Abstract: A method, system, and computer usable program product for local synchronization in a memory hierarchy in a multi-core data processing system are provided in the illustrative embodiments. A request to acquire a reservation for a reservation granule is received at a first core. The reservation is acquired in a first local cache associated with the first core in response to a cache line including the reservation granule being present and writable in the first local cache. A conditional store request to store at the reservation granule is received at the first core. A determination is made whether the reservation remains held at the first local cache. The store operation is performed at the first local cache responsive to reservation remaining held at the first local cache.

Abstract: A data storage system in which each of a set of disks has a first portion for either popular data or error correction information (parity or Reed-Solomon code symbols) associated with the popular data, and a second portion used for other data. A disk controller connected to the set of disks maintains a first popular data block in the first portion of a first of the set of disks and a second popular data block in the first portion of a second of the set of disks. The system maintains at least two of the disks in an active state. The active disks are selected to insure that any data in the popular data blocks can be determined from data stored in the active disks. An additional disk is maintained in an active state if write access is permitted or if there is an additional popular data block.

Abstract: A data storage system includes a set of disks where each disk is operable in a plurality of discrete angular velocity levels. A disk controller controls the angular velocity of each active disk. The controller replicates a first portion of data on a plurality of the disks stores a second class of data in the set of disks without replication. The disk controller routes data requests to one of the active disks based, at least in part, on the current loading of the active disks to maintain balanced loading on the active disks. The disk controller alters the angular velocity of at least one of the active disks upon detecting that the latency of one or more of the data requests differs from a specified threshold. In this manner, the disk controller maintains the angular velocity of the active disks at approximately the same minimum angular velocity needed to attain acceptable performance. The disk controller may replicate the first portion of data on each of the disks in the set of disks.

Abstract: A system and method for operating a server cluster that includes a set of server devices connected to a local area network (LAN). Each server device maintains a directory of the contents of its file cache. When a decrease in server cluster traffic is detected, a server device on the server cluster is selected for powering down. Prior to powering down a server device, the device's file cache directory is broadcast over the LAN to each of the other server devices on the cluster. If a subsequent request for a file stored in the powered-down server's file cache is received by the cluster, the request is routed to one of the remaining active server devices. This server device then retrieves the requested file from the powered-down server's file cache over the LAN. Prior to broadcasting the file cache directory, pending client requests on the selected server device are completed. The powered-down server may continue to provide power to its NIC and system memory while the processor is deactivated.

Abstract: A method and system for responding to a client request for information from a service device in a data processing network. Initially, the server receives the information request from the client and determines that the requested information is not in the server's cache. The server then generates a storage request and sends the storage request to a network attached storage device. The storage device retrieves the requested information and generates a set of packets containing the requested information. The storage device then sending the generated packets simultaneously to the client to satisfy the client request and to the server to refresh the server cache. The storage request may include protocol information corresponding to the client-server connection that the storage device uses to replicate the client-server protocol stack in the generated packet. Sending the generated packet may comprise including a multicast address in the packet.

Abstract: A disk controller connected to a set of disks in a data storage system is configured to mirror a first class of data on multiple disks while storing a second class of data without mirroring. The controller maintains at least one of the disks containing the first class of data in an operational state while maintaining at least one of the remaining disks in a low power state. The first class of data may contain popular files while the second class of data contains unpopular files. The first data class may be mirrored on each of the disks. The data in the second class may be stored with parity where the parity information is maintained on a single disk. The operational disk may be changed following access to data from the second class of data that is stored on one of the low power state disks.

Abstract: A data processing network in which console interactions are communicated to and from server appliances over the network. The system may include a server appliance configured to re-direct serial port transactions to a network port. The service appliance may include a mechanism for transmitting and receiving console data and control information via the network. The system further includes a console server for accepting and displaying console traffic that is sent over the network by a server appliance and for transmitting commands entered by a user back to the server appliance for processing.

Abstract: A data processing system for communicating over a network includes at least two network connection devices. The system selects one of the connection devices in response to a request for a network communication. The system may deactivate at least one of the connection devices not selected and initiate the requested network communication using the selected connection device. The system may select the optimal connection device for the particular communication based upon a set of parameters associated with each connection device. These parameters may include connection device's bandwidth, cost, and security. One or more of the parameter values for each connection devices may be negotiated with a service provider. The system may deactivate at least one of the connection devices by turning off all of the connection devices not selected for the network communication. The system may further turn off the selected connection device after the communication is terminated.

Abstract: A disk controller connected to a set of disks in a data storage system is configured to mirror a first class of data on multiple disks while storing a second class of data without mirroring. The controller maintains at least one of the disks containing the first class of data in an operational state while maintaining at least one of the remaining disks in a low power state. The first class of data may contain popular files while the second class of data contains unpopular files. The first data class may be mirrored on each of the disks. The data in the second class may be stored with parity where the parity information is maintained on a single disk. The operational disk may be changed following access to data from the second class of data that is stored on one of the low power state disks.

Abstract: The distribution of power dissipation within cluster systems is managed by a combination of inter-node and intra-node policies. The inter-node policy consists of subdividing the nodes within the cluster into three sets, namely the “Operational” set, the “Standby” set and the “Hibernating” set. Nodes in the Operational set continue to function and execute computation in response to user requests. Nodes in the Standby set have their processors in the low-energy standby mode and are ready to resume the computation immediately. Nodes in the Hibernating set are turned off to further conserve energy, and they need a relatively longer time to resume operation than nodes in the Standby set. The inter-node policy further distributes the computation among nodes in the Operational set such that each node in the set consumes the same amount of energy.

Abstract: A system and method for operating a server cluster that includes a set of server devices connected to a local area network (LAN). Each server device maintains a directory of the contents of its file cache. When a decrease in server cluster traffic is detected, a server device on the server cluster is selected for powering down. Prior to powering down a server device, the device's file cache directory is broadcast over the LAN to each of the other server devices on the cluster. If a subsequent request for a file stored in the powered-down server's file cache is received by the cluster, the request is routed to one of the remaining active server devices. This server device then retrieves the requested file from the powered-down server's file cache over the LAN. Prior to broadcasting the file cache directory, pending client requests on the selected server device are completed. The powered-down server may continue to provide power to its NIC and system memory while the processor is deactivated.

Abstract: A method and system for responding to a client request for information from a service device in a data processing network. Initially, the server receives the information request from the client and determines that the requested information is not in the server's cache. The server then generates a storage request and sends the storage request to a network attached storage device. The storage device retrieves the requested information and generates a set of packets containing the requested information. The storage device then sending the generated packets simultaneously to the client to satisfy the client request and to the server to refresh the server cache. The storage request may include protocol information corresponding to the client-server connection that the storage device uses to replicate the client-server protocol stack in the generated packet. Sending the generated packet may comprise including a multicast address in the packet.

Abstract: A data processing network in which console interactions are communicated to and from server appliances over the network. The system may include a server appliance configured to re-direct serial port transactions to a network port. The service appliance may include a mechanism for transmitting and receiving console data and control information via the network. The system further includes a console server for accepting and displaying console traffic that is sent over the network by a server appliance and for transmitting commands entered by a user back to the sever appliance for processing.

Abstract: Configuration changes are dynamically applied to a cluster multiprocessing system by enqueuing a configuration change event. When the configuration change event is processed, the prior configuration is backed up and each software component applies a relevant portion of a configuration change transaction in an ordered, synchronized manner. Each software component applies its portion of the transaction either by reinitialization or a logged transition operation. If the configuration change transaction fails, the software components roll back the portions, of the configuration change already applied in an ordered, synchronized manner to restore the prior configuration. Multiple events for different configuration changes may be enqueued.