Computer Science Principles The Internet - 03 Addressing and Routing Information - 06 Packets and fault tolerance

Messages are broken down into small pieces of data called a packet.

A packet consists of about a kilobyte of information, but if you're sending a long email, an audio file, or an image, that would require that the data is broken down into dozens or even hundreds of packets, or more.

When these packets are sent out from the server, there are lots of conditions that can happen as they all navigate the roads of the internet.

The chains of the packets will rarely stay together as a group, and will break apart, and take different paths along the internet.

For instance, if we have a message that needs to be broken down into five packets, they start at the same location.

But, as they are transmitted, they have to navigate the internet along with trillions of other packets of information, which can create traffic and bottlenecks at certain points.

For instance, two packets might have been managed by a node on a network, but that node might have been overwhelmed by another request, making it unavailable to manage the third packet.

Since all the packets are individual units, they can follow different paths, so the third packet takes a different path.

Along the way, all the packets might be on a different path, but they all ultimately get to the same destination.

But, due to traffic and other conditions, they have to take different routes.

The key is the header, which each node examines and determines the next location for the packet to go, inching it closer to its destination.

In all, this takes just a dozen or so milliseconds to complete.

Something you can do to see how this works is to ping the server.

This is a common command in Windows, Linux, and Mac OS that sends test packets to a server, and then requests the server to send the information back.

Using a ping, you can test to see how long it takes to reach, or ping, the server, and if any of the packets are lost along the way.

So, you could ping Google.com, and it will send a handful of packets to Google.com, resolving the domain to the IP address.

It then sends a series of packets, in this case four, each of which are 32 bytes in size.

When the packet is returned, you are told how long it took for the round trip.

At the end, you get a report that says how many packets were sent, how many were received back, and how many were lost in the internet for some reason.

Being able to navigate through various conditions on the internet is called fault tolerance.

Because of how the network is configured, if there was a massive network disruption from a natural disaster, or some other event, the network is able to compensate for it and re-route packets in different ways to get the information to the destination.

Depending on the magnitude of the outage, however, that could mean that the rest of the network, or certain areas, might be overwhelmed due to the additional traffic it needs to manage.

As a system is able to recover, and come back online, the network can then resume normal operations.