Abstract: A barrier synchronization system in a distributed computing environment. A group of members, including a plurality of related processes executing on one or more processors of the distributed computing environment, is provided a protocol having a dynamic number of synchronization phases. The protocol is voted on by one or more members of the group to determine how the protocol should proceed for the group of members. The protocol may complete successfully, continue to another synchronization phase or be rejected. The group of members determines how many synchronization phases the protocol is to include.

Abstract: Actions responsive to changes in the state of independent process groups whose process members are spread across a group of nodes in a distributed computing environment. The process groups are arranged in a hierarchy having source process groups and target process groups immediately above and dependent upon the source process groups. A change in the state or membership of a process group will result in a group services function of the nodes in the group injecting the change into the source group members of the affected group. When the action is successfully completed, the change is reflected up the hierarchy to the target group member or members. Upon successful action by the target group, the target group becomes the new source group which in turn may have target groups, and the changes are again reflected up the hierarchy to the new target groups.

Abstract: An application programming interface that unifies a plurality of mechanisms into a single framework. The interface includes a mechanism for communicating between members of a process group of related processes, and a mechanism for synchronizing the related processes of the process group. Additionally, the application programming interface may include mechanisms for managing membership of the process group or a processor group of processors, and/or a mechanism for controlling a group state value for the process group.

Abstract: A barrier synchronization program product in a distributed computing environment. A group of members, including a plurality of related processes executing on one or more processors of the distributed computing environment, is provided a protocol having a dynamic number of synchronization phases. The protocol is voted on by one or more members of the group to determine how the protocol should proceed for the group of members. The protocol may complete successfully, continue to another synchronization phase or be rejected. The group of members determines how many synchronization phases the protocol is to include.

Abstract: Actions responsive to changes in the state of independent process groups whose process members are spread across a group of nodes in a distributed computing environment. The process groups are arranged in a hierarchy having source process groups and target process groups immediately above and dependent upon the source process groups. A change in the state or membership of a process group will result in a group services function of the nodes in the group injecting the change into the source group members of the affected group. When the action is successfully completed, the change is reflected up the hierarchy to the target group member or members. Upon successful action by the target group, the target group becomes the new source group which in turn may have target groups, and the changes are again reflected up the hierarchy to the new target groups.

Abstract: Actions responsive to changes in the state of independent process groups whose process members are spread across a group of nodes in a distributed computing environment. The process groups are arranged in a hierarchy having source process groups and target process groups immediately above and dependent upon the source process groups. A change in the state or membership of a process group will result in a group services function of the nodes in the group injecting the change into the source group members of the affected group. When the action is successfully completed, the change is reflected up the hierarchy to the target group member or members. Upon successful action by the target group, the target group becomes the new source group which in turn may have target groups, and the changes are again reflected up the hierarchy to the new target groups.

Abstract: In accordance with a preferred embodiment of the present invention, a method and an apparatus is disclosed for providing a corresponding view of a partitioned distributed parallel processing environment. The parallel processing system itself is comprised of a plurality of nodes connected to one another and to a central control element via a network. The central control element is used for communication and status maintenance of the system environment. The apparatus and method provides for separate storing of system data pertaining to different sub-environments in a system data repository located in the central control element. Each data pertaining exclusively to a sub-environment is identified as such and limited access to that data is provided only to the nodes residing in that environment. A system view is constructed based on the data in this repository. Once a view is constructed, the view is identified and stored separately as belonging to a specific sub-environment.

Abstract: Managing membership of a domain of processors in a distributed computing environment including determining the lowest numbered node available in the domain and periodically sending grovel messages from the other available nodes to the lowest numbered node in the domain. Responsive to receiving all of the grovel messages by the lowest numbered node, the lowest numbered node performs a two phase commit whereby the lowest numbered node becomes a name server for managing the processors in the domain. The two phase commit includes sending an insert phase 1 (P1) message to the other available nodes from the lowest numbered node and, responsive to the insert P1 message, sending an acknowledge (ACK) message from the other available nodes to the lowest numbered node. Responsive to the ACK messages, a commit message is sent to the other available nodes notifying them that the lowest numbered node has become the name server.

Abstract: A method of deciphering all communications for a given sub-environment in a partitioned parallel processing environment is suggested through the use of daemons. The parallel processing environment is a UNIX based operating environment comprising of a plurality of nodes connected through a network to one another and to a central control element used for controlling all communications. The parallel processing environment initially operates as a single computing unit, which is later partitioned into two or more isolated sub-environments. The method comprises of storing all system data in a repository provided in the central control element, each data being organized as belonging exclusively to one sub-environment, or globally to all sub-environments. The central control element is then apportioned, so that the data belonging exclusively to one sub-environment is located at a given address, each sub-environment being assigned a different address.

Abstract: In accordance with a preferred embodiment of the present invention, an application and method of preserving persistent logical partitions, is suggested for use in a partitioned massively parallel processing system environment. All system data, including configuration data is stored in a central control element in a special configuration object class file. The configuration object class file maps the resources to different sub-environments. Upon a system shut-down or re-booting of any resource, the configuration information stored in the central control element is recovered to re-establish the sub-environments integrity.

Abstract: In accordance with a preferred embodiment of the present invention, a method and apparatus is disclosed for partitioning a massively distributed parallel processor into at least two isolated sub-environments. The parallel processing environment is comprised of a plurality of nodes connected to one another via a network in such a way as to form a single system unit initially. A central control element is provided for controlling and maintaining all communications and status information between nodes. The nodes and the central control element are in processing communication with one another via the network. All system data is deposited in a system data repository located in the central control element. Once the environment is partitioned, the system data stored in this repository is organized into system or partitioned classes, with the partitioned classes only being accessible selectively to nodes residing in one environment or another.

Abstract: A barrier synchronization technique in a distributed computing environment. A group of members, including a plurality of related processes executing on one or more processors of the distributed computing environment, is provided a protocol having a dynamic number of synchronization phases. The protocol is voted on by one or more members of the group to determine how the protocol should proceed for the group of members. The protocol may complete successfully, continue to another synchronization phase or be rejected. The group of members determines how many synchronization phases the protocol is to include.

Abstract: A system for managing membership of process groups of a distributed computing environment. Requests to alter membership of a first process group are received and presented to the first process group as a batched request. However, requests to change the membership of a second process group are received and presented to the second process group serially. Thus, the manner in which requests are presented can be different for various process groups. Additionally, different types of requests can be presented to the same process group in a differing manner. For instance, requests to join a process group can be presented as a batched request and requests to leave the process group can be presented serially, or vice versa.