A distributed system for creating a checkpoint for a plurality of processes running on the distributed system. The distributed system includes a plurality of compute nodes with an operating system executing on each compute node. A checkpoint library resides at the user level on each of the compute nodes, and the checkpoint library is transparent to the operating system residing on the same compute node and to the other compute nodes. Each checkpoint library uses a windowed messaging logging protocol for checkpointing of the distributed system. Processes participating in a distributed computation on the distributed system may be migrated from one compute node to another compute node in the distributed system by re-mapping of hardware addresses using the checkpoint library.

Transparent checkpointing and process migration in a distributed system

A node ( 101 , FIG. 1 ), within a networked computer system ( 100 ), is capable of supporting communications with other nodes relating to operating multiple application instances in a master-slave configuration. Each node periodically generates and sends ( 520, 526 , FIG. 5 ) a Heartbeat message (FIG. 8 ) that indicates the operational status for one or more application instances being managed by the node. When a node receives a Heartbeat message from a remote node, it evaluates (FIG. 10 ) the Heartbeat information for each application instance reported in the message, and takes any appropriate actions. The node also determines ( 1206 , FIG. 12 ) whether new configuration information should be obtained for each of the application instances the node is managing, and requests ( 1210 , FIG. 12 ) that new configuration information, when necessary.

Method and apparatus for exchanging heartbeat messages and configuration information between nodes operating in a master-slave configuration

A program control device operating in a distributed environment comprised of hosts instantiating copies of a scalable application, the program control device controlling start up, shutdown or movement of a selected one of the copies based on signals generated in response to first information regarding performance of all copies of the scalable application and second infomation regarding performance of the hosts.

Program control for resource management architecture and corresponding programs therefor

A system and method for providing process persistence in a virtual machine are described. A virtual persistent heap may be provided. The virtual persistent heap may enable the checkpointing of the state of the computation of a virtual machine, including processes executing within the virtual machine, to a persistent storage such as a disk or flash device for future resumption of the computation from the checkpoint. The Virtual Persistent Heap also may enable the migration of the virtual machine computation states, and thus the migration of executing processes, from one machine to another. The saved state of the virtual machine heap may also provide the ability to restart the virtual machine after a system crash or shutdown to the last saved persistent state, and to thus restart a process that was running within the virtual machine prior to the system crash or shutdown to a checkpointed state of the process stored in the virtual persistent heap. This persistent feature is important for small consumer and appliance devices, as these appliances may be shutdown and restarted often.

Process persistence in a virtual machine

Determining metadata associated with a graph-based computation includes functionally transforming metadata associated with a first portion (2310) of a graph to generate transformed metadata associated with a second portion (2316) of the graph; determining a third portion (2317) of the graph related to the second portion of the graph; and propagating the transformed metadata from the second portion of the graph to the third portion of the graph.

Managing metadata for graph-based computations

A checkpoint of a parallel program is taken in order to provide a consistent state of the program in the event the program is to be restarted. Each process of the parallel program is responsible for taking its own checkpoint, however, the timing of when the checkpoint is to be taken by each process is the responsibility of a coordinating process. During the checkpointing, various data is written to a checkpoint file. This data includes, for instance, in-transit message data, a data section, file offsets, signal state, executable information, stack contents and register contents. The checkpoint file can be stored either in local or global storage. When it is stored in global storage, migration of the program is facilitated. When a parallel program is to be restarted, each process of the program initiates its own restart. The restart logic restores the process to the state at which the checkpoint was taken.

Method of performing checkpoint/restart of a parallel program

Program products for performing checkpoint/restart of a parallel program

The management and use of replicated distributed transactions is facilitated. A distributed synchronous transaction system protocol is provided to manage the replication of distributed transactions for client application instances. The distributed synchronous transaction system allows transactions to be replicated without having the client application instances be aware of other instances to receive the transaction. Further, if a failure occurs during processing of a distributed replicated transaction, the distributed synchronous transaction system manages the recovery of the failure.

Synchronous replication of transactions in a distributed system

A transaction is initiated within the distributed computer environment, through an entity of a client application. The initiated transaction is copied to at least one other entity of the client application, whose presence is hidden before initiating the transaction. Independent claims are also included for the following: (a) Synchronous transaction duplication execution system; (b) Storage medium with synchronous transaction duplication execution program

Synchronous transaction duplication execution in distributed computer environment, involves hiding presence of specific entity of client application, before initiating transaction within distributed environment

Anton Prenneis

Papers

Method of performing checkpoint/restart of a parallel program

Program products for performing checkpoint/restart of a parallel program

Synchronous replication of transactions in a distributed system

Synchronous transaction duplication execution in distributed computer environment, involves hiding presence of specific entity of client application, before initiating transaction within distributed environment