Method and device for creating network card queue under RDMA (Remote Direct Memory Access) network

Abstract

The embodiment of the invention provides a network card queue creating method and device under an RDMA network, and the method comprises the steps: receiving a queue creating request which carries queue address information and queue port information; creating a network card queue pool and at least two network card queues corresponding to the queue creation request according to the queue address information and the queue port information; and associating the at least two network card queues with the network card queue pool to serve as the network card queues of the network card queue pool.

Description

Technical field [0001] Embodiments of this specification relate to data communication technology, and in particular, to network card queue creation methods and devices under RDMA networks. Background [0002] With the development of the Internet, more and more enterprises support employees online office, online conferences have become the mainstream of the future. In this case, global users have further surprising the demand for cloud computing, which also prompted major Infrastructure suppliers continue to expand data center size and quantity to meet the current and future needs. The current data center uses ordinary commercial server and network equipment. The single data center is very large, and the number of devices is tens of thousands to hundreds of thousands. The data center has increased with more network equipment failure (life expires, hardware BUG, burst power failure, conventional fault probability, natural disaster, etc.), which typically leads to usability jitter, ...

AI Technical Summary

Problems solved by technology

The current data center fully uses common commercial servers and network equipment. The scale of a single data center is very large. bugs, sudden power outages, regular failure probability, natural disasters, etc.), these network failures usually cause the availability of cloud services running on the data center to fluctuate, the service is temporarily offline, and even the service is unavailable for a long time.

At present, when a fault occurs, the operation and maintenance system will alarm, and the operation and maintenance engineer will manually isolate the faulty device and switch the faulty traffic to the available device. This process usually takes minutes, which is quite long for service recovery.

Images