Okay, hello Cloud Gurus, and welcome to this lecture.

This lecture, before we get started on Route 53,

and creating our own domain name,
we are going to need to create
an Elastic Load Balancer as a part of that.
And I just wanted to go through
the theory of Elastic Load Balancers.
So what is an Elastic Load Balancer,
or what's a load balancer?
Well, the hint is in the name.
A load balancer simply helps us
balance our load across multiple different servers.
So let's say you've got a whole bunch of Web Servers,
and you want to equally balance the load between them,
you deploy a load balancer,
and then using various types of technology
it will help you balance a load across those Web Servers,
so no one Web Server will get overwhelmed,
and so hopefully your application won't crash.
So that's all a load balancer is.
Now, with AWS there's three different types
of load balancers.
We've got our Application Load Balancers.
And our Application Load Balancer operates
at the OSI Layer Seven.
So if you're not familiar with the OSI Layers,
just go ahead and type it into Google
and click on the Wikipedia article.
It basically will go on and explain
all the different layers of computing.
So, Application Load Balancers can
actually make very clever decisions.
They can actually go all the way down to the
application layer, and they can load balance based off that.
If it could see that packets are destined for,
let's say, sales.acloud.guru it would then send
those packets directly to the sales Web Servers
or application servers.
If it was a different sub-domain,
it could go to a completely different sub-domain.
It can also look at whether or not
someone's making a purchase.
Maybe that would go to a whole bunch
of different application servers.
My point being is that Application Load Balancers
can see all the way up to the application layer,
and they can make very clever routing or rooting
decisions, based on that, you know, information.
We then have out Network Load Balancers,
this operates at Layer Four.
This is where you basically want super fast performance,
and super fast speed.
It is basically AWS's most expensive load balancer
but it is what you would use in production.
Especially is latency is an issue, you want to use
Network Load Balancers.
Then we have out Classic Load Balancers.
And our Classic Load Balancers are pretty much,
you know, they don't recommend them anymore.
But it is there for legacy purposes.
Interestingly the Developer Associate exam
actually tests you mostly on Classic Load Balancers
even though AWS, themselves, are pushing you to
either use Application Load Balancers or
Network Load Balancers.
So, we will have a quick look at
Classic Load Balancers as well.
But in the next lecture, what were going to
go ahead and create is an Application Load Balancer.
So, lets start with Application Load Balancers.
And they're best suited for load balancing
of HTTP and HTTPS traffic.
They operate at Layer Seven and are application-aware.
They are intelligent, and you can
create advanced request routing,
sending specified requests to specific Web Servers.
So here we've got a Tesla Model X
lets say that we have an application inside our Model X.
We can send all out Model X traffic to a specific
set of Web Servers that are only designed for Model X.
And then if we had a Tesla Model S or
a Model 3, for example, we could get our
Application Load Balancers to send them to completely
different EC2 instances depending on the packets.
Basically the Application Load Balancer
could be configured so it could figure out
what type of car was sending the request.
And then it could send that to a group of web-servers
that is designed specifically to
respond to traffic for that car.
We then have out Network Load Balancers.
So, Network Load Balancers are best suited for
load balancing of TCP traffic
where extreme performance is required.
And again I love Teslas, so here is a Tesla Roadster,
it is going to be capable of doing
zero to 60 miles per hour in under two seconds.
So think of extreme performance.
That is going to be one of the, basically, the fastest
production car ever made.
So, operating at the connection Level Layer 4
Network Load Balancers are capable of handling
millions of requests per second,
while maintaining ultra- low latencies.
So, use a Network Load Balancer for an extreme performance.
Do bear in mind, that it is going to be the most costly
load balancer of the three different load balancers.
So, think of extreme performance.
Think of a Tesla Roadster.
You're going to get extreme performance out of them,
but they are the most expensive Teslas that you can buy.
We then have Classic Load Balancers.
We've got our old Banger.
So Classic Load Balancers are
the legacy Elastic Load Balancers.
You can load balance HTTP and HTTPS applications
using Layer Seven specific features,
such as the X-Forwarded-For and sticky sessions.
But you can also use strict Layer Four load balancing
for applications that rely purely on the TCP protocol.
So essentially, you can either do
Layer Seven or Layer Four using a Classic Load Balancer,
but Layer Seven is not anywhere near as intelligent
as an Application Load Balancer.
It is literally just using
X-Forwarded-For and sticky sessions.
And we'll come to X-Forward-For in one more slide,
but before we do that I just
want to talk about load balancer errors.
So, your Classis Load Balancer, if your application
stops responding the Elastic Load Balancer,
or Classic Load Balancer, responds with a 504 error.
This means that the application is having issues.
This could either be at the web-server layer
or at the database layer, but basically it's not
the load balancer itself that's having the issue.
It's something below the load balancer, it's just not
getting a response and that means you need to trouble shoot
either the Web Server layer or the Database layer.
You need to basically identify where the application
is failing and scale it up or out where possible.
So, that's if you get a 504 error, its basically
a gateway timeout error.
Moving on to X-Forwarded-For.
So, we've got our user here.
This is their public IP address.
I've just made one up, it's 124.12.3.231.
They then do a DNS request and they
hit our load balancer.
Our load balancer will take that request
and it will use the private IP address,
so, 10.0.0.23, and then send that on to
our EC2 instance.
Now, our EC2 instance will only see
that private IP address, or see 10.0.0.23.
Now this can cause problems because you probably
want to know where in the world your user is,
where they're coming from, et cetera.
So how can you get the IPv4 public address when your
Elastic Load Balancer is sending you the private address?
Well, really easy, you just got to go and look for it
in the X-Forwarded-For Header, and you will
be able to see the public IP address there, so 124.12.3.231.
So just remember that going into your exam.
So, speaking of exam tips, so basically, just remember
that there's three different types of load balancers.
You've got your Application Load Balancers,
your Network Load Balancers,
and your Classic Load Balancers.
Remember that a 504 error means
that the gateway has timed out.
And this means that the application is not responding
within the idle timeout period.
So how do you troubleshoot this?
We'll go ahead and trouble shoot the application.
Is the Web Server down?
Or is the Database Server down?
Do you need to scale out?
Do you need to scale up? Et cetera.
If you need the IPv4 address of your end user,
using Classic Load Balancers then you just need
to use the X-Forwarded-For header.
So that is load balancers in a nut shell.
Remember these tips going into your exam and you'll do fine.
In the next lecture, what we're going to do, is were going
to register a domain name using Route 53.
We're then going to point that domain name to
an Application Load Balancer, which will have some
EC2 instances inside it.
And then, hopefully, we'll be able to resolve
our domain name all the way down to our EC2 instance.
So, if you've got the time please
join me in the next lecture, thank you.