ARMAN HEZARKHANI:
Hello, everybody. So my name's Arman
Hezarkhani, and today I'm going be talking to
you all about Cloud. So to start out, my name-- is that the-- OK, cool. So we're going to
talk about Cloud. So first of all,
my name is Arman. I'm a student at Carnegie
Mellon University, and I'm going to be
a senior in the fall. I study electrical and
computer engineering. And as Damien mentioned,
I'm the head TA for one of the biggest computer
science courses at that school. This slide isn't correct, but
I'm just going to keep going. So in addition to that,
I'm an intern at Google. So currently I'm what's
called the developer programs engineer. And what that means is that I
write code every single day. The code I write, however,
is not for production. So the code I write doesn't
go into Google Cloud Platform. The code I write is basically
to enhance the developer experience. So I would write a code sample
that would go on GitHub, and developers
could see and learn how to use a specific API. And so this is my second
DPE internship at Google. This year I'm on G Suite, which
many of you know and love. And last year I was on
GCP, but I was specifically in the context of EDU. So I've experienced talking
to not only students, but also a lot of faculty
members, CTOs, and CIOs of universities about how
to get GCP in a university. And so today I'm going
to be talking about GCP, and I'm going to be answering
three questions for you all. First of all, I'm going
to answer, what is Cloud? And so who here
has heard of Cloud? Right. We hear it every day. We hear all these new
startups spinning up that are going to be
cloud-based startups. And they love that
buzz word, right? And we hear every
day that Cloud is going to be very important to
our futures, but a lot of us don't actually know
what Cloud truly is. And so I'm going to
hopefully clarify some of the misconceptions
and a lot of the questions that we have in our heads
about, what is Cloud? In addition to that, I'm
going to say, what is GCP? So I just introduced
Cloud, and then I'm going to plug GCP into that mix. So how do GCP's tools fit into
this idea and story of Cloud that I've just told? And lastly, I'm going
to answer the question, how do I get started? So you all have laptops out. That's great. Because we're actually
going to get hands on, and we're going to
log into the console, and we're going to spin
up a virtual machine. So everyone's moving. You don't need to
get them out yet. I will tell you when. For now, just sit back,
relax, take some notes. As Aaron was saying,
you can take out a pen and pencil the
old fashioned way. But for now, we're just
going to be talking. And so the first question I'm
going to answer, as I said, is what is Cloud? So before I answer that
question specifically, let's talk about a little
bit of history. So I'm going to go through
three different stages of, like, how companies and
people use computers, and then it's going to end,
as you can see, in Cloud. And then we're going to get
into what that actually is. So first of all, we're going
to talk about the world as we used to know it,
which is when companies were in the on premise world. So basically, this
was back in the day. Let's say I'm a startup,
and I'm spinning up the newest, like,
social media website. Let's call it, like,
facenotebook.com or something. And I have five users. At that point, I have one
server sitting in my closet, and I have one person
just managing that server. Everything's fine because
I only have five users. The next day, though, word
spreads and I have 1,000 users, and then my website
keeps crashing. And I'm like, why
is that happening? And does anyone know the answer? It's cause I only
have one server, and one server is not going
to handle that traffic. So I have to do
something called scaling. And nowadays, we
think about scaling-- again, that's another buzzword
that startups love to use. But back in the day,
scaling was this-- it was I had to go to a store,
buy a brand new computer, put it on top of the other
computer in my closet, download code on that,
do a bunch of other stuff to add it to my
network, and then I could now handle 1,000 users. And that took almost
an entire day. That was a huge headache. And so that was just-- that's basically one of the
things that lead to Cloud. So that pain point is what
led to the next thing, called timesharing. So we have startups that
have these computers, but on the other
side of the aisle we have these huge companies. Think, like, IBM, right? And they have a lot of money. They can buy gigantic
sets of computers. They don't need to
guess and say, OK, I have 1,000 users today. What if I have a
million tomorrow? And then I'm going to buy that
many computers to basically manage that scale. IBM, they can just go out and
buy a million computers, right? But what if they
over-budgeted, and now they have, like, 200 computers
that are not being used? Then one smart person said,
hey, let's rent these out. Let's allow other companies
like these startups to rent out these computers
for certain amounts of time so that then it's
cheaper for them and we're making passive income. And so that's how we got to
the idea called timesharing. And then this got
more and more popular, and then that just
became mainstream. It became very normal for
startups, smaller companies, to go to large companies
and rent out their computers for a certain amount of time. And then after a while,
these bigger companies said, oh, what if we
could make more money off of this timesharing thing? What if we provided
additional services on top of this infrastructure
that we're selling? And these additional
services, this suite of services that was created,
is what's called Cloud. And so I didn't necessarily get
into the details of how people use Cloud, but now I'm going
to get into that to hopefully clarify what Cloud really is. Going one level deeper. So Cloud is oftentimes broken up
into three different umbrellas of tools. And so the first set of tools is
what's known as Infrastructure as a Service. Infrastructure as a
Service, that's timesharing. That is, I have a computer, and
I'm not using it all the time. So let me sell it for a
certain amount of time to these other companies. I have storage disks. Let me sell this infrastructure. It's literally hardware
that you can touch and feel. So I'm selling that. And then, as I said,
these companies said, what if we can
add additional tooling to this infrastructure? What if instead of saying,
OK, I have 10 users today, and then I have
100 users tomorrow, and making the
companies manually scale that on our network, what if we
just had a tool that said, hey, startups-- whether you have 10 users
or a million users-- we will handle that for you. You just need to write
the code for your website, and we will fully scale for you. So that's basically
infrastructure plus a little bit of magic. I throw those stars
up there to, like-- they're a little bit of
magic in the background. Because it's basically
infrastructure with some software involved,
but it's quite magical. I like to think of it
that way, at least. And then lastly is
Software as a Service. So these are basically
tools that you know. Think Google Docs, Google
Drive, Slides, and Sheets. There are also other companies
that are in the Software as a Service space
like Salesforce, for example, if
you've heard of them. And so this is basically-- these
are the three different tiers or three different
umbrellas of Cloud. So now, what is Cloud? I basically just
gave some background, then I talked about
the three umbrellas. So what is it? In one sentence, Cloud
is a set of tools that helps you spend
less time managing and spend more time creating. So let me say it again. Cloud is just a set of tools
that helps you, the developer, spend less time
managing and more time being creative as a developer. So I don't need to
hire an entire staff to scale my infrastructure. I can just pay Google
a little extra money, and they'll fully scale for me. So I don't need to
put time into that. I can put time into
building new features, talking to my customers. That's what Cloud is. And so now, what is GCP? GCP is obviously a cloud
tool, but it's many things. And so before I get into
the specifics of GCP, let me talk about
Google Cloud as a whole. So Google Cloud is the mixture
of both GCP on the right and G Suite on the left. And so GCP, in a basic sense, is
the infrastructure and platform of the service
side of the house. So those are things
like, like I said, infrastructure and platform,
where you can run code, store data, things like that. And then G Suite, as
you all probably know, is things like Google
Docs, Google Drive. That is the software as a
service side of the house. And so within the
umbrella of Google Cloud, this is how we divide things up. On the right in
GCP land, we have infrastructure and platform, and
on the left, we have software. Today we're only
talking about GCP, but as you all probably know,
G Suite also has a ton of APIs that you can use to
enhance the GCP experience. And they use a lot of the
same consoles, for example. And so you should really
look into that as well. So on GCP, you can
do a lot of things. First of all, you can run code. So you can spin up
a virtual machine, and you can literally
just run a giant script that'll take a week. Instead of running it
on your own computer and worrying about the battery
dying and this and that, you can just put it on a
virtual machine in the Cloud, or you can deploy
a web application and have the code run on Cloud. You can store data. So we have SQL databases. We have NoSQL databases. We have file storage. All of these things I use
almost every single day to build web apps-- all these tools. And again, it's mostly all
fully managed and fully scaled. We have tools for
data migration. So this is for,
like, companies that have a bunch of
data in, let's say, disks in some closet
somewhere, and they want to get that
information onto the Cloud. We have tools that
allow people to do that. We have tools for networking. So in less than a
minute, you can spin up a world wide network-- literally from here to India. I could spin that up
in less than a second. We have tools for networking-- I mean, sorry-- for monitoring. So let's say you have all
these tools in your Cloud, in your platform, and you want
to monitor how they're working. If one thing crashes, let's
see why that happened. We have tools that
allow you to do that. We have tools that
help you build Internet of Things projects. So you should look
into IoT Core. Also we have tools that
help administrators manage this entire platform,
because projects like we're going to see later on
with big companies like Spotify and Coca-Cola, they have
a lot of stuff going on in their Cloud, so they need
some way to manage all of that. We have tools that
help people do that. We have tools to
build your own API-- it's called Apiary--
and other things. And lastly, we have
tools that help you build security
and administration into your platform. So you can say, OK, I have
user data in my SQL table. My entire company
doesn't need to see what's going on in there. So you can manage who can see
what within your platform using the IAM and Admin tools. We also have some
open source tools like TensorFlow, which is--
we have some sessions on that. It's a machine learning
platform, totally open source. Many other companies
use it as well. We have Kubernetes, as
well as other open source tools that help with security
and management, as well. So now I've gone
over, what is Cloud? And I've talked
about what GCP is. So now we're going to talk about
how some companies are actually deploying these
tools and using them to help them spend less
time managing and more time creating more. So if we could just play the
video now, that'd be great. [VIDEO PLAYBACK] - We wanted to make
this marketing campaign for Coca-Cola something that
every single person in this world, no matter who you are,
could participate in the World Cup experience. Submit your photo. You could be on the pitch at the
opening ceremony of the World Cup. - We're talking about 3
and 1/2 million images from over 200 countries. Different formats, different
users, different sources. - Fans could actually
find their photo to know, OK, I was here
on the flag in the opening of the World Cup. - It was a gigantic scale. The agency that was creating the
photo mosaic printed in fabric. We recreated the
equivalent digitally. We did not have time to
set up an infrastructure and set up the servers and
configure the machines. By using Google
Cloud platform, we went straight into
building on top of it. Without that, my
first deliverable would be impossible. - You have to predict what
the loads are going to be. There's a lot of uncertainty. When you look at a cloud
provider like Google, the opportunity there
is for auto scaling. You don't have to worry
about disaster recovery. You don't have to worry
about backup because it's provided to you as a service. - We wanted to
send everyone that had an image in that flag a
notification the very moment the flag was opened
physically in the field. We would have a lot of people
doing that simultaneously. They're not taking
turns to use it. So we were monitoring the
transmission on TV and waiting. I remember clearly seeing
the spikes and just hoping, please, please scale up. Scale up. And it did. - Fundamentally, it gets down
to one thing, and it's one word. It's trust. The same infrastructure that
powers the Google search engine and search experience is
what's powering the cloud that Coca-Cola is running on. - The biggest digital
marketing initiative. - To participate is like a
once in a lifetime opportunity. [END PLAYBACK] ARMAN HEZARKHANI: I heard
someone trying to clap. You can go ahead and clap. All right. One more over here. OK. Anyway. So who's heard of Coca-Cola? Right. So Coca-Cola, one of
the biggest companies on the face of the earth,
put their biggest marketing campaign in history in
the hands of Google. And as you saw, it went
very, very successfully. And so that's just one example. And now we're going
to see another one. So let's see how Spotify
used GCP to enhance the music experience for everybody. Can we roll? [VIDEO PLAYBACK] - The mission at
Spotify is really about being a great
musical companion and making it more personal. It's not going to be the
streaming music experience or the online music experience. It's going to be your
music experience. - We create the tools for
people to actually swim in that sea of
music data and find the information that
really corresponds to why they're there. - Quite while ago, we've made
a decision to focus on data. - Depending on the
complexity of the query, it could take a whole day. Sometimes it would
have to run overnight. - And being able to store
that and process that has been a very big challenge. Moving to Google's
Cloud platform enabled us to take use of the
scale that Google can provide. - The reality is it's quite hard
to build great data centers, and we saw an opportunity to
spend our focus elsewhere. And that was really the driver
into Google's Cloud platform. - With Google Cloud tools,
we can rerun the same query, and it might take a
few seconds, minutes. - The less time
that we can spend solving problems
that are already solved, like scaling
our Hadoop cluster when Google has
already figured out how to scale it for us,
the more time and energy we can spend on turning
our data into value. And that's really a
game changer for us. When you're talking
about a technology like BigQuery, when you're
talking about Dataflow, you will feel the Spotify
product evolve, making the user experience great. - Once you break
down the data, you realize that there are nuances. It's not a straightforward path. It's a mosaic rich beyond
the most popular artists. It's a lot more than that. Music can transcend boundaries
of language, culture. If it moves you, then it's
there, and it's there forever. [MUSIC PLAYING] [END PLAYBACK] ARMAN HEZARKHANI: So Spotify,
another gigantic company that we all use, put their
success in Google's hands. They used Google's
tools to basically get a bunch of their
data from users and provide tools to their
users, like The Daily Mix, or whatever else playlists
that are custom made for you. And if you heard in
the video, basically the CTO said, why would
we try to solve a problem that's already been solved? Google was building the biggest
search engine in the world. And as a matter
of fact, they just built one of the greatest sets
of data centers in the world. So why would another
company put time into doing that
when they could just put more time into creating? And so that's how
companies use GCP. But how are students using GCP? That's what we're
worried about today. So now we're going to
go through some projects that students actually
built, and we're going to see what tools they used. So first, I'm going
to talk about a tool that we found at the
University of Arizona at a hackathon
called Hack Arizona. So a few of us were at the
hackathon, and we were judging, and we saw this
group of students. It was about three students,
and they were freshmen. And they'd just gotten
back from a trip to Europe, and they were telling us about
the horror stories of finding tickets to and from Europe. If anybody's ever tried to find
tickets to Europe as a student, you know how difficult
it is to find tickets that are within our budget. And they basically
wanted to solve this tool using technology. And so they showed up to
this hackathon and they said, how can we
build a tool that will allow students
to find cheap flights to and from Europe? And if you know anything
about searching for flights, you know that there is an
API out there or a few APIs out there where
you can literally scour the internet for every
flight that's out there. It includes where it's going
from and where it's going to. It includes the
price, the times, all the information you need. And so they basically
queried this API, and now they had a gigantic
data set and a very difficult problem. Because parsing this--
basically, it's like a tree. It's like a tree with
nodes of different places where you're going. And parsing this and
trying to find the cheapest flight, that's what's called
the traveling salesman problem, and that is a very, very
computationally heavy problem, as some of you may
know if you study CS. And so they thought, how
are we going to do this? Because no one's ever
solved this problem in less than exponential
time, and so it's very, very computationally heavy. Let's just throw it
on Google Cloud's VMs. They spun up, like, the
biggest and most expensive virtual machine that we had. It had, like, 16
cores or something. It was insane. But they spun it up. And when you go onto
this web application, you don't even notice that
it's doing this gigantic search because of how strong
and powerful Google's infrastructure is. And the code that they wrote
was also pretty efficient. And so that's one example of
a thing that we could all use and we all could have
thought of as well, but it relies on GCP
to make it usable, to make it something that
we would actually use. And it saves us time. Next, I want to talk
about a team that was from Columbia University. They were basically solved
with this problem, which is, how do we allow two
people to communicate if one person is typing and
one person is on the phone? And so the problem
that they very quickly faced was, what
language do these two users speak to each other? And I'm not talking about,
like, spoken language. I'm talking about
one person's writing, the other one's speaking. Literally, how do we
put these two together? And they also found
GCP at this hackathon, and this is what
they came up with. So this is a
diagram of the stack that they built.
And I'm basically going to walk through
what's going on here. So on the top right,
you'll see Twilio. That's another
company that builds an API that allows you to
use cell phones via code, basically. And it spins you
up a phone number, and so you can
call, you can text, and do a bunch of
things from the API. On the left, we
have Hangouts Chat, which is a Google
product, and then we have Messenger, which
is a Facebook product. Those are both chat
messaging apps. So let's start with the right. Let's start with,
how do I allow a user to speak and turn that into
text and send that over to a chat room? So we start with Twilio
sending a request basically to an instance, which is
basically a virtual machine or just, like, a computer
on Google Compute Engine. So they had a server running
on Google Compute Engine that was running a Flask server. So Flask is a Python
framework, and it basically accepts this
request from Twilio. And now they have audio message. They have an audio file
that they then send to Google's speech to text API. This API basically uses
Google's deep learning. It's a fully trained model. All you need to do is basically
call a method in your code that says, almost literally,
turn this speech into text. So it does that, and then
it sends the text back. And then now in this server, we
have text, which is basically just a string, and
then we send that to another server that's
also running a Flask app, but that handles the
text side of this stack. And then once we have
the text over there, we send that up
to the chat rooms, whoever this person
is speaking to. So now using this API and
using the Google Compute Engine instance, we were able to
turn the voice into text. So now let's go the
other way around. So we get text into the
server, and then we send that into the Text to Speech API. And if any of you
saw the presentation that Sundar, our CEO, gave
at, I think it was Google I/O, then you probably
saw the fully-- like, the speaking robot. They were using
this deep learning model, which
basically turns text into very recognizable speech. It sounds like you're
talking to a person. And so we call that API, which
sends us back this audio. We send that audio
to the other server, which sends it to Twilio and
will respond on the phone. And so now we have a full
communication platform between text and speech. So I have a little
check mark there. We just accomplished that. Check. So this team built this in
a hackathon using our tools. And if any of you know about
hackathons, which you do, they're usually under 36 hours. So it means that
these tools are pretty easy to use if
somebody could just get it done in under 36 hours. And so we're very
proud of this team. They built something
really cool. And these were two examples of
something that students built. And it's basically
up to y'all, too, to get these tools into the
hands of students throughout the next year. And we're really excited
to see what other projects we can find from your work. And lastly, how
do I get started? So now we're actually going
to get our hands dirty. Now it's time to take
out your laptops. And as we're going
through this demo, there are going
to be some helpers walking around and basically
answering questions if you have them. And in addition to that,
you each have a coupon code in front of you, which
looks something like this. And on the side of
this coupon code, you'll see a little
yellow bar with a number. And I'll talk about
that a little bit later. So right now we're
going to do four things. First, we're going to log in. Then, we're going
to make a project, and I'll explain
what that means. We're going to create
a billing account, and then we're going
to create a VM. So now I'm going to
log onto my computer, and we're going to
do all these things. So if we could just get
the computer on the screen. Please. OK. Cool. So this is my computer. So first and foremost, I
would like you all to log on to console.cloud.google.com. So I'll type it in with you. console.cloud.google.com. And then just press Enter. And you'll come to a page that
says Access Forbidden that will look a little bit-- I'll just come here. That will look a
little bit like this, but it'll have some
other stuff saying, sign up for a $300 credit. It might say that if
it's your first time logging onto the console. But I've already logged on to
this, so it looks like this. The first thing
we're going to do-- also, if you haven't logged
in, just use a Gmail account. So the first thing
we're going to do is we're going to
say Select a Project. And then now we have all
these different projects. And so what is a project? Let's say I'm a
company or a student, and I'm working at
a hackathon, and I want to build that flight tool. So within that flight tool,
they had a web app running, they had some storage
instances running, and they had a few different
Google Cloud Platform resources. So the way that
we house that all together with the permissions
and everything is in a project. So it's quite intuitive. A project is a project. So to make a project,
we click on New Project. And what it's asking us
for is a project name. You can name this
anything you want. And if you're, like,
incorrect or are not in the correct
format, it'll tell you and then you could change it. But I'm just going to
stay with the default. And then it's going to
say No Organization. That's fine for now. And then Create. So what it's doing right
now is it's basically making my project. It takes a minute because
it needs to propagate through all the systems. But as we wait for
that to happen, again, if you have any
questions, there are people standing around who
can answer your questions, so feel free to raise your
hand if you need it. So now we have our project. As you can see, this
is our dashboard, and right now it
is fairly empty. There is no sign of any
errors for whatever. It's all empty. So as you use it, this
stuff will get filled up with some monitoring
and some tools that will help you see all of
your resources at a glance. But the one thing I'm going
to highlight right now is at the top left
of the screen. So up here, we see Project Info. And I can zoom into this
a couple clicks as well. So we have three
pieces of information. The two most important
are at the top, and the one most important
is right in the middle, and that is your Project ID. So your project name is
not universally unique, but the Project ID is. And so the project ID oftentimes
when you're using GCP tools, you'll be asked for it. And the way that you'll
find that is you'll just click on Google Cloud
Platform, which will bring you back to the Dashboard. You'll make sure that you're
in the correct project, and then you'll just
find it right here. So if you ever need
your project ID, you can just go right there. So next, let's hook up billing. So GCP costs money. So companies pay us money to
basically use these tools. But as students, as
was mentioned before, we have these free
credits that we can use to use these tools
for free to get started. So these coupons each have
a code on the bottom in-- I guess that's
like an orange hue. And so the way that
you enter these is you go to console.cloud.google.com,
same as before, but you do slash edu. So that's
console.cloud.google.com/edu. And so you go to
this page, and you see that there's a little form
specific to education grants. And these are the ones
that we have here. And so what we do is
we enter this code. So I'm going to do
this live right now. So we can all enter
these together. And then once I'm done entering
it, instead of the red, we see this credit amount
expiration date and course. So Michael, where are you? Wave to us. Hey, Michael. Michael's name is on the board. So now we have this coupon code. One quick thing. Oftentimes, if you're at an
event and you have these codes and you're handing
them out, people will say, oh, like, this code is
still red, even though they're done typing. The first thing that they
probably did was they just had a typo. But otherwise, it could be that
they already used the coupon or they already entered it. And so if you ever need
to debug that problem, it's probably one
of those two things. They either mistyped it or
they already used that code. And so for country of residence,
I'm just going to say US. And then please
email me updates. So these are not, like, very
annoying marketing updates. They're actually things
specific to education. So I love education and I'm
going to click Yes, as you all of education. And then I click
Accept and Continue. So what this does
is it creates what's called a billing account. And so the idea of a billing
account versus a project are-- they're kind of two
different things. So every project needs a
billing account hooked up to it, but I can have one
billing account that's hooked up to several projects. So let's say I enter
my credit card. And I have one project for
a side project I'm doing, and then another one
for a class project. I can have that one
billing account tied to these multiple projects. And so the way that you hook
them up to the actual project is you click on this dropdown
menu right at the top, and then you go down to
Manage Billing Accounts. So now we're on the
Manage Billing Accounts. And then up here we see-- I'll zoom in a little. We see My Billing
Accounts and My Projects. Right now we want to
go to the Projects so that we can specify what
billing account we want to use. So we go to My Projects. And over here we see this is
the project that I just spun up. Daring Diode. Yeah. MICHAEL HAMAMOTO: Really quick. If anyone's having
trouble when they're trying to create a new project
and it's saying organization and there's the red
text, what that means, without getting into specifics,
try using your personal Gmail account. Some people's school domains,
like your .edu account, aren't activated to use GCP. And so use your personal
Gmail if that's the case. And then once you've done that,
if you're still having issues, raise your hand, and we'll
come and try and help you. ARMAN HEZARKHANI: Cool. Yeah, and that's also a
very good point, Michael. That's something that a lot
of people at your school will also face. We run into this with
faculty all the time. And so if that is
an issue, that means that your G Suite
administrator, so the IT people at your school,
didn't allow GCP to be used in this organization. So if you just use
your Gmail account, everything will be fine. So we're here,
and right now this is already hooked up
to the billing account because I already had one. But for you to add
a billing account, you click on this
dot dot dot thing, and you click Change
Billing Account, and then you can just pick one. So this says No Billing
Accounts Available because I already have one. But you click on that dot dot
dot, you click Change Billing Account, and then you would
click the actual billing account that you want
to hook it up to, which would be called Michael
Hamamoto or something. And so now we go
back to our project, and we have a billing
account in here, and we have our project set up. So let's actually do
something in here. Let's actually build a resource. So we're going to do what's
called spinning up a VM. So we're just going to
create a virtual machine. And for those of you who don't
know what a virtual machine is, it's basically, like,
an abstraction away from an actual computer. So you all have computers
in front of you, and they all have
CPUs in them that's actually running the code. So we have a bunch
of CPUs at GCP, but we don't sell
those, exactly. We sell what's called a
virtual machine, which is running on this actual CPU. So we can have like three or
four VMs running on one CPU, and they're kind of
separated operating systems. So we're going to do that. First thing we're going to do is
click on the top left hamburger menu. So this is like if you
ever need anything, you just click on this
and it's probably there. This is, like, the answer
to all of my problems ever. That's not exactly
true, but I hope it was. So you're over here. As you can see, a
lot of the tools I spoke about earlier like
IAM, and Admin, and App Engine, and storage, and all these
things, they're all here. So if you ever want to
use one of these tools, you click on the hamburger menu,
and then you scroll down to it and find it there. What we're going to
do now, though, we're going to use Compute
Engine, which shorthand, you'd just say GCE. So you go to Compute Engine,
you just click on that, and then it says Compute
Engine is Getting Ready. So as you can see,
there's a lot of times when GCP is just kind
of propagating data through its systems. And so that's what
it's doing right now. It's just getting ready for us. But once that happens, then we
can just get ready and do that. And again, I see, like,
people are continuing to raise their hands. Keep doing that. Our helpers are running around
and answering questions. Just keep your hand up
if you have a question, and we'll get to you. So we are still
waiting and waiting. But once it gets ready,
we will continue. Anybody have a joke? I guess that was kind of a joke. Nice. We did it. Oh, you have a joke? AUDIENCE: Yeah. ARMAN HEZARKHANI: Go for it. AUDIENCE: Why does
Waldo wear stripes? Because he doesn't
want to be spotted. ARMAN HEZARKHANI: Let's give
that a round of applause. That was really something. OK. So I'm going to refresh
and see if that helped. So it's still getting ready. We have time for one
more joke, I think. Anybody? Oh, we got one over here. Yeah. Stand up for us. What's your name? AUDIENCE: It's more like a joke
that I wouldn't want anyone to clap to. ARMAN HEZARKHANI: OK. You sure you want to say it? AUDIENCE: [INAUDIBLE] ARMAN HEZARKHANI: What was that? AUDIENCE: [INAUDIBLE] ARMAN HEZARKHANI: What? Not going to repeat that one. OK. So this is still getting ready. Let's go to-- do I have
a different project? So is anybody's ready? OK. Cool. It's ready. OK. So whose is ready right now? Awesome. That's, like, a little
bit of the crowd. And then I'm also going to
point you to a tutorial for this later on after this. So if you fall behind, you could
just go through that as well. So basically right now,
we click on Create. And this is basically
the form that will help us create this instance. So the first thing we
need to do is zoom in. But then after we're
done zooming in, we specify the name
of the instance. So I'm just going to
leave this as it is. Instance one-- I could
make this Arman's Instance. So I guess I'm not
going to leave it as-- anyway, it's called
Arman's Instance now. And then now we
select the region, which basically means
where in the world this computer is
actually going to sit. So you can tell it--
if you're in London and all your users
are in London, and you want this computer
to actually be closer to them because of, like,
actual networking speed, then you could spin
it up in London. For now, I am in-- I'm not in South
Carolina, actually. I want to go to US west. So let's go LA. And then you can
specify the actual zone, which is like one level
deeper than region. And then now we click
on the machine type. So we can click
anything from micro, which is like a very, very
small computer, all the way up to, wow, 96 CPUs. That is something. So I'm just going to go
lightweight and click one CPU. And I'm not going to
deploy a container engine. I'm just going to spin up
this operating system, which is a Debian operating system. But also what you can do here
is you can specify a specific, what's called, image-- so as a specific operating
system with certain things installed. So you could specify that and
spin up a VM with all of that already configured. And so for security
courses, for example, a lot of times
teachers will say, hey, go to this virtual machine. It has an old version of
some operating system. Try to break into it. And so if you have courses
like that at your school, you can easily
tell the professor, hey, this is an idea that
you might be able to use. And then the rest of it
I'm just going to skip, but there's also other
settings that you can set, like all down here. And I'm just going
to click Create. So this actually
takes a hot second. So I'm going to
open up a new tab and show you about that tutorial
that I was talking about. So I open a new
tab, and I go to-- what is it-- codelabs.develop-- let's just Google "code labs." So it's
codelabs.developers.google.com. So again,
codelabs.developers.google.com. And so this page is going to
be very, very useful to you. So who are the returning
student innovators in the room? Awesome. How many of you have
used a code lab? Awesome. So that's almost everybody. So code labs are
basically tutorials that step you through how
to use a specific tool and perform a specific task. So in this case, we want to
spin up a virtual machine. So if you type in
Virtual Machine, you can start this code lab. So you can start Create
a Virtual Machine. And what this basically does--
starts with an overview, and then you go to
Setup and Requirements, and it basically walks
you through exactly what you need to do to create a code
lab, including the cleanup. So, like, how to delete it. And so these are really
good for if you're holding a session where
you invite some students and you want to teach them
about a specific tool. I know [? Zishan ?]
has done this a bunch. So I recommend you all
use this tool a lot. And if you fell behind,
you could easily come here and continue with
the code lab on your own. And so with that, the
instance has been created. So we click into that, and
now we have a computer. So we can SSH into it. And what this basically
means is it just opens up a console
where I'm literally writing code in the dock-- I mean, in the virtual machine. And so it connects
SSH keys to the VM. It is connecting. And now I can do anything. So I can make a file
called helloworld.py. I can go in there,
and I can say print-- oh, no. Print hello world. Save that. And then I can say
Python helloworld.py, and it runs the code. So you can do anything
in these computers. You can set up a server
on these computers. You can do literally anything. You can train a model, like
a machine learning model, on there. So they're very, very powerful. And then to close it, you
just click X and leave. And then now let's
clean this up. So we don't want to run out
of our entire $100 credit. So we want to go to the top
right and just click Delete, and that will delete the
instance that we just created. And so that's how we
clean up what we just did. And so I recommend that
you tell all of the people that you're working
with to clean up their projects afterwards
so they're not billed. So that is that. Cool. So that's all for the demo. So if we could just go
back to the slide deck. So now let's just do a
little bit of a recap. So we started out telling the
story of on prem to Cloud. We gave a little bit of
history and kind of tied all the loose ends of
how we actually got here. Then we talked about the three
different umbrellas of Cloud-- Infrastructure as a Service,
Platform as a Service, and Software as a Service. Then we talked about
GCP versus G Suite. So we broke up the
infrastructure and Platform as a Service under
GCP and we said GCE covers the Software
as a Service side. We also talked about
GCP's tools and use cases by going through the different
areas that GCP covers and wrapping up how both
companies and students use these tools in meaningful ways. And lastly, as we
just did, we tried it. We actually opened
up our laptops, we walked through the code, and
we spun up a virtual machine. So that is all for me. I want to thank
you all for coming. I think we have a
little bit of time left, so you all can continue
working and trying to finish up that demo that we just did. But I want to thank you
all for listening to me. [MUSIC PLAYING] SPEAKER: Right on. Thanks, Arman. We appreciate it. All right, everyone,
we are back on time. It's very exciting. So as we move into lunch,
we are eating at the time that we had planned to eat lunch
at, which is very exciting. You can actually leave your
stuff in the room if you want. We have security. Maybe lock your screens. Don't just, like, leave
it totally unlocked, but you can leave your stuff
in the room if you want. You're going to be eating
lunch upstairs, same place you had breakfast. Hold on a minute. You'll be eating with real
life Google engineers. So actually being able
to interact with them and learn from them. There is no assigned
seating, so make sure you sit with people
that look like they're going to be interesting,
which I can assure you they all will be. And lastly, we will see
you in an hour or so. 10 to 1:00 is when
you'll be here. Be safe upstairs
as you eat food. I mean, don't run
each other over. Perfect. Have a good one. [MUSIC PLAYING]
