Routing Policies

Video Activity
Join over 3 million cybersecurity professionals advancing their career
Sign up with
Required fields are marked with an *
or

Already have an account? Sign In »

Time
19 hours 19 minutes
Difficulty
Intermediate
CEU/CPE
20
Video Transcription
00:00
>> Hello everybody and welcome back.
00:00
In this lecture, we're going to be
00:00
talking about routing policies.
00:00
The learning objectives here are going to be
00:00
to help you understand what routing policies are,
00:00
we're going to review the different types of
00:00
routing policies available to you,
00:00
and then we'll discuss how
00:00
these policies apply and when to use them.
00:00
Route 53 routing policies.
00:00
Basically, this allows you to define
00:00
how DNS queries are handled when it comes to Route 53.
00:00
This is not the same thing as load balancer routing.
00:00
This is going to be specific
00:00
to DNS and how DNS responses to these queries.
00:00
The different types you have
00:00
are simple, weighted, failover,
00:00
latency based, geo-location,
00:00
multi-valued answer, and geoproximity.
00:00
Beginning off, let's go ahead and talk about simple.
00:00
This is typically the one that we're going to be
00:00
using but if you have
00:00
a very specific need and you are handling DNS queries,
00:00
then we can use one of the other ones.
00:00
But if you're not, if you don't have
00:00
one of those specific needs,
00:00
then simple is perfectly fine.
00:00
Simple typically route the traffic to a single resource.
00:00
So if you have one web server,
00:00
you're not going to be leveraging other end points.
00:00
Maybe you'll say you have one web server,
00:00
it's only here in the United States and you're not
00:00
planning on rebuilding the site in Chinese,
00:00
then you're not going to have a single resource
00:00
over in China.
00:00
If you're just having one here in
00:00
the United States and its English,
00:00
then simple is the way to go.
00:00
You can specify multiple values,
00:00
but only one is chosen at random.
00:00
Alias enabled can only specify
00:00
one AWS resource and
00:00
you don't have any health checks in the situation.
00:00
A weighted policy controls
00:00
the percentages of requests going to the resources.
00:00
When would you use this?
00:00
Weighted policy is when you would
00:00
have maybe two different servers
00:00
and they both have the same website,
00:00
but you're not trying to overload the website.
00:00
This is common in e-commerce and stuff like that.
00:00
Maybe one is bigger than the other
00:00
and you want to equally or you
00:00
want to distribute the weights of
00:00
the traffic based on the capacity of sets server.
00:00
So let's say the beefier server,
00:00
the server that can handle more of
00:00
the requests because it has more resources within,
00:00
can take on 70 percent of the traffic and
00:00
the remaining 30 percent will go to
00:00
the smaller server that
00:00
you haven't invested as much money in,
00:00
but you still need it because throwing
00:00
a 100 percent of
00:00
the traffic at one single server instance,
00:00
which is completely overworking,
00:00
it would crash the server.
00:00
This is a way to offload it and deal
00:00
with some of that cost at scale.
00:00
This is going to leverage
00:00
relative weight distribution using policies.
00:00
Like I said, you can distribute that way.
00:00
The DNS records must be the same name and
00:00
type and you can
00:00
leverage health checks and you can do load balancing.
00:00
Latency based routing policy redirects
00:00
the resource that has the least latency, so the closest.
00:00
This is going to be done based on where
00:00
the end-user is trying to surf from.
00:00
If you have an end-user in Florida,
00:00
then they're going to be
00:00
routed to the computer resource
00:00
that's closest to Florida.
00:00
They're not going to be sent to
00:00
the web server that would be in Oregon,
00:00
for example, they would find one that's closer.
00:00
The reason why you want to do this is to avoid latency.
00:00
Latency is that slowness,
00:00
it's that lagginess that takes
00:00
place whenever we're doing something on the Internet.
00:00
Ideally, we want to
00:00
avoid latency because it helps with SEO,
00:00
it helps with user engagement.
00:00
People want things fast, they want it now.
00:00
So going with latency based routing
00:00
is a great way to make everyone happy.
00:00
Using this policy,
00:00
it's very helpful to
00:00
the users because we're making them a priority
00:00
and it's based on
00:00
the traffic between the users like I said.
00:00
We can leverage health checks
00:00
to make sure that the endpoints that
00:00
the end-users are searching are available and if we need
00:00
to rebuild them we can rebuild
00:00
them automatically based on those health checks.
00:00
Geo-location is another routing policy that we have here.
00:00
It's based on the user's location as
00:00
well and we can specify the location by the continent,
00:00
the country, or the state.
00:00
Geo-location could be helpful for multi-language sites.
00:00
It could be the same site,
00:00
but we want to offer it in different languages.
00:00
Well, we can use geo-location to
00:00
help route that users to
00:00
the correct domain based on
00:00
their location and the language that they speak.
00:00
This is very popular with
00:00
something called website localization,
00:00
which is exactly what I just described.
00:00
Also restricting content, which sometimes
00:00
takes place based on which country you're coming from.
00:00
If you're in one of the Korean countries or in China,
00:00
there's certain areas where they don't
00:00
allow the population to access certain sites.
00:00
You may have heard of Google
00:00
being restricted in certain areas,
00:00
or Facebook being restricted in certain areas.
00:00
That is how that's done, is through geo-location.
00:00
This can be associated with health checks as well.
00:00
Routing failover. This is
00:00
another routing policy that we can use.
00:00
This is used to route traffic to
00:00
a new AWS resource whenever the primary resource fails.
00:00
This can be used with health checks as well so.
00:00
Pretty self-explanatory there.
00:00
Geoproximity is another routing policy here.
00:00
This routes traffic to the resources based on
00:00
the geographic location of the user and the resources.
00:00
This gives us the ability to shift
00:00
more or less traffic to resources.
00:00
You can use 1-99 to increase traffic
00:00
or negative 1 to negative 99 to decrease traffic,
00:00
and you can do this based on geoproximity.
00:00
One example that I've seen of this being
00:00
used is with a mobile application that
00:00
users were using to
00:00
access a gas station to purchase sandwiches and coffee.
00:00
It was one of those larger gas stations
00:00
and it's here in the United States.
00:00
Now, if you're thinking,
00:00
why would somebody buy that from
00:00
a gas station? Well, it's a nice place.
00:00
It's like a Starbucks inside the gas station basically.
00:00
They leveraged geoproximity to
00:00
allow the users to access the mobile app and
00:00
select what kind of sandwiches they wanted to
00:00
purchase and have it
00:00
ready before they arrive at the gas station.
00:00
It's all done through geoproximity because it's
00:00
selecting the nearest gas station to them.
00:00
That's how that's done.
00:00
You can do that and you can leverage
00:00
the ability to shift more or less traffic depending
00:00
on the resources and so forth
00:00
right here through the routing policy,
00:00
which is pretty interesting.
00:00
Another routing policy is multi-value.
00:00
This is used when routing
00:00
traffic to multiple different resources.
00:00
You can also return multiple values and resources,
00:00
and can be associated with
00:00
health checks and you can leverage up
00:00
to eight health records
00:00
that are returns for each multi-value query.
00:00
That about wraps up this lecture on
00:00
the various different types
00:00
of routing policies available to.
00:00
Hopefully you found this helpful.
00:00
There is going to be a lab
00:00
where you're going to get a chance to
00:00
leverage health checks and
00:00
you're going to get a chance to play around with DNS.
00:00
I encourage you to use that and if you have
00:00
any questions or you want
00:00
a little bit more clarity on
00:00
Route 53 and routing policies,
00:00
make sure to review
00:00
the documentation because there's a lot of
00:00
very valuable information on
00:00
how these routing policies are applied.
00:00
They can provide multiple examples and
00:00
even give you some ideas on
00:00
how to stand them up yourself.
Up Next