00:00:01
If you were to walk
through a data center,
00:00:03
you will see many of
these racks all together
00:00:06
with a lot of
different equipment
00:00:08
installed in each one of them.
00:00:10
All of these devices
work together
00:00:12
to be able to take data
from one part of the network
00:00:15
and move it to another
part of the network.
00:00:17
Each of these devices is
installed for a specific reason,
00:00:20
and it's helpful
if we understand
00:00:22
why we installed that
particular piece of equipment
00:00:25
to begin with.
00:00:26
Over time, we may be installing
more of this existing
00:00:29
equipment into our
racks, or we may
00:00:31
be installing new technology.
00:00:32
So in this video, we'll look
at different types of devices
00:00:36
and how we might use
them in our data center.
00:00:39
Let's start with one of the most
common devices that you'll find,
00:00:42
which is a router.
00:00:44
A router allows us to take
data on one IP subnet and route
00:00:48
that information to a
different IP subnet.
00:00:52
These may be subnets that
are next to each other
00:00:54
in the same data center,
or these IP subnets
00:00:57
may be located in different
parts of the world.
00:01:00
We refer to a router as
an OSI layer 3 device.
00:01:04
At the OSI layer 3,
or network layer,
00:01:06
we're referring to IP addresses.
00:01:09
And IP addresses is exactly
what's used by a router
00:01:12
to be able to determine the
next hop for this information.
00:01:16
You may sometimes see
this routing functionality
00:01:18
also included inside
of an existing switch.
00:01:22
And we'll often refer to these
as layer 3 switches, which,
00:01:26
of course, is referring to
that OSI layer 3 functionality.
00:01:29
It's not that the switch
itself is now operating
00:01:32
at a different OSI layer.
00:01:34
It's just, within that
same piece of equipment,
00:01:36
we have both a layer 2
switch and a layer 3 router.
00:01:40
So we've abbreviated
that as a layer 3 switch.
00:01:43
These routers often connect many
different types of networks.
00:01:47
So we may be connecting a
Local Area Network, or a LAN,
00:01:51
to a Wide Area
Network, or a WAN.
00:01:54
These might also be
copper-based connections
00:01:56
or fiber-based connections.
00:01:58
So we may have routers with
many different connections
00:02:01
or interfaces on them,
and we're connecting
00:02:03
many different diverse
networks to all
00:02:06
of those different interfaces.
00:02:08
Another common device
is a network switch.
00:02:11
Switches operate at
the MAC address layer
00:02:14
to be able to forward traffic.
00:02:15
So we'll often refer
to that as an OSI layer
00:02:18
2 or datalink device.
00:02:20
These operate
mostly in hardware.
00:02:23
The hardware inside
of these devices
00:02:25
is referred to as an ASIC, that
is, an Application-Specific
00:02:29
Integrated Circuit.
00:02:30
There are many different
functions and capabilities
00:02:33
inside of these
switches, especially
00:02:35
if you're using one
designed for the enterprise.
00:02:38
For example, many
of these switches
00:02:40
have the ability
to include power
00:02:42
on the same wires as
your ethernet connection,
00:02:45
and we refer to that as
Power Over Ethernet, or POE.
00:02:49
And as we mentioned
before, you may
00:02:51
hear folks refer to
this as a layer 3 switch
00:02:54
if the switch includes some
type of routing functionality
00:02:57
built into the device itself.
00:03:00
Security on our networks
is also important.
00:03:03
That's why you probably are
using a firewall at home
00:03:06
and you most certainly have
a firewall in your office.
00:03:10
A traditional firewall
allows you to filter traffic
00:03:12
based on a TCP or
UDP port number,
00:03:15
but if you have a
more modern firewall,
00:03:18
you're probably using a
Next-Generation Firewall,
00:03:21
or NGFW, which is able to
identify applications traversing
00:03:26
your network and
allow you to manage
00:03:28
whether that application should
be allowed or not allowed
00:03:31
on your network.
00:03:33
Most firewalls also have
additional functionality.
00:03:36
For example, it's
common to find firewalls
00:03:38
that will allow us
to encrypt traffic
00:03:40
traversing the network through a
Virtual Private Network, or VPN.
00:03:44
It's very common to have a
firewall at one remote site
00:03:47
and a firewall at
another remote site
00:03:50
and be able to create
an encrypted tunnel
00:03:52
between those firewalls
using this VPN functionality.
00:03:56
And most firewalls
can also operate
00:03:58
as a layer 3 device, which
means the firewall themselves
00:04:01
can act as a router.
00:04:03
That's because they
are often sitting right
00:04:05
between the ingress and
egress point of your network,
00:04:09
where all the traffic on
the inside of your network
00:04:11
is going to the outside
or internet connection
00:04:14
and your internet
traffic is coming inbound
00:04:17
to your local network.
00:04:18
We rely on the firewall
to be able to manage
00:04:21
the communication
between the inside
00:04:23
and the outside of the network.
00:04:25
To be able to perform
this functionality,
00:04:28
many firewalls also provide
Network Address Translation,
00:04:31
or NAT.
00:04:32
And because they
are a router, it's
00:04:34
very common to have dynamic
routing protocols supported
00:04:37
inside of the firewall as well.
00:04:40
Many data centers might also
have standalone IDS or IPS
00:04:44
devices, although much
of that functionality
00:04:46
is also integrated into the more
modern next-generation firewall.
00:04:50
IDS refers to an Intrusion
Detection System,
00:04:54
and the IPS refers to an
Intrusion Prevention System.
00:04:58
Both of these work
in similar ways.
00:04:59
They're looking for attacks
that are inbound to your network
00:05:03
and are able to identify,
alert, and in many cases,
00:05:06
prevent that attack from
gaining access to your network.
00:05:10
These are commonly
known attack types.
00:05:12
These might be exploits
against operating systems
00:05:14
or the applications
that you're using,
00:05:17
and they might take advantage
of known vulnerabilities
00:05:20
with those
applications or systems
00:05:22
by taking advantage of a buffer
overflow, a cross-site scripting
00:05:25
vulnerability, or other
known vulnerabilities
00:05:28
to those systems.
00:05:29
If you're using an
intrusion detection system,
00:05:32
it's able to alarm or
alert if it ever sees
00:05:35
any of these inbound attacks.
00:05:38
If you're using an
intrusion prevention system,
00:05:41
it's able to go a step further
and block that particular attack
00:05:44
before it gets inside
of your network.
00:05:47
Since an intrusion
detection system is not
00:05:50
able to block that
traffic, it's very common
00:05:53
to see an intrusion prevention
system used on our enterprise
00:05:56
networks.
00:05:58
If you've ever
used a website that
00:06:00
may be accessed by millions
of people every day,
00:06:03
you may be wondering how that
site is able to remain up
00:06:07
and running without
any type of downtime.
00:06:09
In most cases, it's because that
site is using a load balancer
00:06:13
to be able to
distribute that load
00:06:15
across multiple
physical servers.
00:06:18
As the end user, you may have no
idea that this load balancing is
00:06:21
taking place, but
if you were to look
00:06:23
at the data center
for this organization,
00:06:26
you might find a large number of
web servers or database servers
00:06:30
in farms that can be used in
conjunction with this load
00:06:33
balancer to maintain
uptime and availability.
00:06:36
These load balancers are
also very good at identifying
00:06:39
any outages to these servers.
00:06:41
So if one of the servers happens
to fail due to a hardware
00:06:45
error or some type
of software problem,
00:06:47
the load balancer can
recognize the issue,
00:06:49
take that server
out of the rotation,
00:06:52
and continue to provide
access to these services using
00:06:55
the remaining devices that are
connected to the load balancer.
00:06:59
Here's a common design
for a load balancer
00:07:02
where users on the
internet would be accessing
00:07:04
a service at a location.
00:07:06
To the end user's
perspective, they're
00:07:08
accessing a single
server, but they're really
00:07:10
accessing a load balancer
that is distributing
00:07:13
that load between multiple
servers inside of that company's
00:07:16
data center.
00:07:17
These load balancers can also
optimize the communication.
00:07:21
For example, it may
perform TCP offloading
00:07:23
so that the communication
to all of these servers
00:07:25
on the inside of the
network are occurring
00:07:27
as quickly as possible.
00:07:29
These load balancers
can also act
00:07:31
as an SSL offload, which
means that they will provide
00:07:34
the encryption and decryption
capabilities instead
00:07:38
of having the servers
themselves manage that process.
00:07:41
Data might also be cached
on the load balancer
00:07:43
so requests made to
the load balancer
00:07:45
can be answered
immediately instead
00:07:47
of going all the way down to
the server to provide that data.
00:07:50
And load balancers are also
very good at prioritizing
00:07:53
different types of
traffic over others.
00:07:56
There might be
certain web pages that
00:07:58
have higher access than
others, and you can commonly
00:08:01
perform that prioritization
using Quality of Service,
00:08:04
or QOS.
00:08:05
Load balancers can also provide
application-centric load
00:08:08
balancing, where
certain pages may
00:08:11
be located on certain
servers and all
00:08:13
of the requests to those
pages would go exclusively
00:08:16
to those individual servers.
00:08:19
Many organizations have security
concerns about individual users
00:08:23
being able to
directly communicate
00:08:25
with a server or service
that's on the internet.
00:08:28
One of the ways that the
organization can manage
00:08:30
these connections is
by putting a device
00:08:32
in the middle of this
conversation called a proxy.
00:08:36
This proxy is responsible for
taking the user's request,
00:08:39
performing that request
on their behalf,
00:08:42
receiving the answer
to that request,
00:08:44
verifying that the
answer doesn't contain
00:08:47
some type of malicious
software or malicious code,
00:08:50
and then providing that
answer to the end user.
00:08:53
That is the purpose
of a proxy, to sit
00:08:56
in the middle of
the communication
00:08:58
and make that communication
on the user's behalf.
00:09:01
Since the proxy is sitting in
the middle of the conversation,
00:09:04
it's a perfect place to
do caching so the user can
00:09:07
make a request to a web server.
00:09:09
If that request has already
been cached by the proxy server,
00:09:12
the answer can go
right back to the user
00:09:14
without having to
access the internet.
00:09:16
We might also provide access
control from the proxy server
00:09:20
so that we can request
a username and password
00:09:22
from the user in order to
gain access to the internet.
00:09:26
From that point, we
might want to filter URLs
00:09:28
or perform some type of
content scanning to make sure
00:09:31
that the user is not receiving
any type of malicious software.
00:09:35
Some proxies require you to
configure the operating system
00:09:38
or the applications that you're
using to identify the proxy
00:09:42
and be able to use that to
send and receive communication.
00:09:46
But not all proxies work
in that explicit manner.
00:09:49
There are also
transparent proxies
00:09:52
that will work
invisibly without making
00:09:54
any changes to the operating
system or the applications
00:09:57
in use.
00:09:58
It's very common to store
documents and other files
00:10:02
on centralized
storage facilities
00:10:04
inside of our data centers.
00:10:06
One type of storage
is referred to as
00:10:08
a Network-Attached
Storage, or NAS.
00:10:11
We often refer to this
network-attached storage
00:10:14
as providing file-level access.
00:10:16
That means that if we wanted
to gain access to information
00:10:19
within a file, we need
to pull the entire file
00:10:21
across the network into
the memory of our system.
00:10:24
And when we're writing
information or changing
00:10:26
information in
that file, we will
00:10:28
need to write the entire
file back to the NAS.
00:10:31
A more efficient
way of communication
00:10:33
might be through the use of a
Storage Area Network, or a SAN.
00:10:38
This is very similar to
reading and writing information
00:10:41
from a local
storage drive, where
00:10:43
instead of copying
the entire file
00:10:45
to be able to change just a
bit of information within it,
00:10:49
we have block-level
access, which
00:10:51
means that we can change
just the blocks that
00:10:53
have been modified.
00:10:54
And when you have
very large files,
00:10:56
this can be a very
efficient way to modify
00:10:58
just a little bit of information
within that very large document.
00:11:03
Whether you're using
a NAS or a SAN,
00:11:05
you're probably transferring a
lot of files to these systems.
00:11:09
And for that reason,
we want to be sure
00:11:11
that we're using the
most efficient method
00:11:13
of communication.
00:11:14
It's very common, for
example, to put the NAS
00:11:16
or the SAN on its
own isolated network,
00:11:19
and it's commonly a network
that has very high bandwidths.
00:11:23
If you're in your office
and you look at the ceiling,
00:11:26
you might see a
device like this.
00:11:27
This is an access point.
00:11:29
This device allows us to
communicate wirelessly
00:11:31
from our device to the
rest of the network.
00:11:34
This is not the
wireless router that you
00:11:37
might be using at home, which is
a router and a wireless access
00:11:41
point and a switch
in the same device.
00:11:44
When you're in larger
enterprise environments,
00:11:46
you're usually using a
device that is purpose
00:11:49
built for a single function.
00:11:51
And having an access
point means that we're
00:11:53
using this for wireless
communication and wireless
00:11:56
communication only.
00:11:58
On the other side
of this wireless
00:12:00
access point is very commonly
an ethernet connection.
00:12:03
So this is bridging
communication
00:12:05
between the wireless network
and the wired ethernet network.
00:12:09
That's why we refer to access
points as an OSI layer 2
00:12:13
device, or a data
link layer device,
00:12:16
because it's making that
translation between the 802.11
00:12:19
wireless network and the
802.3 ethernet network.
00:12:24
In most businesses of
any size, you probably
00:12:27
have more than one access point.
00:12:29
That's because you probably have
a very large building or series
00:12:32
of buildings, and
you need to be sure
00:12:34
that everyone is able to access
that wireless network wherever
00:12:38
they happen to be inside
of these buildings.
00:12:40
But this means that
we have to manage
00:12:42
many different wireless
access points wherever
00:12:45
they might be in
our local network
00:12:48
or in a remote site network.
00:12:50
And we might need to manage
security settings, access
00:12:53
policies, and other
configuration parameters
00:12:55
within that access point.
00:12:58
We also have users that may
be very mobile and moving
00:13:01
between different
parts of the building
00:13:02
or moving from one
building to the other.
00:13:04
And we need to make sure
that they can seamlessly
00:13:07
roam from one access
point to the other
00:13:09
so that they are always
connected to the network.
00:13:12
Instead of connecting to
each individual access point
00:13:16
to make these configuration
changes or manage this process,
00:13:20
we can have a centralized
management tool
00:13:22
that allows us to manage
all of our access points
00:13:25
from one central place.
00:13:27
This is a wireless
LAN controller,
00:13:29
and it gives us that
single pane of glass
00:13:31
so that we can manage the entire
infrastructure while we're
00:13:35
sitting in one chair.
00:13:36
From this single device, we
can deploy new access points
00:13:40
with a full configuration.
00:13:41
We might want to set up
performance or security
00:13:44
monitoring and be
alerted if we happen
00:13:46
to see anything across
any of our access points.
00:13:49
We can also take any changes
that we need to make and deploy
00:13:52
those automatically to
all of our access points
00:13:55
with one click of the mouse.
00:13:57
This also allows
us, very commonly,
00:13:59
to create reports on how much
our access points are being used
00:14:03
and be able to understand if
we need to update or change any
00:14:06
of our access point locations.
00:14:08
These are often
proprietary systems.
00:14:10
So if we have an
access point from one
00:14:13
particular
manufacturer, then we're
00:14:15
also using the wireless
LAN controller that is also
00:14:18
from that same manufacturer.