A cluster data processing method, system, device and medium

Abstract

The invention discloses a cluster data processing method, comprising the following steps: receiving parameters corresponding to data to be written; screening out several parameters from the parameters and converting the screened parameters into data in a key-value pair structure; storing the data of the key-value pair structure in the first storage space; in response to receiving a broadcast of data merging sent by other nodes, reading the data in the first storage space on the other nodes; storing the read data in the first storage space; The data is merged into the local first storage space. The invention also discloses a system, a computer device and a readable storage medium. The solution proposed by the present invention can solve the problem of data loss in a cluster in a high availability mode, and reduce maintenance costs and technical risks.

Description

technical field [0001] The present invention relates to the field of clusters, in particular to a cluster data processing method, system, device and storage medium. Background technique [0002] The artificial intelligence platform is based on the InfluxDB database to store monitoring data. In business scenarios, it is necessary to ensure that data is not lost and services can run uninterrupted. However, the non-enterprise version of InfluxDB does not support high availability; based on cost control and business scenario requirements, existing technologies can The scheme of writing multiple data at the same time is adopted to ensure the redundant backup of data, and the available services of InfluxDB are controlled and switched through the business layer code. [0003] However, according to InfluxDB's own architecture and actual business data analysis, it is found that the data in META is most likely to be lost in the user's actual scenario, and the loss will lead to abnorma...

AI Technical Summary

Problems solved by technology

[0002] The artificial intelligence platform stores monitoring data based on the InfluxDB database. In business scenarios, it is necessary to ensure that data is not lost and services run uninterrupted. However, the non-enterprise version of InfluxDB does not support high availability; based on cost control and business scenario requirements, existing technologies can Use the scheme of writing multiple data at the same time to ensure the redundant backup of data, and switch the available services of InfluxDB through the business layer code control

[0003] However, according to the analysis of InfluxDB's own architecture and actual business data, it is found that the data in META is most likely to be lost in the actual scene of the user, and the loss will lead to abnormal business services (the InfluxDB data in each node is chaotic, and the business data will display exceptions such as negative numbers)

Images