Skip to main content

What is Difference Between Hardware and Software Switching

If I summarize the answer of this question in one line then I would say Hardware Switching is performed by ASICs and Software Switching is performed by CPUs. But let’s have some explanation of both to clearly understand the difference.
The term hardware-switching refers to the act of processing packets at any Layers 2 through 7, via specialized hardware components referred to as Application-Specific Integrated Circuits (ASIC). ASICs can generally reach throughput at wire speed without performance degradation for advanced features such as QoS marking, ACL processing, or IP rewriting.
Other terms used to describe hardware-switching are in-hardware, using ASICs, and hardware-based.
Multilayer switching (MLS) is another term commonly used to describe hardware-switching.  MLS describes the capability to route and switch frames at line-rate (the speed of all ports sending traffic at the same time, full-duplex, at the maximum speed of the interface) with advanced features such as Network Address Translation (NAT), QoS, access controls, and so on using ASICs.

Switching and routing traffic via hardware-switching is considerably faster than the traditional software-switching of frames via a CPU. Many ASICs, especially ASICs for Layer 3 routing, use specialized memory referred to as ternary content addressable memory (TCAM) along with packet-matching algorithms to achieve high performance, whereas CPUs simply use higher processing rates to achieve greater degrees of performance. Generally, ASICs can achieve higher performance and availability than CPUs.
In addition, ASICs scale easily in switching architecture, whereas CPUs do not. ASICs integrate not only on Supervisor Engines, but also on individual line modules of Catalyst switches to hardware-switch packets in a distributed manner.
ASICs do have memory limitations. For example, the Catalyst 6500 family of switches can accommodate ACLs with a larger number of entries compared to the Catalyst 3560E family of switches due to the larger ASIC memory on the Catalyst 6500 family of switches. Generally, the size of the ASIC memory is relative to the cost and application of the switch. Furthermore, ASICs do not support all the features of the traditional Cisco IOS. For instance, the Catalyst 6500 family of switches with a Supervisor Engine 720 and an MSFC3 (Multilayer Switch Feature Card) must software-switch all packets requiring Network Address Translation (NAT) without the use of specialized line modules. As products continue to evolve and memory becomes cheaper, ASICs gain additional memory and feature support.
Labels:

Comments

Post a Comment

Popular posts from this blog

How to import Putty Saved Connections to mRemoteNG

Just started using mRemoteNG and its being very cool to connect to different remote connection with different protocols e.g Window Remote Desktop, VNC to Linux, SSH, HTTP connection etc. from a single application. As new user I configured some remote desktop connection which was quite easy to figure out. But when I wanted to add SSH connections, it came in my mind to import all of the saved connections in the putty. But I couldn't figure it out how can it be done, though it was quite easy and here are the steps. Open your mRemoteNG Create a folder if you want segregation of multiple networks Create a new connection Enter the IP address of remote server under connection in Config pane Under the config pane, select protocol " SSH version 2 ".  Once you select protocol to SSH version 2 you are given option to import putty sessions, as shown in the snap below. In the above snap, I have imported CSR-AWS session from my saved sessions in Putty.
1 comment
Read more

BGP Local Preference Controlling Outbound Traffic in BGP

In BGP, Local Preference is used to control the outbound traffic path. It helps you decide which egress point (exit point) should be used when you have multiple connections to external networks, such as ISPs. Local Preference is an attribute that is local to your AS and is shared with all iBGP peers but not with eBGP neighbors. Higher Local Preference = More preferred outbound path. Example Scenario : You have two external links: ISP1 (via CE1) and ISP2 (via CE2). You want traffic to prefer ISP1 for all outbound traffic. Network Topology : CE1 (connected to ISP1): 10.0.1.1/30 CE2 (connected to ISP2): 10.0.2.1/30 iBGP Router (Internal) connected to both CE1 (10.0.1.2/30) and CE2 (10.0.2.2/30). Configuration on CE1 (Higher Local Preference) : Create a route map to set the local preference to 200 for routes learned from CE1: route-map SET_LOCAL_PREF permit 10 set local-preference 200 In the BGP configuration for CE1, apply this route map to the neighbor: router bgp 65001 ne...
Post a Comment
Read more

BGP MED: Managing Inbound Traffic with Multi-Exit Discriminator

The Multi-Exit Discriminator (MED) is used in BGP to control inbound traffic into your AS. It tells a neighboring AS which entry point into your network it should prefer when there are multiple links between your AS and the neighboring AS. The lower the MED value , the more preferred the path. MED is only honored between the same neighboring AS . Example Scenario : You are connected to ISP1 via two routers, CE1 and CE2 , and want to control which router ISP1 uses to send traffic into your AS. Network Topology : CE1 (connected to ISP1): 10.0.1.1/30 CE2 (connected to ISP1): 10.0.2.1/30 iBGP Router (Internal) connected to both CE1 (10.0.1.2/30) and CE2 (10.0.2.2/30). Configuration on CE1 (Lower MED, More Preferred) : Create a route map to set the MED to 50 for CE1: route-map SET_MED permit 10 set metric 50 Apply this route map to the neighbor in the BGP configuration for CE1: router bgp 65001 neighbor 10.0.1.1 remote-as 65000 neighbor 10.0.1.1 route-map SET_MED out Configuratio...
Post a Comment
Read more