|Computer Networking FAQ #18|
<< Previous Question -- Return to FAQ Index -- Next Question >>
This one gets asked on many entry-level and intermediate-level networking exams.
Q. "What is the 5-4-3-2-1 rule of network design?"A. The 5-4-3-2-1 rule embodies a simple recipe for network design. It may not be easy to find examples in practice, but this rule neatly ties together several important elements of design theory... (See below)
... . To understand this rule, it's first necessary to understand the concepts of collision domains and propagation delay. Collision domains are portions of a network. When a network packet is transmitted over Ethernet, for example, it is possible for another packet from a different source to be transmitted close enough in time to the first packet to cause a collision on the wire. The total range over which a packet can travel and potentially collide with another is its collision domain.
Propagation delays are a property of the physical medium (e.g., Ethernet). Propagation delays help determine how much of a time difference between the sending of two packets on a collision domain is "close enough" to actually cause a collision. The greater the propagation delay, the increased likelihood of collisons.
The 5-4-3-2-1 rule limits the range of a collision domain by limiting the propagation delay to a "reasonable" amount of time. The rule breaks down as follows:
5 - the number of network segments
4 - the number of repeaters needed to join the segments into one collision domain
3 - the number of network segments that have active (transmitting) devices attached
2 - the number of segments that do not have active devices attached
1 - the number of collision domains
Because the last two elements of the recipe follow naturally from the others, this rule is sometimes also known as the "5-4-3" rule for short.