A method and a device for processing a flow event

Abstract

The invention relates to a method and a device for processing a flow event, which are used for avoiding overflow and omission of the flow event. The method comprises the following steps of: designinga double-layer cache pool of a stream queue and a database; after a flow event is captured, firstly determining a flow queue corresponding to an event type of a flow event; if no free space exists inthe stream queue, storing the flow event into a corresponding storage space of the flow queue in a database, reading the flow event into the flow queue to wait for processing when it is determined that a preset condition is met, and meanwhile, adaptively adjusting the queue length of the flow queue according to the use condition of the flow queue. Therefore, when a high-concurrency flow event arrives, the flow event can be rapidly received and stored, loss and omission of the flow event caused by overflow of the flow queue are prevented, the processing performance and the processing efficiencyof the system are ensured, meanwhile, the size of a flow event storage space can be flexibly controlled, and resource waste is reduced.

Description

technical field [0001] The invention relates to computer technology, in particular to a method and device for processing stream events. Background technique [0002] At present, the concurrent processing methods of real-time streaming events can be divided into network processing methods based on web architecture and client processing methods based on operating system according to the system architecture. [0003] The network processing method based on the web architecture is relatively mature, and usually obtains information related to flow events by monitoring network ports and processes them. However, the network processing method technology based on the web architecture is difficult to deal with real-time streaming events of big data. Taking sleep monitoring data as an example, the reasons are as follows: [0004] The real-time display of big data requires sufficient data storage space. Even if the sleep monitoring data is stored in time slices, the size of each segment...

AI Technical Summary

Problems solved by technology

[0015] However, the client-side processing method based on the operating system also has the following defects: the size of the message queue is fixed, and it is easy to cause overflow and omission of flow events

The length of a single queue is generally limited by memory

[0017] When storing high-concurrency flow events, it is easy for the system queue to have insufficient storage space when the flow events are intensively captured, resulting in the overflow of the flow events captured later and being discarded by the system, eventually resulting in the failure of system events to respond

[0018] Correspondingly, when the system events are calm, it is easy to cause empty queues and occupy memory space, resulting in waste of resources

Images