When you're designing, building, or
troubleshooting a network, you'll be
using a network topology type that can
vary depending on the technology you're
using. If you're looking to understand
the way the data is flowing, or you want
to plan to create a new type of network,
then it's useful to understand what
these topologies might be. This is also
useful during the troubleshooting
process because you'll know exactly what
flow the traffic takes to get between
point A and point B. One of the most
popular types of topologies is the star
topology. You might also hear this
referred to as a hub and spoke, where the
hub is in the middle and the spokes are
along the outside. This is a topology
that you'll find on almost any network,
regardless of the size of the network.
And you'll also find that most devices
are connecting back to this central hub
of the star. For example, a switched
Ethernet network has the Ethernet switch
in the middle of this star. And then all
of the devices run directly back to this
particular switch. They're not connecting
to each other. They're instead connecting
back to the central switch in this star
topology.
Although we don't often see a ring
topology used on our local area networks,
it's still a topology type that's used
quite often for wide area networks. So
although many people will remember the
older Token Ring technologies that we
used to run inside of our local area
networks, we don't generally find those
ring networks any longer. But if you're
connecting over a metropolitan area
network or a wide area network, we use
ring networks extensively. It's not
because a ring technology somehow lends
itself to work better over a wide area
network. It's because we can create
additional redundancy using the ring
topology.
For example, a very common way to send
traffic over a ring network is to have
the traffic simply go in a circle. Now, if
we're on this wide area network and
there's construction going on and
someone happens to sever a fiber
connection that's being used for this
metropolitan area network, then we're not
going to be able to send that traffic
through the rest of that ring. But the
devices that are on either side of that
severed link recognize that traffic is
no longer able to traverse that
connection, and instead will loop back
the connection on those individual
endpoints. So instead of having data go
around a ring, the data will instead go
as far as it can around the ring and
then loop back to get to the other side
of the ring, maintaining uptime and
availability even in the case where part
of that ring may be severed.
Early types of Ethernet networks were
not switched Ethernet, but were instead
run over coax. And this coax was quite
simply a cable that was run down the
middle of the room, very similar to this
cable. This is a bus network. And although
it was commonly used on those early
Ethernet networks, we can still find
modern networks that use the same bus topology.
One problem with bus networks is that it
is a single cable that is running either
through the walls or down the center of
the room. And if we happen to have a
break in this cable, you can see
immediately that it would suddenly
segment the network into different
pieces, or in some cases, cause no data to
be transferred across the network. That's
one of the reasons we moved away from
bus networks for our local area networks--
because one single disconnect could
cause an outage for everybody else on the network.
In our modern automobiles, we have bus
networks that we use extensively. These
are Controller Area Network buses, or CAN
bus connections, and they're used to
connect all of the different sensors and
controllers inside of our automobiles to
be able to make all of our cars much
safer to drive on the roads.
Another popular topology, especially in
larger networks, is to create a mesh
between devices or a mesh between sites.
We may have devices that are connected
in different locations, and we might want
to connect them all together. But instead
of having a single connection to a
particular site, we may want to create
multiple connections to mesh these
together. That way, if we do lose any one
of these network links, we're able to
work around that problem by simply using
one of the redundant connections.
You'll commonly use this type of mesh
design if you're creating redundancy or
fault tolerance, or perhaps you're
designing a load-balanced network and you
can use different parts of the network
to share that load.
Probably the most common place to find a
wired mesh network is over a wide area
network, where you can create multiple
links to other sites so that you can
have a primary connection from one site
to the other, and then a backup or
secondary connection that you can use if
you run into problems.
When you start combining these different
topologies together, you create a hybrid
network. A hybrid network is more than
one of these topology types all working
together. For example, you might have
three remote sites all connecting
devices together using a switched
Ethernet or star network. And then you
may be connecting those together over a
wide area network that uses a ring topology.
If you're using a wireless network, you
may be communicating in a number of
different ways. If you're using an access
point, you're probably communicating over
an infrastructure connection. This means
that all of the devices on your network
are communicating through an access
point. This is probably the most common
way to use wireless connections, but it's
not the only way to communicate over a
wireless link. If you just have two
devices and there's no access point that
you can use, you can connect directly
from one device to another using ad hoc
networking. You don't need an access
point or any other type of wireless
infrastructure. You simply have
one device communicate directly to another
device over this wireless connection.
And if you've added Internet of Things
devices, which are commonly wireless
devices that control our lights, our door
locks, or the air conditioning systems,
then you're probably using a mesh
network where all of these devices can
communicate to all the other devices
simultaneously to create an
interconnected mesh of communication
between all of these IoT devices.
One of the advantages of these mesh
wireless networks is that it allows many
devices to communicate to each other,
even if those devices are very far apart
from each other. This also allows the
mesh network to self-heal. So if you turn
off one of those IoT devices, the
remaining devices will self-heal and
redesign themselves into a mesh network
that will allow them to continue the communication.