Using Amazon Web Services (AWS) for Labs
Atefeh Sohrabizadeh, Eddie Huang
Acknowledgment
We would like to thank Yuze Chi and Peng Wei for creating these slides.
Outline
• Signup for an AWS Account
• Signup for AWS Educate Credit
•Create an AWS EC2 instance
Outline
• Signup for an AWS Account
• Signup for AWS Educate Credit
•Create an AWS EC2 instance
Create an AWS Account• Or login to it if you have already signed up for it
• Use this link for AWS Account Signup
• Unfortunately, your credit card info is needed, so be very careful against using up your $100 credit and being billed on your credit card!
Signup or Login
Outline
• Signup for an AWS Account
• Signup for AWS Educate Credit
•Create an AWS EC2 instance
AWS Educate
•$100 in AWS credits per student OR
•An AWS Educate Starter Account with a $100 in AWS credits• Cannot be used for the Amazon EC2 instances we’ll use
•Only students from member universities get the $100 credits, so please signup AWS Educate with your @ucla.edu email
•Access to AWS Technical Essentials Training Course
•Use this link for AWS Educate signup or search “AWS Educate” on Google to find it by yourself
Join AWS Educate
•Or login to it if you have already signed up for it
Signup or Login
Join AWS Educate
Signup as a Student
Join AWS Educate
Use your @ucla.edu email, since students from member universities get much more credits, do NOT use @g.ucla.edu or @cs.ucla.edu or anything else, or you will NOT receive the correct amount of credits
Write the complete name of the university (without comma) to get the full 100$ credit“University of California Los Angeles” IMPORTANT: do NOT enter "UCLA"
Join AWS Educate•Be really careful with this step!
• If selecting an AWS Educate Starter Account, you will only get $100 credits, and NO ACCESS to F1!
It’s not there anymore
You will probably only see this one
Join AWS Educate• An email will be sent to your @ucla.edu address for verification• And then, the application will be reviewed
• And finally, an approval email with the $100 promotion code will be sent to your @ucla.edu email
• Follow the instruction to redeem your credit
Outline
• Signup for an AWS Account
• Signup for an AWS Education Account
•Create an AWS EC2 instance
Create an AWS EC2 instance
•Once your account is active, head over to the AWS console, and click on EC2.
Under “Services”, find and click EC2
Select AMI
• We’ll use different AMIs for different purposes
• For CPU and GPU we use Ubuntu 18.04
• For FPGA we use FPGA Developer AMI or Merlin AMI
Select AMI
Select the Instance• We’ll use different instances
for different purposes
• For CPU we use m5.2xlarge (Use this for now)
• For GPU we use g3s.xlarge
• For FPGA development we use m5.2xlarge
• For FPGA deployment we use f1.2xlarge
Create and keep your key pair
Create a key pair and use it for SSH
Using key pair to log into instance
Type in your linux console:
ssh -i <your_keypair_file.pem> ubuntu@<ip address>
Once you SSH…• Check the price of the instance your are using. keep it
on only when you are working on the labs.•When you need to take a break, make sure you stop
the instance!• You’ll be charged for storage, which is not expensive but not
0• You can start the instance at any time
•When you’re done for the lab, make SURE you terminate the instance!• You’ll not be charged for this instance anymore• Your data will be erased
• For Ubuntu the default username is ‘ubuntu’• FPGA AMI uses CentOS, the default username is
‘centos’.
Accounts
•By now you will have two different accounts• One is your own account• The other is the AWS educate classroom account
• To access this account, log out of your own account first
• Click AWS Console
• You’ll be redirected
• This account is linkedto this class
• Use this account first!
Instance Limitations
•By default you may not be able to launch instances
•Use your educate account to get started• with m5.2xlarge
•Request limit increase for your own account ASAP• Mention that you are registered for CS 133 at UCLA and
we’ll use the following instances (subject to change)• c5.4xlarge for CPU
• g3s.large for GPU
• f1.2xlarge for FPGA testing
Use GitHub for CS133
•Register a GitHub account if you haven’t
•Use git and GitHub’s private repos to manage your code• Avoid data loss & ease code migration
•We will provide detailed instructions in lab specs