Today, more and more enterprises are taking advantage of data center virtualization to reduce costs and simplify data flow, and to ensure that data can move seamlessly between data centers and business locations, network capacity requirements will increase significantly. To accommodate this trend, the industry is looking for ways to increase network capacity by connecting as many network devices as possible, such as MLAG and vPC, two link aggregation technologies that connect multiple Ethernet switches to increase network capacity. So do you know them? What is the difference between the two? This article will give a comprehensive introduction to this.
What is MLAG?
MLAG (Multi-Chassis Link Aggregation) is a non-standard protocol that enables Layer 2 multipathing from hosts for extra bandwidth or link resiliency, since it is a public protocol, it is customizable by each vendor The mechanism implementation supports MLAG. MLAG refers to the aggregation of two or more switches across devices to form an M-LAG dual-active system, which means that it allows two or more physical switches to present a set of parallel links as a single aggregated link, and Allows hosts to uplink to two switches for physical diversity while only having to manage one bonded interface. In turn, these two switches can use MLAG to connect to other switches and forward all data.
In MLAG links, Link Aggregation Control Protocol (LACP) is usually used for north-south negotiation between hosts and MLAG virtual switches or between MLAG virtual switches. East-west proprietary protocols are used between switches belonging to the same MLAG active-active system. As shown in Figure 1 below, multiple FS S5800-48F4S Ethernet switches are used to form a MLAG active-active system. Among them, the four leaf switches in the MLAG dual-active system establish redundant connections from the server to the switch, and the LAG group is formed by means of static link aggregation or negotiation based on the LACP protocol.
Figure 1: MLAG active-active system
Advantages of MLAG
Evenly distribute traffic to each switch by using LAG;
Increased bandwidth for north-south and east-west directions simply by bundling more links into the LAG group;
Provides stability through dual management and control planes;
Support upgrading one switch at a time without affecting other devices;
Port capacity expansion is simple and free – you can expand the system ports by creating another MLAG active-active system to another switch to add east-west transmission of another switch.
What are vPCs?
vPC, or Virtual Link Aggregation, is a Cisco Nexus series-specific technology that is difficult to configure on other types of switches. Of course, this is not absolute. In addition to Cisco, some suppliers can also provide switches that support vPC, such as Fast (FS) N series switches. So what exactly is vPC? vPC technology allows the physical link connected to two different Cisco switches to be treated as a logical port aggregation link, and then connected to other devices (such as switches, servers or any other network devices that support IEEE 802.3ad PortChannels).
vPC allows Layer 2 PortChannels (port channels) to be created across two switches, as shown in the figure below, vPC is used to create inter-switch links across two switches, while keeping the two control planes of these switches separate. After enabling the vPC function, you need to create a link for peer survival detection (that is, keepalive) to send heartbeats between devices. The vPC domain includes vPC devices, vPC peer survival detection links, vPC peer interconnection links, and all PortChannels in the vPC domain. Note: Only one vPC domain can be specified on each device.
Figure 2: vPC working topology diagram
Advantages of vPC
Allows a device to connect to two different upstream devices using PortChannel;
Eliminate spanning tree protocol blocking ports;
Provides a loop-free topology;
All available upstream bandwidth can be used;
When the link or device fails, it can achieve fast convergence, which is faster than spanning tree;
Provide link-level resiliency;
Help ensure high availability.
What is the difference between MLAG and VPC?
As can be seen from the above, both MLAG and vPC can be used to create a port group between two switches, and can provide Layer 2 multipath selection. In an MLAG active-active system or vPC domain, each switch is independently managed and configured to forward/route traffic without passing through the master switch. So what is the difference between MLAG and vPC?
Obviously, the biggest difference between them is the difficulty of implementation. MLAG is a public protocol, and almost every manufacturer can use a custom mechanism to support MLAG; while vPC is a protocol dedicated to Cisco Nexus, not all manufacturers can use this technology; so the installation of MLAG is relatively easier than vPC .
Network engineers who want to deploy vPC should study the vendor’s vPC design guide before building a vPC domain. When configuring vPC, they must ensure that they are using the same series of Cisco Nexus switches, such as on Nexus 7000 series or Nexus 5000 series switches Configure vPC, but you cannot configure vPC on Nexus 7000 series and Nexus 5000 series switches. Also, the vPC side switches must be running the same NX-OS version (except for non-disruptive upgrades). In addition, the vPC peer link must be at least two 10G Ethernet interfaces.
vPC is more advanced than MLAG. vPC can support Layer 2 and Layer 3 multi-paths. Users can increase network redundancy when there are multiple optional paths. At the same time, they can increase bandwidth, enable multiple parallel paths between multiple nodes, and load balance traffic. If you want to enable Layer 3 multipathing, you can use Active-Active Gateway Protocol (MAGP). Typically, vPC is used for data center (Nexus devices running NX-OS or ACI mode) switches, while MLAG can be used for most distributed applications or data center switches.
For data centers and cloud computing networks that require higher network bandwidth and reliability, both MLAG and vPC are ideal technologies and are easy to implement. MLAG takes full advantage of link aggregation and distributes it on a pair of data center switches to provide higher redundancy and realize a highly resilient multipath network.
vPC is more suitable for non-blocking path diversity, where a virtual machine may be located at any given location. Whether you choose MLAG or vPC, you need to find out whether your network equipment can support MLAG or vPC, and then decide which function to choose and configure based on the network architecture and your own specific business traffic requirements.
ThinkMo CCNA Dump exam information exchange group：
The complete EVE_NG file, free learning PDF and PPT that can be used directly, as well as video explaining the technical points are all here!