Inter-node bus accessing priority deploying method based on multiple servers

Abstract

The invention provides an inter-node bus accessing priority deploying method based on multiple servers. Firstly, according to the server node bus communication requirements, grouping integration is carried out on key physical signals participating in bus communication logic through a programmable logic controller; meanwhile, transmission starting monitoring on high speed data of all parallel nodes is carried out; the actual value cooperation requirements such as time sequences and triggering priority requirements of the bus signals are achieved in a DSP in a built-in mode. The principle is that in the bus idle stage, buses are sequentially switched and connected among the nodes according to a natural sequence in order to guarantee the data response instantaneity; in the bus busy stage, besides consideration to the natural sequence, judgment on the waiting time is further added, and the nodes with the long waiting time preferentially have access in bus communication. Thus, the time sequences and the triggering requirements of the bus signals are achieved through DSP software languages, bus communication signals collected by the programmable logic controller are transmitted to a DSP data processing center, a DSP judges the actual values such as the time sequences and the triggering requirements of each participant signal according to the preset timing sequence requirements, and whether the bus switching requirements are met or not is further judged, and resource balanced configuration is achieved.

Description

technical field [0001] The present invention relates to the field of computer communication, and specifically uses a priority allocation method based on bus access between multi-server nodes to solve the problem that the server system cannot communicate according to the current node-to-point communication data transmission process between the current server system nodes. , Real-time automatic deployment of computing resources when the bus is idle. The method ensures the real-time performance, reliability and stability of server node communication. Background technique [0002] Today's server systems often require multiple server nodes to cooperate to complete computing tasks, and the data and management channels between multiple server nodes are usually mounted and communicated equally on the bus. The server needs long-term uninterrupted operation and stability, and the reliability requirements of node communication are constantly increasing. In order to ensure the stable d...

AI Technical Summary

Problems solved by technology

[0005] The present invention is based on the timing theory of bus communication, and uses a priority allocation method based on bus access between multi-server nodes to solve the problem that the server system cannot Real-time automatic allocation of computing resources according to node and bus idle conditions

Images