公开(公告)号：GB2531341A

公开(公告)日：2016-04-20

申请号：GB201418447

申请日：2014-10-17

Applicant: IBM

Inventor： WAKELIN PHILIP ,  ROEGER BRAD ,  BROOKS MICHAEL DAVID ,  HE JENNY JING ,  ALLEN GERALD MARTYN

IPC: H04L45/24 ,  H04L45/28

Abstract: The invention provides reconnection of a client to a server in a transaction processing server cluster having a single address in order to resolve partially completed transactions, such as in doubt units-of-work. The method carried out at a client includes: establishing a connection with a first server in the server cluster and requesting units of work at the first server; detecting a failure of the connection to the first server; establishing a connection with a second server in the server cluster and requesting units of work at the second server; receiving a reconnection request from the first server, wherein the first server has unresolved units of work; terminating the connection with the second server including completing current units of work; and reconnecting with the first server and resynchronizing the unresolved units of work with the first server. The reconnection request may include a flag indicating that it is for a previously established connection and the server cluster to which the first server belongs.

公开(公告)号：GB2527798A

公开(公告)日：2016-01-06

申请号：GB201411823

申请日：2014-07-02

Applicant: IBM

Inventor： BROOKS MICHAEL ,  ALLEN GERALD ,  HE JENNY JING ,  WAKELIN PHILIP

IPC: G06F11/20 ,  H04L45/28

Abstract: Disclosed is a method of recovering during synchronization operations. The system includes a first recovery management connector at a first region of the transaction processing system which carries out the method by identifying a failure of a network connection during synchronization operations with a partner region, querying connections defined to the transaction processing system to identify a replacement connection that is connected to the partner region using a different network type to the failed network connection, the replacement connection having a second recovery management connector in the first region. Then collecting relevant data for an interrupted synchronization operation, issuing a command to link to a specific program at a first recovery management connector in the partner region, the command including the relevant data for the interrupted synchronization operation, and using the second recovery management connector at the first region to communicate the command and a response from the partner region via the replacement connection in order to continue the interrupted synchronization operation.

公开(公告)号：GB2503920A

公开(公告)日：2014-01-15

申请号：GB201212426

申请日：2012-07-12

Applicant: IBM

Inventor： WEBSTER MATTHEW ALEXANDER ,  BOHM FRASER ,  MITCHELL IAN ,  WAKELIN PHILIP

IPC: G06F9/54

Abstract: An application server identifies a software object to be executed and an application context which enables a service. An entry point of the object is identified and a tag identifying the context is associated with the entry point. The object is executed using the entry point and the tag. The tag may be passed directly to the object via a memory reference. Alternatively, the tag may be placed in a file referenced by the object. The object may be application software. The application context may identify objects, which provide a service. The objects may execute in a cloud computing environment.