Network Security Internet Technology Development Database Servers Mobile Phone Android Software Apple Software Computer Software News IT Information

In addition to Weibo, there is also WeChat

Please pay attention

WeChat public account

Shulou

What are the technologies of layer 2 network in data center?

2025-03-28 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > Network Security >

Share

Shulou(Shulou.com)06/01 Report--

Data center network is mainly divided into computing network, storage network and Internet network, and computing network is the core network of data center network.

Most of the traditional data center networks adopt the second and third layer architecture similar to the campus network.

Advantages:

Easy to deploy, in line with the business characteristics of data center partition and sub-module

However, with the evolution of data centers from traditional data centers to virtualized data centers and cloud data centers, this two-tier and three-tier hybrid network architecture appears to be inadequate. Because the current data center adopts virtualization technology, the upper virtual machines migrate dynamically with the change of demand, and the key of dynamic migration is to ensure that the business is not interrupted, the IP address remains the same, and the running state remains the same, so the virtual machine migration can only be carried out in the two-tier domain, not praise the two-tier domain migration. In order to improve the redundancy of layer 2 network, equipment redundancy and link redundancy are adopted. In this way, it is necessary to use STP technology to prevent loops. Due to the performance limitations of STP, the layer 2 network broken by STP is usually no more than 50 network devices. In this way, the migration scope of VM is limited, and the application is greatly limited. in order to realize the large-scale cross-regional migration of VM, all the servers that may be involved are required to be connected to a larger layer 2 network domain, which is the big layer 2 network.

It is impossible to use STP technology to realize layer 2 network, so we need to adopt new technology. The current layer 2 network technology is mainly as follows:

1. Network equipment virtualization technology

With the combination of network equipment virtualization and link aggregation technology, the original multi-device multi-link structure can be transformed into a logical single-device single-link architecture, which eliminates the emergence of loops, so it is no longer limited by the broken-down protocol, thus realizing the large layer 2 network.

The virtualization technologies of network devices mainly include:

Frame device: VSS,IRF2,CSS

Istack technology for cassette equipment

Frame and box: SVF technology

Advantages: simple network logic, simple management and maintenance

Disadvantages: the scale of the network is relatively small compared with other technologies, and the network equipment virtualization technology is private to each manufacturer, so it can only use devices from the same manufacturer for networking, usually using small and medium-sized POD layer 2 networks.

2. TRILL/SPB and other routed layer 2 forwarding technologies.

The starting point is also to solve the loop problem, which is not to block the loop like STP, nor to put an end to the loop like virtualization, but to learn from the logic breaking mode of the three-layer network.

Layer 3 networks also have loops, so why not block redundant links? because layer 3 networks can rely on routing protocols to collect, synchronize and update network topologies, so that every network node can achieve the best forwarding path. therefore, even if there is a loop, there will not be a forwarding loop, so as to achieve logical breakage, the routing layer 2 forwarding technology is borrowed from this idea. The forwarding mechanism of layer 3 routing is introduced into layer 2 to solve the loop problem, so as to avoid the traditional layer 2 broken protocols such as STP, and then realize layer 2.

The frame structure of TRILL:

TRILL can be collected, synchronized and updated through the IS-IS routing protocol

When the virtual machine is migrated in the TRILL network, the forwarding table on each switch can be updated automatically through IS-IS, so the IP address of the virtual machine can be kept unchanged and the dynamic migration can be realized. The large layer 2 network constructed by TRILL is larger, and TRILL is the standard protocol of IETF, so it is easy to realize the interconnection between various vendors, which is suitable for building large POD or layer 2 network of the whole data center.

However, TRILL is a new technology, and the deployment of TRILL requires new hardware and software equipment, which may require greater investment costs.

Network equipment virtualization and routed layer 2 forwarding technologies are large layer 2 network technologies introduced by network manufacturers, while the third technology is VXLAN and NVGRE Overly technologies proposed by IT manufacturers.

The so-called Overly technology uses tunnel encapsulation technology to transparently transmit the original data packet (layer 2 message) sent by the source host in the existing network after secondary encapsulation, then unencapsulate it when it arrives at the destination, and then forward it to the destination host, so as to realize the layer 2 communication between hosts. It is equivalent to a large layer 2 network superimposed on the bearer network, so called Overly,Overly technology is equivalent to virtualizing the bearer network into a large layer 2 switch. The migration of VM is like running from one port of the switch to another port, and the state remains the same. The Overly scheme is put forward by IT manufacturers, such as VXLAN technology proposed by VMware and NVGRE technology proposed by Microsoft. The reason why IT manufacturers put forward such a scheme is to get rid of the technical dependence on network vendors and independently implement a large layer 2 network, because in the Overly scheme, the bearer network only needs to meet the basic switching and forwarding, and the encapsulation and unencapsulation of the original message can be completed by the virtual switch. No need to rely on network devices

Vxlan is packaged with MAC in UDP.

1. The layer 2 message sent by the virtual machine is encapsulated by VXLAN in the virtual machine switch.

Overly technology does not depend on the bearer network, and can make full use of the existing basic network to realize the large layer 2 network, and has great advantages in supporting SDN and multi-tenant. it is the hottest technology in the current large layer 2 network, which can realize the layer 2 network of the whole data center or even cross the data center, but there are two control layers in Overly technology: Overly network and Underlay bearer network. The management and maintenance and fault location are relatively complex, and the operation and maintenance work is also relatively large.

EVN technology: it is not to build a large layer 2 network of a single data center, but to realize a large layer 2 network interconnection across data centers. Traditional data center layer 2 interconnection usually uses VPLS (VirtualPrivateLanService) or enhanced VPLS OVER GRE, but these technologies have the problems of complex configuration, low bandwidth utilization, high network deployment cost and high consumption of network resources. EVN is a layer 2 interconnection × × technology based on VXLAN tunnel. It can be considered as an extension of VXLAN. EVN technology transmits MAC address information between layer 2 networks by using MP-BGP protocol, and realizes layer 2 packet forwarding by generating MAC address table entries. EVN supports automatic establishment of VXLAN tunnel, multiple regression access load sharing, BGP route reflector and ARP cache pickup and other functions, effectively solving the problems of layer 2 interconnection technology such as VPLS. It is one of the ideal solutions for second-layer interconnection in data centers.

Summary:

Network device virtualization is suitable for small and medium-sized POD-level Fabric

Routed layer 2 forwarding technology is suitable for large POD level Fabric.

Overly technology is suitable for DC-level Fabric

EVN technology is suitable for DCIFabric.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.

Views: 235

*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.

Share To

Network Security

Wechat

© 2024 shulou.com SLNews company. All rights reserved.

12
Report