Fast fault recovery method and system for multiprocessor signal processing equipment

Abstract

The invention provides a multiprocessor signal processing equipment rapid fault recovery method and a system thereof. The multiprocessor signal processing equipment rapid fault recovery method comprises the following steps: in response to the occurrence of a fault of a first processor, storing a communication middleware state and a fault recovery mark of the first processor in a nonvolatile memory; protecting the code segment of the system firmware operated by the first processor in the memory, so that the first processor is subjected to soft reset; after the first processor is reset and restarted, running the boot firmware, accessing the nonvolatile memory by the boot firmware, acquiring an entry address of the code segment in the memory by the boot firmware in response to a failure recovery mark stored in the nonvolatile memory, and executing the system firmware through an instruction executed from the entry address; enabling the first processor to execute the system firmware to obtain the stored state of the communication middleware so that the first processor can be added into the communication group to which the first processor belongs when the first processor breaks down again; enabling the first processor to execute the system firmware to perform a task processing cycle.

Description

technical field [0001] The present application relates to reliability and fault tolerance, in particular, to a fast fault recovery method and system thereof after a node in a multiprocessor signal processing device fails. Background technique [0002] High-performance signal processing equipment usually adopts an architecture based on multi-core and multi-processor cooperative computing, and the computing tasks for implementing signal processing are shared among multiple processors or processor cores. In addition to computing, multiple processor cores or processors communicate frequently to distribute and / or collect computing tasks, and to synchronize the progress of computing tasks. Run software on the processor core or processor to control the signal processing process. The running software is large in scale, complex in process, and implements a variety of complex algorithms. It also includes communication middleware that manages communication between processors or process...

AI Technical Summary

Problems solved by technology

However, it takes a long time to restart the processor. In a multi-processor system, even if a single processor resumes to the point of failure and continues to run after restarting, all nodes (processor or processing) of the entire system will not be restarted due to other processors. The operating state of the core) is difficult to synchronize

Images