The Agenda

- What Is Layer 3 Switching?

- What is the Difference Between Layer 2 Switching, Layer 3 Switching, and Routing?

What Is Layer 3 Switching?

Recently, the industry has been bombarded with terminology such as Layer 3 switching, Layer 4 switching, multilayer switching, routing switches, switching routers, and gigabit routers. This “techno-jargon” can be confusing to customers and resellers alike.
For purposes of this discussion, all these terms essentially represent the same function, and, as such, the term Layer 3 switching is used to represent them all.
While the performance aspect of Layer 3 switching makes most of the headlines, higher performance in switching packets does not, by itself, promise that all problems are solved in a network. There must be a recognition that application design, mix of network protocols, placement of servers, placement of networking devices, management, as well as the implementation of end-to-end intelligent network services are at least as important—maybe more so—than simply adding more bandwidth and switching capability to the network.

Why Do We Need Layer 3 Switching?

So, why do we need Layer 3 switching? Enterprise networks face unprecedented challenges today. Desktop computing power has tripled in the past two years and shows no sign of leveling off. The proliferation of network-dependent intranet and multimedia applications has increased traffic volumes in many campus networks by an order of magnitude over the past several years. Network managers have responded to this need to move data at greater speeds by moving more desktops to switched 10/100 Mbps and deploying LAN switching at unprecedented levels, both in the data center and in the wiring closets to scale their end-to-end bandwidth. To effectively utilize the increased capacity, they must scale their Layer 3 performance to handle changing traffic patterns. Conventional wisdom that 80 percent of the traffic stays local to the subnet and 20 percent or less traverses across subnets no longer holds. More than half of the traffic volume travels across subnet boundaries. Two factors contribute to these changing traffic patterns.
With Web-based computing, a PC can be both a subscriber and a publisher of information. As a result, information can now come from anywhere in the network, creating massive amounts of traffic that must travel across subnet boundaries. Users hop transparently between servers across the entire enterprise by using hyperlinks, without the need to know where the data is located.
The second factor leading to the loss of locality is the move toward server consolidation. Enterprises are deploying centralized server farms because of the reduced cost of ownership and ease of management. All traffic from the client subnets to these servers must travel across the campus backbone, exacerbating performance problems.
Because of the rising levels of anywhere-to-everywhere communication, Layer 3 switching that can scale with increasing link speeds has become an imperative. Layer 3 switching is required to meet the demands of both client/server and peer-to-peer traffic on the intranet.

What Is Layer 2 Switching?

What is the difference between a Layer 2 switch, a Layer 3 switch, and a router?

A Layer 2 switch is essentially a multiport bridge. Switching and filtering are based on the Layer 2 MAC addresses, and, as such, a Layer 2 switch is completely transparent to network protocols and users’ applications.
Layer 2 switching is the number one choice for providing plug-and-play performance.

What Is Routing?

In contrast to Layer 3 switches, routers make Layer 3 routing decisions by implementing complex routing algorithms and data structures in software. Keep in mind this has little to do with the forwarding aspects of routing.
Routing has two basic functions, path determination, using a variety of metrics, and forwarding packets from one network to another.
The path determination function enables a router to evaluate the available paths to a destination and to establish the preferred handling of a packet.
Data can take different paths to get from a source to a destination. At Layer 3, routers really help determine which path. The network administrator configures the router enabling it to make an intelligent decision as to where the router should send information through the cloud.
The network layer sends packets from source network to destination network.
After the router determines which path to use, it can proceed with switching the packet: taking the packet it accepted on one interface and forwarding it to another interface or port that reflects the best path to the packet’s destination.

The Agenda

- What Is Layer 3 Switching?

- What is the Difference Between Layer 2 Switching, Layer 3 Switching, and Routing?

What Is Layer 3 Switching?

Recently, the industry has been bombarded with terminology such as Layer 3 switching, Layer 4 switching, multilayer switching, routing switches, switching routers, and gigabit routers. This “techno-jargon” can be confusing to customers and resellers alike.
For purposes of this discussion, all these terms essentially represent the same function, and, as such, the term Layer 3 switching is used to represent them all.
While the performance aspect of Layer 3 switching makes most of the headlines, higher performance in switching packets does not, by itself, promise that all problems are solved in a network. There must be a recognition that application design, mix of network protocols, placement of servers, placement of networking devices, management, as well as the implementation of end-to-end intelligent network services are at least as important—maybe more so—than simply adding more bandwidth and switching capability to the network.

Why Do We Need Layer 3 Switching?

So, why do we need Layer 3 switching? Enterprise networks face unprecedented challenges today. Desktop computing power has tripled in the past two years and shows no sign of leveling off. The proliferation of network-dependent intranet and multimedia applications has increased traffic volumes in many campus networks by an order of magnitude over the past several years. Network managers have responded to this need to move data at greater speeds by moving more desktops to switched 10/100 Mbps and deploying LAN switching at unprecedented levels, both in the data center and in the wiring closets to scale their end-to-end bandwidth. To effectively utilize the increased capacity, they must scale their Layer 3 performance to handle changing traffic patterns. Conventional wisdom that 80 percent of the traffic stays local to the subnet and 20 percent or less traverses across subnets no longer holds. More than half of the traffic volume travels across subnet boundaries. Two factors contribute to these changing traffic patterns.
With Web-based computing, a PC can be both a subscriber and a publisher of information. As a result, information can now come from anywhere in the network, creating massive amounts of traffic that must travel across subnet boundaries. Users hop transparently between servers across the entire enterprise by using hyperlinks, without the need to know where the data is located.
The second factor leading to the loss of locality is the move toward server consolidation. Enterprises are deploying centralized server farms because of the reduced cost of ownership and ease of management. All traffic from the client subnets to these servers must travel across the campus backbone, exacerbating performance problems.
Because of the rising levels of anywhere-to-everywhere communication, Layer 3 switching that can scale with increasing link speeds has become an imperative. Layer 3 switching is required to meet the demands of both client/server and peer-to-peer traffic on the intranet.

What Is Layer 2 Switching?

What is the difference between a Layer 2 switch, a Layer 3 switch, and a router?

A Layer 2 switch is essentially a multiport bridge. Switching and filtering are based on the Layer 2 MAC addresses, and, as such, a Layer 2 switch is completely transparent to network protocols and users’ applications.
Layer 2 switching is the number one choice for providing plug-and-play performance.

What Is Routing?

In contrast to Layer 3 switches, routers make Layer 3 routing decisions by implementing complex routing algorithms and data structures in software. Keep in mind this has little to do with the forwarding aspects of routing.
Routing has two basic functions, path determination, using a variety of metrics, and forwarding packets from one network to another.
The path determination function enables a router to evaluate the available paths to a destination and to establish the preferred handling of a packet.
Data can take different paths to get from a source to a destination. At Layer 3, routers really help determine which path. The network administrator configures the router enabling it to make an intelligent decision as to where the router should send information through the cloud.
The network layer sends packets from source network to destination network.
After the router determines which path to use, it can proceed with switching the packet: taking the packet it accepted on one interface and forwarding it to another interface or port that reflects the best path to the packet’s destination.

Packet Manipulation at Layer 3

How does Layer 3 switching differ from Layer 2 switching? Layer 3 switching requires rewriting the packet. This implies decrementing the TTL field, modifying the MAC addresses, changing the VLAN-ID and recomputing the FCS. Doing all these actions at wire speed is difficult which is why an ASIC is necessary.
True Layer 3 switching has all the advantage of routing, therefore it is rich in feature and performance.
Layer 2 switching, on the contrary, does not require packet rewriting. Without packet rewriting, no matter how you call it (e.g. virtual routing) it is NOT routing.

What Is Layer 3 Switching?

Layer 3 switching is hardware-based routing. The packet forwarding is handled by specialized hardware, usually ASICs.
A Layer 3 switch can make switching and filtering decisions on both Layer 2 and Layer 3 addresses and can dynamically decide whether to route or switch incoming traffic.
Multilayer switching combines the ease of use of Layer 2 switching with the stability and security of Layer 3 routing.
To make Layer 3 switching decisions, routing table information must be assembled and exchanged between routing entities.
Route calculation is performed by one or more route processors that reside in routers or other devices. These route processors periodically distribute their routing tables to multilayer LAN switches to allow them to make very fast switching decisions.
Layer 3 switching is the favorite for highly scalable, resilient networking.

A Layer 3 Switch Has Two Distinct Components

ASICs:

- High-performance, hardware-based Layer 3 switching and services with consistent low latency

Routing software:

- Routing protocols to provide scalability

- Backbone redundancy
- Dynamic load balancing and fast convergence in the backbone
- Reachability information

- Multiprotocol support for the campus

What Is the Difference Between Layer 3 Switching and Routing?

Layer 3 switches tend to have packet switching throughputs in the millions of packets per second (pps), while traditional general-purpose routers have evolved from the 100,000 pps range to over a million pps. Aggregate performance is one of the key differences between Layer 3 switches and traditional routers.
Traditional routers still offer key features used typically in WAN environments. However, many of those features, such as multicast routing, multiprotocol routing, IBM feature sets, routing protocol stability, are still key for Layer 3 switches/campus routers.

A Layer 3 or a Layer 2 Switch?— Scalability Advantages

Let’s look more closely at when a customer might choose a Layer 3 switch over a traditional Layer 2 switch. Layer 3 switches offer considerable advantages depending on the customer’s requirements.

Scalability— For customers with large networks that need increased performance to handle the changing traffic patterns of today’s new applications, Layer 3 switches offer increased scalability. Clearly a network of hubs does not scale. While bridges helped, they were not sufficient to handle networks of many thousands of users and devices. Routers were the solution as they kept broadcasts local to a segment. Layer 3 switches avoid the problems associated with flat bridged or switched designs using traditional routing mechanisms allowing customers to scale their network infrastructure.
Layer 3 switches also utilize routing protocols thus avoiding the slow convergence problem of Spanning Tree Protocol and lack of load-balancing across multiple paths.

Advanced services— Layer 3 switches also offer the benefit of broader intelligent network services. These services permit applications to run on the network as well as enable the creation of a cost-effective, operational environment to support day-to-day operations and management of the enterprise intranet.

Other Advantages

Other advantages include:

Security—Layer 3 switches provide enhanced security functions to protect corporate information while allowing appropriate access. Access control lists are supported by Layer 3 switches with no performance degradation. Layer 3 switching is able to enforce the multiple levels of security traditionally only found on routers on every packet of the flow at wire speed.

Management—Networks that use a multilayer model are by nature hierarchical. This type of infrastructure is easier to manage as problems are more easily isolated.

Redundancy/resiliency—Some Layer 3 switches offer significant redundancy and resiliency options not available with Layer 2 switches. Default gateway redundancy is provided by HSRP that enables Cisco switches to transparently switch over to the hot standby backup router instantly when the primary router goes off line, eliminating a single point of failure in the network. UplinkFast provides alternative paths when a primary link fails. Load balancing is achieved by intelligent Layer 3 routing protocols.

While there are obvious advantages to a Layer 3 switch over a Layer 2 switch, other factors needed to be considered as well. Layer 3 switches are more expensive than Layer 2 switches and are more complex. Depending on the size of a customer’s network, the cost and complexity may not justify a Layer 3 switch. However, for customers with larger networks in need of enhanced scalability, Layer 3 switches will actually simplify network infrastructure.

Not All Layer 3 Switches Are Created Equal

At its most basic, Layer 3 packet switching or forwarding is common across all vendors platforms, with perhaps exceptions in their multicast or DHCP services behavior.
The more scalable, flexible, and adaptable Layer 3 switches also offer a variety of routing protocols and services for topology discovery, load balancing, and resiliency. Buying a Layer 3 switch without the richness and depth of routing protocols is somewhat akin to a driverless car. The car can certainly travel very fast in the direction that it is pointed, but the intelligence lies in the driver, who needs to make all the decisions about where it should go and when to stop and turn. The more flexible and resilient these capabilities, the better reliability and adaptability the switch offers.
Finally, there are services. All the queuing, filtering, classification, multiprotocol, route summarization and redistribution functions, plus additional debugging, statistics gathering and event logging services is what lets network managers deploy solutions that rise to the future challenges of mobility, multiservice, multimedia, and service level agreements for business critical applications.

---

- Summary -

- Layer 3 switching is ASIC-based routing

- Traditional routers are better for WAN aggregation

- Layer 3 switches are more appropriate for scaling Layer 3 performance

- Layer 2 switches are more appropriate when the additional cost and complexity are not warranted