(dramatic music)
- Greetings everyone,
Raymond Lacoste here,
with Stormwind Studios,
and I wanna take this opportunity to save you money.
Now how am I gonna save you money?
Well, I am going to give you six tips
to save more money with AWS.
Why?
Because I saved 50% because it was on sale.
So how are you gonna save money with AWS?
Well first and foremost,
you wanna pick the right size EC2 instance.
It's as simple as this.
Don't use an instance
that is gonna give you more than you need.
More what?
More CPU, more memory, more storage, more networking.
If you use an instance
that gives you more than what you need,
then you are paying for resources
that you don't actually need.
So don't over provision,
and also, don't under provision.
So to make sure you don't over
or under provision your resources,
you wanna adapt.
So if you realize you've over or under provisioned,
scale up, scale down, scale in, scale out.
That's right, adapt as the needs of your business change.
I want you to take a look at reserved or spot pricing.
A lot of people are only familiar with on-demand pricing,
because they haven't taken the time
to learn about what is reserved pricing,
or what is spot pricing when it comes to AWS.
Well, reserved pricing
is essentially a contract between you and AWS,
that states for the next year, or three years,
you can do an all, partial, or no up-front payments.
And, for 24 hours a day, seven days a week,
365 days a year,
for that one or three year timeframe,
you will pay this amount of money on an hourly basis
for that instance.
You might be saying, "Wow, gee, that's a lot of money."
Well wait, if you compare that same timeframe
to on-demand pricing,
you could save up to 75% if you use a reserved instances,
over on-demand pricing,
over that period of time, so plan ahead.
Spot pricing.
Oh, this is an interesting one,
it's like playing the stock market.
What I'm gonna do,
is I'm gonna set a certanin price limit,
that I wanna pay for using an EC2 instance.
If the price is below that value,
then the instances are gonna be launched,
and I will be able to use those instances
for what I wanna use them for.
However, if the price goes above that value,
so let's say I pick 10 cents,
and the price goes above 10 cents per hour,
then those instances are gonna be terminated,
or no instances are gonna be launched.
Why?
Because it's above my price.
So I bid on spare EC2 computing capacity
that's not being used in that moment in time,
and I could save up to 90% over on-demand pricing.
That's right, up to 90%.
That's a huge cost savings.
So in your organization,
if you plan properly,
you're gonna have a mixture of on-demand pricing,
reserved pricing, and spot pricing,
so that way you can save as much money as possible,
running those EC2 instances.
Now, do not pay for something you are not using.
How many times have you told somebody in your household,
"Shut off the light when you're exiting the room!"
Right?
Your child leaves their bedroom, they leave the light on,
eventually you realize the light's on,
and you remind them again, "I don't wanna pay
"for something I'm not using."
Meaning, I don't wanna pay for that electricity,
if nobody is in that room.
Same thing here.
Terminate instances that are not being used.
Now if you're using reserved pricing,
you're gonna pay for it no matter what,
whether it's running or not.
That's the idea behind reserved pricing.
But you're gonna save a lot of money.
So I'm talking about on-demand instances here.
If there are on-demand instances running,
and you're not using them to their potential,
or using them at all,
terminate them, get rid of them!
'Cause you don't wanna pay for something
you are not actually using at that moment in time.
Now, EBS volumes are gonna be used with our EC2 instances.
So if you have EBS volumes
that are no longer attached to an EC2 instance,
or you have snapshots of EBS volumes
that are no longer needed,
then you're paying for storage that is not needed anymore.
So if you come across any EBS volumes
or snapshots in your account,
that you're no longer utilizing, or no longer need,
make sure you delete them,
so you don't pay for that storage anymore.
I really like being advised
on things that I, well, need to be advised about,
and we have this tool that's given to us by AWS
called the Trusted Advisor.
And I want you to use Trusted Advisor.
AWS wants you to use Trusted Advisor.
Why is that?
Because Trusted Advisor is going to scour your account,
and locate resources, and services, and features,
that you can save money on.
What?
Did Raymond just say that properly?
I did.
Trusted Advisor will go through your account,
and find areas that you can save money on,
as well as other things.
But I really like the saving money-aspect of it.
So AWS created a tool,
that will allow me to pay them less money.
Huh, that is great.
I love this tool.
So take advantage of Trusted Advisor.
One of the things that people get, or forget about,
that's what I'm trying to say, forget about,
is elastic IP addresses.
So here's an example.
Elastic IP addresses is simply a public IP address
that's given to you by AWS.
And I might need one, two, three, four, five, whatever,
and then I associate those public IP addresses
with EC2 instances.
If I am using those public IP addresses on EC2 instances,
I don't pay for them.
That's right, I don't pay for them if I'm using them.
I only pay for them if they're associated with my account,
and I'm not using them.
So AWS does this
so you're not hoarding public IPB 4 addresses.
They'll start charging you for them
if you're not actually using them.
So Trusted Advisor will look to see
if you have any unused elastic IP addresses,
and warn you about them,
so that way you can give them back
so you no longer have to pay for them.
So there's a great advantage of using Trusted Advisor.
AutoScale, use CloudWatch, what does that mean?
Well CloudWatch is gonna allow me to monitor my resources.
So let's say we have instances running,
and they're in a group,
and the CPU utilization gets to 80%.
Well, then we can use AutoScaling to scale up,
so that way that as a group,
we add more servers to the group,
and then CPU utilization goes down.
But we can also use this in reverse,
meaning that if we have a whole bunch of servers
that are a part of an AutoScaling group,
and resource utilization drops,
let's say down to 20% amongst the group,
then we take a server away.
So every time as a group,
if those servers combined equal 20%,
take a server away, take a server away, take a server away,
so that way we don't have to pay for servers and resources
that we don't really need to pay for.
So we can grow and shrink the number of instances
to meet the needs of our organization automatically,
and save us money.
And lastly here, Cost Explorer.
With Cost Explorer,
you're gonna have the ability to keep track of costs,
you'll be able to keep track of usage trends.
You'll even be able to keep an eye on your billing.
So set up billing alerts,
so that way there you'll be notified
if your bills get out of control.
Meaning that you get a bill that's gonna be coming in,
that's way more than you anticipated.
So you might say, "Well, on average,
"we only spend $200, or not even on average,
"maximum we've ever spent in the last year,
"has ever only been 150 bucks."
I am going to set a billing alert at $200.
So if you ever go over $200,
I'm gonna be notified right away,
so that way we can take action
and adjust resource usage as needed.
So folks, I really hope you will take advantage
of my tips for saving money with AWS.
Until next time folks,
everyone here at Stormwind Studios
wishes you the utmost fantastic journey with AWS.
Take care everyone, and bye for now.
Không có nhận xét nào:
Đăng nhận xét