With the continuous expansion of virtualization technologies such as cloud computing and containers

the east-west traffic in cloud

data center and enterprise network is increasing rapidly. If the virtual network traffic is not collected

80% of the user’s network traffic will be in a "black box"

unable to secure the east-west vector in cloud platforms. In the information security technology: general requirements for classified protection of cyber security version 2

cloud computing security extensions

it is clearly required to detect and report traffic anomalies between virtual machines and between hosts and virtual machines. Based on the analysis of the existing east-west traffic detection solutions in this field

this paper studies and proposes a non-invasive traffic collection method which is deeply integrated with the cloud platform’s own architecture. This overall solution of attack detection has been improved by emerging technologies such as big data analysis and machine learning algorithm. In addition

it also explains the automatic adaptation of the integration process with the cloud platform business.