Abstract: A fault tolerant computing system comprises a plurality of processing nodes interconnected by a communication medium for parallel-running identical application programs. A fault detector is connected to the processing nodes via the communication medium for periodically collecting configuration status data from the processing nodes and mutually verifying the collected configuration status data for detecting an abnormal node. In one preferred embodiment of this invention, the system operates in a version diversity mode in which the processing nodes are configured in a substantially equal configuration and the application programs are identical programs of uniquely different software versions. In a second preferred embodiment, the system operates in a configuration diversity mode in which the processing nodes are respectively configured in uniquely different configurations.

Abstract: The present disclosure is directed toward a method for restoring a computer processor to a previous state. Described is a processor/memory architecture that may store successive instructions/data into a pushdown stack. As instructions are loaded and executed, the loading and executing of new instructions may be suspended. The instruction execution and memory stack then may be restored to a previous processor state in terms of instructions, processor memory state, register values, etc.

Abstract: Enhancements to the functionality of a file system are provided. A file system can provide an additional level of error correction, in addition to the error correction provided by a data storage device. An extension to the metadata area in a file system can store index information about data files that the file system can use to reduce the time it takes to access the files. A file system can use extended metadata to store conditional access information for data files stored in the file system. A data storage device can also examine and characterize the data being stored in order to categorize a user's files automatically. Modules can be loaded into a file system to provide new functionality to the file system. The modules can load additional modules to extend their functionality.

Abstract: In one aspect of the invention, a method is provided. The method may include: (1) storing a snapshot of a system state of a node; (2) executing a job on the node; and (3) restoring the node to the system state using the stored snapshot of the system state.

Abstract: A method and apparatus for delaying the process of freeing data blocks during the processing of a non-volatile transactions log are disclosed. Accordingly, after a system failure, when a storage system is replaying transactions in a non-volatile transactions log, the storage system flushes the data generated from the transactions to disk storage without freeing and data blocks which, as a result of processing the transactions log, are no longer storing data that is part of the active file system. Consequently, the replay of the transactions log takes less time, and the storage system is brought online more quickly.

Abstract: A mechanism is provided for providing reliability of communication. A first processor determines a current state of links coupled to ports of a first processor of the data processing system. Each port of the first processor comprises a plurality of links to a corresponding port on a second processor of the data processing system. The current state of the links indicates a level of error associated with each link. The first processor determines, for each link, if a level of error associated with the link exceeds a threshold. For each link whose level of error exceeds the threshold, the first processor tags the link with an error identifier in a switch associated with the ports of the first processor. The first processor reduces a level of usage for transmitting data on ports associated with links tagged with the error identifier.

Abstract: In general, the invention facilitates diagnosing fault conditions, such as flapping, by permitting users to request information for specific components in a network device such as a router. The invention also facilitates the diagnosis of other fault conditions, including, but not limited to, excessive numbers of dropped packets, hard drive crashes, high temperature readings, and inactive interface cards. A user may obtain a targeted log containing information relating to selected fault conditions or other network device events, rather than a system log containing information relating to all network device events, some of which may not be of interest to the user. The targeted log may be parsed and analyzed with greater ease than the system log.

Abstract: A transport-neutral in-order delivery in a distributed environment is provided. Typically, in-order delivery guarantees that sequential orders received by a transport engine are sent out in the same order they are received. Such delivery may be forwarded either to another transport engine or to some application. In case there is a failure of delivery of messages in a stream, the messages are either resubmitted, suspended, or moved to backup. A user or administrator can configure the desired action. Additionally, any stream can be manually aborted or a specified port can be unenlisted. Deliverable streams of messages are locked on to a back-end transport engines or applications and dequeued sequentially unless one of the above mentioned failure scenarios occurs.

Abstract: Disclosed is a method for replicating a software application in a multi-computer architecture (cluster). The software application is executed on a first computer of the cluster, which represents a primary or operation node, and is replicated on at least one other computer of the cluster, which represents a secondary node, including replication of the resources associated with the software application. The method includes streamlined updating of the replicated resources via a dynamic introspection mechanism supplying the structure of the application that is to be replicated and the dynamic graph of the implemented resources and dependencies.

Abstract: A method and system for testing a modular data-processing component. Register information associated with a modular data-processing component to be tested at a test location can be identified and stored. The modular data-processing component can then be tested and removed from said test location. Thereafter, the register information can be retrieved and provided for use with testing of a new data-processing component at said test location without losing said register information during testing of multiple modular data-processing components. The register information can be, for example, PCI configuration data and the modular data-processing component can be an HAB.

Abstract: A computing device comprising a plurality of devices and a plurality of visual indicators is disclosed. Each of the plurality of visual indicators is interconnected with a power source and is associated with a different one of the plurality of devices. None of the plurality of visual indicators comprises a capacitor.

Abstract: A circuit and method for providing a power-on self test capability for peripheral devices that allows direct testing of address-line data. The preferred embodiment includes a multiplexer circuit that allows the read address line outputs of a microprocessor to be directly returned to the data inputs of the microprocessor, thus providing a direct verification of the integrity of the read-address connection.

Abstract: This invention relates to a method for the transmission of logging data, within a cluster of computers, enabling all or part of the events constituting the running of a user process to be memorized or reproduced. The method applies in particular to the logging data representing the internal events of this process. When the functioning of the primary user memory space is managed by system software managing first software and hardware communication resources (OPsock, OPM) to enable an exchange of data with another node (SB), this method comprises transmitting the logging data from the first user memory space (OPU) to second software and hardware communication resources (HSI) in order to transmit them to this secondary node at a rhythm globally equivalent to that of the execution of the logged process.

Abstract: A method and an apparatus to aggressively detect media errors on storage devices with negligible performance impact have been disclosed. In one embodiment, the method includes computing a first data rate at or above which a scrub on a set of storage devices can be completed within a predetermined period of time, and running the scrub on the storage devices at a second data rate without interfering with the storage devices servicing client requests to access the storage devices, the second data rate being at or above the first data rate.

Abstract: In response to determining that computer generation of an abstract workflow plan for a network-accessible service resulted in failure, a method attempts to automatically recover from the failure. Where attempting to automatically recover from this failure is unsuccessful, the method applies remedy rules, to assist in debugging the failure. In response to determining that computer assignment of a physical plan for the abstract workflow plan resulted in failure, the method attempts to automatically recover from the failure. Where attempting to automatically recover from this failure is unsuccessful, the method applies remedy rules, to assist in debugging the failure. In response to determining that computer evaluation of a runtime configuration for the physical plan resulted in a failure, the method attempts to automatically recover from the failure. Where attempting to automatically recover from this failure is unsuccessful, the method applies remedy rules, to assist in debugging the failure.

Abstract: The present embodiments perform active volume restores on actively running initial operating systems in an active partition. The methods identify available memory within a partition, duplicate at least a portion of an initial operating system image into the identified available memory providing a duplicate operating system, remap access from an initial operating system to the at least the duplicate operating system image, store a temporary operating system on to the partition, reboot into the temporary operating system, operate through the temporary operating system and store a restore operating system into the partition; and reboot into the restore operating system.

Abstract: An apparatus and method for estimating the proficiency of a software test according to EMS messages extracted from a code base. Embodiments of the invention are generally directed to providing some measurement of the proficiency of a test script for testing a software application. In one embodiment, data is collected on event messages generated during a test of a software application to form event message statistics. In one embodiment, a measurement is computed to identify an amount or percentage of software application code tested during the test script. A code base of the software application may include a central repository having a substantial portion of the event messages that may be issued by the software application in response to error conditions. In one embodiment, these event messages are taken from the central repository and stored within a database. Other embodiments are described and claimed.

Abstract: A memory system permitting a number of alternative memory blocks to be made ready in order to extend the rewritable life and thereby contributing to enhanced reliability of information storage is to be provided. The memory system is provided with a nonvolatile memory having a plurality of data blocks in predetermined physical address units and a controller for controlling the nonvolatile memory in response to an access request from outside. Each of the data blocks has areas for holding a rewrite count and error check information regarding each data area. The controller, in a read operation on the nonvolatile memory, checks for any error in the area subject to the read according to error check information and, when there is any error, if the rewrite count is greater than a predetermined value, will replace the pertinent data block with another data block or if it is not greater, correct data in the data block pertaining to the error.