Can someone tell what is the differece between these terms when you look at various switch performane criteria - forwarding bandwidth, wire speed, total bandwidth, backplane bandwidth etc. I am not sure about relation b/w these terms. Do some of them have same meanings?
For routers I have seen "wire speed " term and for switches I have seen other terms.
'Speed' is a bit of a misnomer here. Speed is a practical measurement of bandwidth for a given application. I think whatever you read may have been cross referencing bandwidth and latency, which are more closely intertwined.
Bandwidth is a measure of how much data can pass in a given interval. This is measured in bps (bits per second). Using the age-old analogy of the water hose, it's how fat the hose is, and by extension how fast the water is traveling through it.
Latency is a measure of how long it takes a unit put into the system on one end to come out on the other end. In the example of the water hose, while the water might be rushing out of the hose, if it has to travel many many miles, a gallon put in at the source may take a long time to come out the destination. Latency is usually measured in milliseconds (ms). Usually applications that strive for low latency are time sensitive.
For transferring a file via FTP, bandwidth is a concern. You want the transfer to go fast, which means pushing a lot of data.
For talking on the phone over a VOIP network, latency is a concern. The packets are small, but you need them to arrive on time. High latency will result in a delay between when the speaker speaks and the receiver hears.
'Wire speed' or 'wire rate' simply means that you can take two ports of the same bandwidth/speed, and pump data between them with no packet loss. For example, if I have a fastE port, and data is being switches from that fastE port to another fastE port, I will always get out what I put in. If there is a bottleneck somewhere inside the device (for example, the CPU on a lower-end router), then its possible I will drop packets along the way, and not be able to forward at 'wire rate'. So, the claim of 'wire rate' implies that there is a non-blocking end-to-end data path, and that the forwarding engine is capable of making enough forwarding decisions on packets to not congest that data path (including features such as policing, shaping, queueing, etc). Please keep in mind that this only takes into consideration the data path between two ports on the system. A totally non-blocking system on all ports simultaneously only exists in the dreams of hardware design teams - so take the term 'wire rate/wire speed' with a bit of salt.
Backplane bandwidth is a measure of the bandwidth of the internal architecture of the switch. The term is thrown around pretty loosely to refer to anything internal to the switch that doesn't refer directly to the packet forwarding rate of the forwarding engine.
Total bandwidth - I can't really help you with this one without knowing what it was used in reference to. Usually when I see a vendor talk about 'total' bandwidth it is used as a pretty marketing number, but as I mentioned, it really depends on what you are describing.
I'd say the best two metrics to look at when evaluating the capacity of a switch are:
1.) Forwarding rate
Usually this is the measure of how many 64-byte packets the forwarding engine can make decisions on. This is measured in packets-per-second (pps).
2.) Backplane bandwidth
This is most often the measure of the total switching capacity of the system internally.
Ryan's post defines the terms, but I suspect, although not explicitly asked, you desire a better understanding of a network device's performance for informed product selections.
First realize that almost all network devices can send or receive a single frame at the bandwidth of the link. However, many devices can not sustain "line rate" or "wire speed". Both mean the link is running at 100% with back-to-back frames/packets.
For Ethernet, "line rate" or "wire speed" varies based on the frame/packet size.
Packet Size (Bytes) 64 128 256 512 1024 1518
Theoretical Maximum Kpps 1488 845 453 235 120 81
Second, devices with multiple ports need to be able to move traffic from any port to any other port, but often internal "backplane" bandwidth doesn't always match the combination of all the port's bandwidth.
It's not uncommon to find network devices that can not sustain the forwarding rate and/or bandwith between ports if all ports are sustaining 100% usage. However, real world networks rarely have such high loads. If you determine whether a device's performance capacity will support your load, you can often purchase a less expensive device which will serve just as well as a more expensive device assuming the more expensive device's extra capacity is never used.
Both the 2821 and 2851 provide gig Ethernet ports, but the pps rating of the former is 170 Kpps vs. 220 Kpps. Both can't really sustain 100 Mbps full duplex, let alone gig, but the latter should be able to handle about 29% more traffic than the former.
Both the 24 and 48 gig port versions of the 3560G offer 38.7 Mpps and 32 Gbps fabric. The 48 gig port 4948 offers 72 Mpps and 96 Gbps fabric. Although the 4948, I believe, is a "wire speed" L3 switch, you may not need to purchase that level of performance.
Also not asked, but you should be aware of, how internal bandwidth between ports is implemented can be imporatant. The two most common are a shared bus, sort of single super speed internal link, or a "fabric", usually some type of mesh or cross bar between ports or groups of ports. Also whether the internal bandwidth is "blocking", especially head-of-line-blocking, or "non blocking" can be important.
We are pleased to announce availability of Beta software for 16.6.3.
16.6.3 will be the second rebuild on the 16.6 release train targeted
towards Catalyst 9500/9400/9300/3850/3650 switching platforms. We are
looking for early feedback from customers befor...
Introduction Featured Speakers Luis Espejel is the Telecommunications
Manager of IENova, an Oil & Gas company. Currently he works with Cisco
IOS® and Cisco IOS XE platforms, and NX to some extent. He has also
worked as a Senior Engineer with the Routing P...
In this session you can learn more about Layer 3 multicast and the best
practices to identify possible threats and take security measures. It
provides an overview of basic multicast, the best security practices for
use of this technology, and recommendati...