Abstract: 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.

Abstract: An apparatus and method for controlling resources in a computing system including receiving an allocation request for a resource; determining whether an allocation limit for the resource has been reached; and, restricting access to the resource upon determination that the allocation limit has been reached.

Abstract: In a method and apparatus for retrieving data from a snapshot data storage system, for each epoch, a snapshot including (i) all changed data, and (ii) an index is created. The index includes an entry for each page that has changed during the epoch. For rapidly retrieving the data as of any given time, the method creates a hash table that includes an entry for each data page that has changed since the baseline was created. The hash table entry indicates the epoch in which the data most recently changed and an offset corresponding to the location of the changed data in the corresponding snapshot. The hash table is created by inserting an entry for each page in the most recent index, and then examining the remaining indices for all other snapshots from the most recent to the oldest snapshot and adding any non-duplicate entries into the table.

Abstract: 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.

Abstract: Log-Based Rollback Recovery for system failures. The system includes a storage medium, and a component configured to transition through a series of states. The component is further configured to record in the storage medium the state of the component every time the component communicates with another component in the system, the system being configured to recover the most recent state recorded in the storage medium following a failure of the component.