Multipoint fault treatment method and device

Abstract

The invention provides a multipoint fault processing method and device, relating to the technical field of computer. The multipoint fault processing method and device are used for preventing the repeated switching of the single board and the ping-pong treatment. The multipoint fault processing method comprises the following steps: pre-classifying the ports while the first switching board performs the multipoint fault treatment; performing the fault detection according to the classes of the ports respectively; calculating the local sum of the first state weights according to the state weights corresponding to the port classes in the normal state; and comparing the sum of the first state weights with the sum of the second state weights of the second switching board and selecting the switching board corresponding to the larger value as the main switching board. Because the state weights can reflect the whole working state of the switch board, therefore, when a multipoint fault occurs, the switching board with better working condition can be quickly selected as the main switching board, thus, the ping-pong switching of the single board or the system deadlock can be avoided, and the availability of the system can be maximized.

Description

technical field [0001] The invention relates to computer technology, in particular to a multi-point fault processing method and device. Background technique [0002] In an advanced telecommunications computing architecture (Advanced Telecommunications Computing Architecture, ATCA) based subrack system, usually a plurality of racks are arranged, and each rack is arranged with a plurality of racks (specifically, figure 1 As shown), the motherboard in a chassis is usually provided with multiple slots (for example, 14 slots, specifically as figure 2 As shown), each slot is inserted with a single board, and the logic function is defined for each slot inside the chassis, such as a global single board for global management, a switching board for data exchange, and so on. The chassis are interconnected through switch boards to ensure communication and data transmission between the chassis. [0003] Although the standard ATCA architecture defines the interconnection specifications...

AI Technical Summary

Problems solved by technology

[0004] In general system design, for a single point of failure in the system, a fault recovery solution will be designed, that is, to perform active-standby fast switching; A single point of failure is processed independently, and there may be conflicts in the processing results of multiple single point faults, resulting in repeated switching of devices, resulting in ping-pong effects or system deadlocks

Especially when two or more fault conditions occur at the same time, if they are all handled according to the single-point fault processing plan and process, the final result often leads to the system being unrecoverable

[0005] For example, after the 7-slot switch board detects a single point failure, it switches to the 8-slot switch board, and the 8-slot switch board also detects a single point fault, and then switches back to the 7-slot switch board, and the 7-slot switch board detects a single point fault again. After a single point of failure, switch to the 8-slot switch board, displaying that it is easy to guide the repeated switching of the device, resulting in a ping-pong effect

[0006] Under the existing technology, due to the complex combination conditions such as the location and time of multiple fault points, it is difficult to design a perfect automatic processing plan for multi-point faults; only rely on the maintenance and management personnel of the equipment according to the situation of multiple fault points Manual judgment for troubleshooting

Manual judgment and processing need to be judged and operated at the near end or at the maintenance station. If there is no on-duty personnel, it often takes a long time to eliminate the fault, which greatly affects the normal working time and fault recovery time indicators of the system equipment.

Images