Plese see the latest BentoML documentation on OCI-container based deployment workflow: https://docs.bentoml.com/

Bentoctl is a CLI tool for deploying your machine-learning models to any cloud platform and serving predictions via REST APIs. It is built on top of BentoML: the unified model serving framework and makes it easy to bring any BentoML packaged model to production.

This repo contains the AWS EC2 deployment operator. This operator defines the terraform configuration for deploying a bento into an EC2 instance.

Note: This operator is compatible with BentoML version 1.0.0 and above. For older versions, please switch to the branch pre-v1.0 and follow the instructions in the README.md.

• Quickstart with bentoctl
• Configuration options
• Troubleshooting

This quickstart will walk you through deploying a bento into an EC2 instance. Make sure to go through the prerequisites section and follow the instructions to set everything up.

1. Bentoml - BentoML version 1.0 and greater. Please follow the Installation guide.
2. Terraform - Terraform is a tool for building, configuring, and managing infrastructure. Installation instruction: www.terraform.io/downloads
3. AWS CLI - installed and configured with an AWS account with permission to EC2 and ECR. Please follow the Installation guide.
4. Docker - Install instruction: docs.docker.com/install
5. A built Bento project. For this guide, we will use the Iris classifier bento from the BentoML quickstart guide. You can also use your own Bentos that are available locally.

1. Install bentoctl via pip

   $ pip install bentoctl
   

2. Install AWS EC2 operator

   Bentoctl will install the official AWS EC2 operator and its dependencies.

   $ bentoctl operator install aws-ec2
   

3. Initialize deployment with bentoctl

   Follow the interactive guide to initialize the deployment project.

   $ bentoctl init
   
   Bentoctl Interactive Deployment Config Builder
   
   Welcome! You are now in interactive mode.
   
   This mode will help you setup the deployment_config.yaml file required for
   deployment. Fill out the appropriate values for the fields.
   
   (deployment config will be saved to: ./deployment_config.yaml)
   
   api_version: v1
   name: iris-classifier-ec2
   operator: aws-ec2
   template: terraform
   spec:
       region: ap-south-1
       instance_type: t2.micro
       ami_id: ami-0a3277ffce9146b74
       enable_gpus: False
       environment_variables:
   filename for deployment_config [deployment_config.yaml]:
   deployment config generated to: deployment_config.yaml
   ✨ generated template files.
     - ./main.tf
     - ./bentoctl.tfvars

   This will also run the bentoctl generate command for you and will generate the main.tf terraform file, which specifies the resources to be created and the bentoctl.tfvars file which contains the values for the variables used in the main.tf file.

4. Build and push AWS EC2 compatible docker image to the registry

   Bentoctl will build and push the EC2 compatible docker image to the AWS ECR repository.

   bentoctl build -b iris_classifier:latest -f deployment_config.yaml
   
   Step 1/22 : FROM bentoml/bento-server:1.0.0a6-python3.8-debian-runtime
    ---> 046bc2e28220
   Step 2/22 : ARG UID=1034
    ---> Using cache
    ---> f44cfa910c52
   Step 3/22 : ARG GID=1034
    ---> Using cache
    ---> e4d5aed007af
   Step 4/22 : RUN groupadd -g $GID -o bentoml && useradd -m -u $UID -g $GID -o -r bentoml
    ---> Using cache
    ---> fa8ddcfa15cf
   ...
   Step 22/22 : CMD ["bentoml", "serve", ".", "--production"]
    ---> Running in 28eccee2f650
    ---> 98bc66e49cd9
   Successfully built 98bc66e49cd9
   Successfully tagged aws-ec2-iris_classifier:kiouq7wmi2gmockr
   🔨 Image build!
   Created the repository iris-classifier-ec2
   The push refers to repository
   [213386773652.dkr.ecr.ap-south-1.amazonaws.com/iris-classifier-ec2]
   kiouq7wmi2gmockr: digest:
   sha256:e1a468e6b9ceeed65b52d0ee2eac9e3cd1a57074eb94db9c263be60e4db98881 size: 3250
   63984d77b4da: Pushed
   2bc5eef20c91: Pushed
   ...
   da0af9cdde98: Layer already exists
   e5baccb54724: Layer already exists
   🚀 Image pushed!
   ✨ generated template files.
     - ./bentoctl.tfvars
     - ./startup_script.sh

   The iris-classifier service is now built and pushed into the container registry and the required terraform files have been created. Now we can use terraform to perform the deployment.

5. Apply Deployment with Terraform

   1. Initialize terraform project. This installs the AWS provider and sets up the terraform folders.

      $ terraform init

   2. Apply terraform project to create EC2 deployment

      $ terraform apply -var-file=bentoctl.tfvars -auto-approve
      
      aws_iam_role.ec2_role: Creating...
      aws_security_group.allow_bentoml: Creating...
      aws_security_group.allow_bentoml: Creation complete after 2s [id=sg-01d5baaa464ff58f9]
      aws_iam_role.ec2_role: Creation complete after 3s [id=iris-classifier-ec2-iam]
      aws_iam_instance_profile.ip: Creating...
      aws_iam_instance_profile.ip: Creation complete after 1s [id=iris-classifier-ec2-instance-profile]
      aws_launch_template.lt: Creating...
      aws_launch_template.lt: Creation complete after 0s [id=lt-09d7717f0f1a56001]
      aws_instance.app_server: Creating...
      aws_instance.app_server: Still creating... [10s elapsed]
      aws_instance.app_server: Still creating... [20s elapsed]
      aws_instance.app_server: Still creating... [30s elapsed]
      aws_instance.app_server: Still creating... [40s elapsed]
      aws_instance.app_server: Creation complete after 43s [id=i-0d9767b74865dc0b0]
      
      Apply complete! Resources: 5 added, 0 changed, 0 destroyed.
      
      Outputs:
      
      ec2_instance_status = "running"
      ec2_ip_address = "13.235.76.37"

6. Test deployed endpoint

   The iris_classifier uses the /classify endpoint for receiving requests so the full URL for the classifier will be in the form {EndpointUrl}/classify.

   URL=$(terraform output -json | jq -r .ec2_ip_address.value)/classify
   curl -i \
     --header "Content-Type: application/json" \
     --request POST \
     --data '[5.1, 3.5, 1.4, 0.2]' \
     $URL
   
   HTTP/2 200
   date: Thu, 14 Apr 2022 23:02:45 GMT
   content-type: application/json
   content-length: 1
   apigw-requestid: Ql8zbicdSK4EM5g=
   
   0%

   Please note that the EC2 instance might take some more time to pull the image and setup the bentoml server. You can check if the server is up by pinging the /livez endpoint.

   URL=$(terraform output -json | jq -r .ec2_ip_address.value)/livez
   curl $URL

   If it doesn't start up after a few minutes refer to the troubleshooting section

7. Delete deployment Use the bentoctl destroy command to remove the registry and the deployment

   bentoctl destroy -f deployment_config.yaml

• region: AWS region for deployment
• instance_type: Instance type for the EC2 deployment. See https://aws.amazon.com/ec2/instance-types/ for the entire list.
• ami_id: Amazon Machine Image (AMI) used for the EC2 instance. Check out https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/AMIs.html for more information. The list of available AMIs for your region is available at https://console.aws.amazon.com/ec2/home#AMICatalog. AMI Ids are region specific, so make sure you get the AMI Id for the region you want to deploy into.

Note: Only Amazon Linux AMIs are supported at the current moment. This is the limitation of the startup_script used.

• ennable_gpus: If using GPU-accelerated instance_types then ennable this.
• environment_variables: List of environment variables that should be passed to the instance.

To run any troubleshooting we will have to connect to the EC2 instance. The EC2 instance created with terraform has EC2-connect configured by default but you will have to open the SSH port to connect to it.

To open the SSH port, open the main.tf that has been generated and add an additional ingress rule into the aws_security_group resource named allow_bentoml.

resource "aws_security_group" "allow_bentoml" {
  name        = "${var.deployment_name}-bentoml-sg"
  description = "SG for bentoml server"

  ingress {
    description      = "HTTP for bentoml"
    from_port        = 80
    to_port          = 80
    protocol         = "tcp"
    cidr_blocks      = ["0.0.0.0/0"]
    ipv6_cidr_blocks = ["::/0"]
  }

  # add this to open SSH port for the EC2 instance.
  ingress {
    description      = "SSH access for server"
    from_port        = 22
    to_port          = 22
    protocol         = "tcp"
    cidr_blocks      = ["0.0.0.0/0"]
    ipv6_cidr_blocks = ["::/0"]
  }

  egress {
    from_port        = 0
    to_port          = 0
    protocol         = "-1"
    cidr_blocks      = ["0.0.0.0/0"]
    ipv6_cidr_blocks = ["::/0"]
  }
}

Now you can run terraform apply again to make the changes to the resources.

terraform apply -var-file=bentoctl.tfvars -auto-approve

You can connect to an instance using the Amazon EC2 console (browser-based client) by selecting the instance from the console and choosing to connect using EC2 Instance Connect. Instance Connect handles the permissions and provides a successful connection.

To connect to your instance using the browser-based client from the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.
2. In the navigation pane, choose Instances.
3. Select the instance and choose Connect.
4. Choose EC2 Instance Connect.
5. Verify the user name and choose Connect to open a terminal window.

For more information check out the official docs

Once connected, make sure the docker container is running by running

docker ps

This will show all the containers running. Ideally, you should see something like

CONTAINER ID   IMAGE                              COMMAND                  CREATED         STATUS        PORTS                                   NAMES
4681b23e0c51   iris_classifier:kiouq7wmi2gmockr   "./env/docker/entryp…"   2 seconds ago   Up 1 second   0.0.0.0:80->3000/tcp, :::80->3000/tcp   bold_kirch

To view the logs from the container, run

docker logs <NAMES>

and it will output the logs from the container.

If the docker container is not running or if the docker command is not available in the ec2 instance then it could be an issue with the initialisation script. You can check the logs for the init-script by running

sudo cat /var/log/cloud-init-output.log