Unilaterally-controlled, time-insensitive, data-link recovery apparatus and method
First Claim
1. An apparatus comprising:
- an actuator comprising a signal generator effective to provide a task initiation signal;
a communication link operably connecting the actuator to a task executing device;
the task executing device adapted to receive the task initiation signal, to execute a task in response thereto, and to return a failure signal in the event of a failure of the task executing device to complete the task;
a logical device operably associated with the actuator, the logical device being adapted to receive the failure signal, automatically select for the actuator a preferred recovery response to the failure signal, and to communicate an instruction signal to the actuator effective to execute the preferred recovery response independent of the time elapsed after the failure; and
a memory device operably connected to the actuator, for storing data structures, the data structures comprising;
a first data set corresponding to execution of the task prior to the failure signal,a second data set corresponding to the task executed after the recovery response, andeach of the first and second data sets being partitioned at an atomic level, independent from redundancies, checkpoints, rollbacks, conversations, and backwards recoveries.
16 Assignments
0 Petitions
Accused Products
Abstract
An apparatus and method for unilaterally-controlled, time-insensitive recovery of disconnected data links provides for user-transparent recovery, without requiring inputs from a user or other intervention. Data and executable operations may be preserved, as re-booting is not required. Optionally, opportunity for user intervention may be provided, at a user'"'"'s request. Recovery may occur automatically by use of expert knowledge applied to reported system errors from the disconnected communication link. An apparatus may rely on a logical device within an actuator or machine of mechanical, pneumatic, hydraulic, electrical or electronic character. A task to be reported to an actuator as successful or failed, may be internal to the actuator, internal to the logical device, or external to both. Data structures stored in a memory device, may implement the invention readily from within an application running on a processor in a computer. Information (data, operational steps) may be preserved at an atomic level of granularity.
61 Citations
29 Claims
-
1. An apparatus comprising:
-
an actuator comprising a signal generator effective to provide a task initiation signal; a communication link operably connecting the actuator to a task executing device; the task executing device adapted to receive the task initiation signal, to execute a task in response thereto, and to return a failure signal in the event of a failure of the task executing device to complete the task; a logical device operably associated with the actuator, the logical device being adapted to receive the failure signal, automatically select for the actuator a preferred recovery response to the failure signal, and to communicate an instruction signal to the actuator effective to execute the preferred recovery response independent of the time elapsed after the failure; and a memory device operably connected to the actuator, for storing data structures, the data structures comprising; a first data set corresponding to execution of the task prior to the failure signal, a second data set corresponding to the task executed after the recovery response, and each of the first and second data sets being partitioned at an atomic level, independent from redundancies, checkpoints, rollbacks, conversations, and backwards recoveries. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
-
-
17. An apparatus comprising:
-
a communication link adapted to communicate data; a processor operably connected to the communication link and programmed to run executables comprising an application executable hosted on the processor and effective to transfer control of the processor to a task, the task being effective to cause a failure signal to indicate a failure of the task to execute completely, and a recovery executable effective to recognize the failure signal, automatically select a recovery method, execute the selected recovery method, preserving a partial data set created by the task, partitioned at an atomic level, and reflecting the state of the task exactly prior to the failure, and cause completion of the task, beginning with executing a step corresponding to the failure independent of the time elapsed after the failure; a first storage device operably connected to the communication link to store the partial data set corresponding to execution of a first portion of the task prior to the failure, and to store, independent from redundancies, checkpoints, rollbacks, conversations, and backwards recoveries, and after execution of the recovery method, a completion data set corresponding to a second portion of the task executed after the failure. - View Dependent Claims (18, 19, 20, 21, 22)
-
-
23. An article comprising a memory device provided with blocks of memory adaptable to store executables and data, the memory device comprising:
-
a first block storing a task executable to be loaded to run on a processor operably connected to the memory device; a second block storing an application, the application having integrated therein a recovery executable to be loaded to run on the processor upon a failure of the task executing device to complete execution of a task; a third block persistently storing first data comprising each atomic unit of data generated by the task executing device, exclusively prior to the failure; a fourth block storing second data generated by the task executable, exclusively after the failure, the first and second data corresponding together comprising a complete data set indistinguishable from a normal data set generated by the task executable executing the task normally without the failure; and a recovery control data block comprising data containing an abort flag used by the recovery executable to recover the task after the failure and effective to disable all further attempts to execute recovery processes. - View Dependent Claims (24, 25)
-
-
26. A method of auto-recovery of a task operating on a computer, the method comprising:
-
providing a processor for operating executables, the processor being operably connected by a datalink to a memory device for storing executables and data; providing an application comprising a plurality of executables executable on the processor, the application also comprising an initiation executable effective to initiate a task executable and comprising a recovery executable effective to recover the task executable following a failure of the task executable to complete execution; operating the application; invoking the task executable, preserving first data generated during execution; determining that a failure of the task executable has occurred; invoking automatically the recovery executable to recover the task to a state existing at the time of the failure; completing the task, generating second data, mutually exclusive of the first data, during execution thereof following the failure, the first and second data forming a non-redundant, complete data set, corresponding to, and indistinguishable from, a normal data set generated by the task operating without failure. - View Dependent Claims (27, 28, 29)
-
Specification