DEPLOYDETAILED.md

Building Docker Image, Putting on ECR, Forcing a Deployment

This verbose and manual document shows show exactly how our CD pipeline works but gives more context by retrieving the AWS inputs from aws-cli. It also can be used to run deployments from local setup. Take note that we are using grep below to whittle down which project and environment we are targeting from all the potential output:

0. Install jq because it's awesome: https://formulae.brew.sh/formula/jq

1. Make sure you've built your local docker branch and it's up to date with any branch changes

   $ docker build -t veda-wfs3-api:latest .

2. Export some os env vars so we can use them filter. Make sure they match the environment you want to work against

    $ export TARGET_PROJECT_NAME=veda-wfs3
    $ export TARGET_ENVIRONMENT=dev

3. Make sure you have an AWS_PROFILE setup that matches the AWS region you want to work with. In the examples below uah2 referes to the UAH account in us-west-2

4. List existing ECR repositories using "aws-cli" and whittle down which one we want to talk to with os env vars:

   $ AWS_PROFILE=uah2 aws ecr describe-repositories 
   {
       "repositories": [
           {
               "repositoryArn": "arn:aws:ecr:us-west-2:359356595137:repository/veda-wfs3-registry-dev",
               "registryId": "359356595137",
               "repositoryName": "veda-wfs3-registry-dev",
               "repositoryUri": "359356595137.dkr.ecr.us-west-2.amazonaws.com/veda-wfs3-registry-dev",
               "createdAt": "2022-12-10T13:46:05-08:00",
               "imageTagMutability": "MUTABLE",
               "imageScanningConfiguration": {
                   "scanOnPush": false
               },
               "encryptionConfiguration": {
                   "encryptionType": "AES256"
               }
           }
       ]
   }
   
   $ AWS_PROFILE=uah2 aws ecr describe-repositories \
       | jq '.repositories | map(.repositoryUri)' \
       | grep $TARGET_PROJECT_NAME | grep $TARGET_ENVIRONMENT
   "359356595137.dkr.ecr.us-west-2.amazonaws.com/veda-wfs3-registry-dev"

5. Login to ECR from awscli:

   $ AWS_PROFILE=uah2 aws ecr describe-repositories \
      | jq '.repositories | map(.repositoryUri)' \
      | grep $TARGET_PROJECT_NAME | grep $TARGET_ENVIRONMENT \
      | xargs -I {} bash -c "AWS_PROFILE=uah2 aws ecr get-login-password | docker login --username AWS --password-stdin {}"

6. Now re-tag the local image we built with the remote ECR repository and tag name:

    $ AWS_PROFILE=uah2 aws ecr describe-repositories \
       | jq '.repositories | map(.repositoryUri)' \
       | grep $TARGET_PROJECT_NAME | grep $TARGET_ENVIRONMENT \
       | xargs -I {} docker images --format "{{json . }}" {} \
       | grep '"Tag":"latest"' \
       | jq '"\(.Repository):\(.Tag)"' \
       | xargs -I{} docker tag veda-wfs3-api:latest {}
   
   # check your work locally
    $ AWS_PROFILE=uah2 aws ecr describe-repositories \
       | jq '.repositories | map(.repositoryUri)' \
       | grep $TARGET_PROJECT_NAME | grep $TARGET_ENVIRONMENT \
       | xargs -I {} docker images --format "{{json . }}" {} \
       | grep '"Tag":"latest"' \
       | jq '"\(.Repository):\(.Tag)"' \
       | jq
   {
     "Containers": "N/A",
     "CreatedAt": "2022-12-12 08:16:23 -0800 PST",
     "CreatedSince": "9 minutes ago",
     "Digest": "<none>",
     "ID": "a0a6c57e40e8",
     "Repository": "359356595137.dkr.ecr.us-west-2.amazonaws.com/veda-wfs3-registry-dev",
     "SharedSize": "N/A",
     "Size": "887MB",
     "Tag": "latest",
     "UniqueSize": "N/A",
     "VirtualSize": "887.2MB"
   }

7. Push the image from local to ECR:

     $ AWS_PROFILE=uah2 aws ecr describe-repositories \
       | jq '.repositories | map(.repositoryUri)' \
       | grep $TARGET_PROJECT_NAME | grep $TARGET_ENVIRONMENT \
       | xargs -I {} docker images --format "{{json . }}" {} \
       | grep '"Tag":"latest"' \
       | jq '"\(.Repository):\(.Tag)"' \
       | xargs -I{} docker push {}
   
   # check your remote work
     $ AWS_PROFILE=uah2 aws ecr describe-repositories \
       | jq '.repositories | map(.repositoryUri)' \
       | grep $TARGET_PROJECT_NAME | grep $TARGET_ENVIRONMENT \
       | AWS_PROFILE=uah2 xargs -I {} aws ecr describe-images --repository-name={}
   {
       "imageDetails": [
           {
               "registryId": "359356595137",
               "repositoryName": "veda-wfs3-registry-dev",
               "imageDigest": "sha256:bf83dd6027aadbf190347529a317966656d875a2aa8b64bbd2cc2589466b68e7",
               "imageTags": [
                   "latest"
               ],
               "imageSizeInBytes": 325163652,
               "imagePushedAt": "2022-12-12T08:35:14-08:00",
               "imageManifestMediaType": "application/vnd.docker.distribution.manifest.v2+json",
               "artifactMediaType": "application/vnd.docker.container.image.v1+json"
           }
       ]
   }

8. Show your existing clusters:

   $ AWS_PROFILE=uah2 aws ecs list-clusters                      
   {
       "clusterArns": [
           "arn:aws:ecs:us-west-2:359356595137:cluster/tf-veda-wfs3-service-dev"
       ]
   }
   
   $ AWS_PROFILE=uah2 aws ecs list-clusters \
     | jq '.clusterArns[0]' \
     | xargs -I{} aws ecs describe-clusters --cluster={}
   {
       "clusters": [
           {
               "clusterArn": "arn:aws:ecs:us-west-2:359356595137:cluster/tf-veda-wfs3-service-dev",
               "clusterName": "tf-veda-wfs3-service-dev",
               "status": "ACTIVE",
               "registeredContainerInstancesCount": 0,
               "runningTasksCount": 0,
               "pendingTasksCount": 0,
               "activeServicesCount": 1,
               "statistics": [],
               "tags": [],
               "settings": [],
               "capacityProviders": [],
               "defaultCapacityProviderStrategy": []
           }
       ],
       "failures": []
   }

9. Once it's there, we can force update the ECS cluster/service/tasks to use it with:

   $ AWS_PROFILE=uah2 aws ecs list-clusters \
    | jq '.clusterArns[0]' \
    | grep $TARGET_PROJECT_NAME | grep $TARGET_ENVIRONMENT \
    | AWS_PROFILE=uah2 xargs -I{}  aws ecs describe-clusters --cluster={} \
    | jq '.clusters[0].clusterName' \
    | AWS_PROFILE=uah2 xargs -I{}  aws ecs update-service --cluster {} --service {} --task-definition {} --force-new-deployment > /dev/null