Application of Network Virtualization Technology in Traditional Network

Application of Network Virtualization Technology in Traditional Network

In recent years, with the explosive growth of network traffic, customers have higher and higher requirements for the stability, reliability, and flexibility of the network architecture. The original fragile traditional network architecture is increasingly difficult to meet the actual needs. Thus, it is urgent to upgrade the traditional network architecture. Based on the shortcomings of the traditional network architecture, this paper adopts a solution based on IRF network virtualization technology for transformation. This technology has excellent characteristics such as high reliability and easy expansion. It is widely used in the upgrading and transformation of traditional networks.

Basic situation of traditional network architecture

Traditional network architectures are usually a star topology. Taking part of the network topology of a campus network as an example, the access layer network consists of four H3C S3600 switches. The aggregation layer network consists of two H3C S5560 switches. The overall network runs the MSTP protocol to eliminate Layer 2 loops, and implements load balancing of different VLAN traffic based on MSTI (Multiple Spanning Tree Instances).

In addition, in order to avoid the occurrence of single point failures in the network system, the VRRP protocol is also configured for gateway devices. This is a redundant backup. Once an aggregation switch fails, all services will be switched to another aggregation switch. Thus, ensuring high reliability of the overall network architecture and load sharing of network traffic.

Analysis of traditional network architecture problems

Since the construction of a campus network, its core equipment and services have been in continuous operation for more than ten years. However, with the expansion of campus network business and scale year by year, the traditional network architecture has gradually exposed some new problems in network operation and maintenance.

  1. Active/standby switching and fault recovery are slow

At present, two aggregation switches use MSTP+VRRP technology to form a dual-machine hot backup system. But, the dual-protocol coordination mechanism is too complicated. Once a fault occurs, the master-standby switchover and fault recovery will take a certain amount of time, usually in seconds.

  1. The scale of the network expands, making it difficult to locate faults

During the operation of the campus network for many years in the project example, the routing information and security policies configured on the two aggregation switches may be inconsistent due to historical reasons. This will make it difficult to accurately locate the fault in the network. Thus, increase the risk of operation and maintenance.

  1. The performance of the aggregation switch is insufficient to meet the needs of the network

In recent years, with the popularity of short online videos and the application of high-definition distance learning systems, campus network data traffic has increased significantly. The problem of insufficient performance of aggregation switches has become increasingly serious. Thus, the network experience of teachers and students has been affected to a certain extent.

Improvement scheme of applying IRF network virtualization technology

  1. Overview of IRF network virtualization technology

IRF (Intelligent Resilient Framework) technology is a network virtualization technology independently developed by H3C. Its core idea is to connect multiple network devices of the same model and software version that support IRF technology through the IRF stacking interface. Then virtualize them into one network device after the necessary configuration. This technology is used to simplify the network topology. Allowing you to realize the collaborative work of multiple network devices in the IRF cluster. It also offers unified management and uninterrupted maintenance at the same time. This is because there are multiple network devices in the IRF cluster as mutual backup. It can also improve the reliability of the network system and overall performance.

  1. IRF network virtualization configuration deployment

In this paper, the HCL (H3C Cloud Lab) simulation software is used to simulate the IRF transformation of the campus network. HCL software is a network simulation software independently developed by H3C. This is used to make up for the lack of experimental conditions in reality. The configuration process and experimental results of the network experiments run on the simulation software are basically consistent with H3C's real network equipment. So, it is widely used in network engineering practice.

IRF virtualization configuration process

The overall process of IRF technology configuration is relatively complicated. Before configuring IRF technology, IRF cables and optical modules need to be connected in advance, and the priority and member number of each member device in the IRF cluster must be specified.

If all the links used for IRF configuration in the IRF cluster are interrupted, there will be two network devices with the same configuration in the entire network. This process is called IRF split. If some necessary detection measures are not taken, IRF split will lead to IP address, Router_ID conflict, route flapping and other network failures in the live network. The solution is to configure the BFD MAD detection function in the IRF cluster. Once the IRF splits, the IRF system will automatically close all the ports in the Slave device within milliseconds. This avoids further expansion of the fault domain in the network. Thus, maintains the network to the maximum extent.

Uninterrupted operation of business.

Because the IRF cluster forms at this time, it only needs configuration on any switch in the IRF cluster. All configuration steps will automatically synchronize with other switches in the IRF cluster.

After the configuration is complete, you can check the general information of the BFD session. At this time, the IRF cluster is running normally. Only the MAD IP address configured on Huiju_01 takes effect, but the MAD IP address configured on Huiju_02 does not take effect. So, the BFD session status is Down, which is normal. Once the IRF splits, the BFD session will be in the Up state instantly. At this time, the MAD detection mechanism will take effect. The IRF system will automatically close all ports on the Huiju_02 device to isolate this device from the live network. Finally, the BFD session status will change to the Down state.


Reasonable use of network virtualization technology to optimize and transform the traditional network architecture can make the network architecture more reliable. It can also make later operation and maintenance more convenient. However, not all network devices support virtualization. Thus, some network devices need to purchase dedicated connection cables and optical transceivers when configuring virtualization. However, in the future network construction and transformation, network virtualization technology will become an important force to promote the transformation of traditional network architecture.

Blog Categories

Recent Posts

Search Site
© 2012-2024    Contact   -   Privacy
magnifier linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram