, and channels are described. The method includes storing a recovery threshold for each event type among a plurality of event types and storing a level-specific logging threshold for each event type, implementing event handlers for each of the channels, the ports, and the adapters of the system, and implementing a threshold manager for the events identified by the event handlers based on the level-specific logging threshold and the recovery threshold for each of the respective event types of each of the events. For any identified event corresponding with a given event type, the implementing the threshold manager includes considering the recovery threshold and the level-specific logging threshold at every level regardless of a level at which the identified event is identified.

Event logging and error recovery

Embodiments are directed to a computer system for managing data transfer. The computer system includes a memory, a processor communicatively coupled to the memory, a send component and a receive component having a message queue and a controller. A link interface communicatively couples the send component to the receive component. The link interface includes a mainline channel and a sideband channel, and the computer system is configured to perform a method. The method includes transmitting mainline channel messages over the mainline channel from the send component to the receive component. The method further includes transmitting sideband channel messages over the sideband channel from the send component to the message queue of the receive component. The method further includes utilizing the controller to control a flow of the sideband channel messages to the message queue without relying on sending feedback to the send component about the flow.

Embodiments relate to two general purpose computers connected in a peer-to-peer mode by connecting a cable (or wireless connection) between universal ports (e.g., PCIe ports) on each computer. A timing protocol utility runs on each computer to time schedule operations performed by its respective computer. Because the system clocks on each peer computer operate independently (asynchronously), they may vary somewhat from each other. To support time synchronized peer-to-peer operations, paired clock value (one for each peer computer) are generated continually and independently by each peer system. Each peer system periodically supplies the paired clock values to its associated timing protocol utility, which uses the paired clock values to time synchronize peer-to-peer computer operations. The timing protocol utilities may also exchange the paired clock values with each other for integrity checking and other operations.

Peer-to-peer networking through universal port connections

A system and method for correlating asynchronous operations via an operation identifier comprises receiving an originating operation from a first system that indicates a change in the first system and generating a first message with respect to the originating operation. The first message is associated with the operation identifier. The system and method further propagates the first message to a second system, which causes a subsequent operation being associated with the operation identifier to be performed by the second system, and correlates the originating operation and the subsequent operation via the operation identifier.

Correlating multiple disjoint events via an operation identifier

Growing community of inventors

Poughkeepsie, NY, United States of America

Richard K Errickson