公开(公告)号：GB2303474A

公开(公告)日：1997-02-19

申请号：GB9514819

申请日：1995-07-19

Applicant: IBM

Inventor： BANKS TIMOTHY W ,  HUNTER IAN ,  NORMINGTON GLYN ,  ZIMMER DENNIS JACK ,  LUPTON PETER JAMES

IPC: G06F12/00 ,  G06F9/46 ,  G06F13/00 ,  G06F15/00 ,  G06F15/16 ,  G06F11/14

Abstract: Disclosed is a method for providing synchronization of a transaction in a data processing system where the transaction involves a plurality of agents participating in the transaction and one coordinator for coordinating the transaction. The agents include at least a middleman which coordinates a set of at least one of the agents. A vote is sent indicating the availability or non-availability to commit from each of the agents to the coordinator. The coordinator determines a commit or backout decision when all the votes are received. A Not Known decision message is sent from the middleman to each of the agents in the set available to commit if the decision is not available at the middleman because of a failure. The Not Known decision message stops resynchronization of each of the available agents. If the middleman has responsibility for resynchronization then, on availability of the decision at the middleman, a resynchronization message is sent to each of the available agents. If the middleman does not have responsibility for resynchronization, then a Known decision message is sent to each of the available agents from the middleman. Receipt of the Known decision message starts resynchronization of each of the available agents.

公开(公告)号：GB2301686A

公开(公告)日：1996-12-11

申请号：GB9511247

申请日：1995-06-03

Applicant: IBM

Inventor： BANKS TIMOTHY W

IPC: G06F12/00 ,  G06F9/46 ,  G06F9/52 ,  G06F15/00 ,  G06F11/14

Abstract: Disclosed is a method for providing synchronisation of a transaction in a data processing system where the transaction involves an initiator node (205), which starts the synchronisation, and a coordinator node (215), which decides the outcome of the transaction. They communicate through a routing node (210) having no resources of its own which are modified in the transaction. The synchronisation is provided by establishing a first conversation (220) between the initiator and routing nodes and a second conversation (225) between the routing and coordinator nodes. A first commit request message (240) is then sent from the initiator node to the routing node. A second commit request message (245) is sent from the routing node to the coordinator node. An additional step is performed of storing asynchronously, checkpoint information to non-volatile storage in the routing node after establishing the first and second conversations and before sending the second commit request message. The routing node acts as just a routing node and does not save checkpoint states as if it were a node with a real need to do so. In the event of failure, the routing node only recovers the route end points of the conversation it was routing when failure occurred. Then resynchronisation protocols are carried out between the end points with the routing node not aware of their contents, except to monitor when it can safely forget its routing information.