trabianmatt / fintech_devcon

zsuite workshop

fintech_devcon

Environment Setup

Step 1 - Create Amazon Web Services (AWS) Account

1. Navigate to AWS Event Page
2. Enter the code 54be-13a09159a4-77
3. Choose how to sign in (AWS Recommends Email One Time Password)
4. Enter the one time password

Step 2 - Create Cloud9 Instance

1. TBC

Step 3 - Configure Cloud9

We'll run a script in Cloud9 to perform the following:

• Removes unnecessary software (PHP, etc)
• Install kubectl Kubernetes CLI
• Pull down and fork helm charts repo (https://github.com/ZSuiteTech/fintech_devcon_charts)

git clone ZSuiteTech/fintech_devcon
cd fintech_devcon
./setup.sh

Instructions:

• Follow instructions in console (copy one-time code) 😊
• Click on link and select "Open"
• Switch to tab/window and follow instructions

Example Output:

! First copy your one-time code: ....-....
- Press Enter to open github.com in your browser... 
! Failed opening a web browser at https://github.com/login/device
  exec: "xdg-open,x-www-browser,www-browser,wslview": executable file not found in $PATH
  Please try entering the URL in your browser manually
  
✓ Authentication complete. Press Enter to continue...

Reference

1. Terraform

Workshop

Step 1: Create EKS cluster using Terraform

1. Create s3 bucket from command prompt 1.

   YOUR_BUCKET_NAME=zstfdemo-ftdc21-${GH_USER,,}`

   2. aws s3 mb s3://$YOUR_BUCKET_NAME

2. Initialize Terraform

   terraform init -backend-config="bucket=$YOUR_BUCKET_NAME"

3. Run Terraform, passing in the variable file

   terraform apply -var-file=demo.tfvars

4. Approve changes by typing yes and pressing Enter.

5. Creating resources will take over 10 minutes.

Step 2: Verify EKS deployment with kubectl

1. Setup kubectl

   export KUBECONFIG=./kubeconfig_zsuite-devcon-<RANDOM>

2. Test command via

   kubectl get namespace

Step 3: Deploy ArgoCD with terraform

1. Replace repoUrl in argocd/manifests/apps.yaml with your forked Helm charts repository

   gh repo view --json url --jq '.url'

2. Uncomment argocd module usage in main.tf

3. Run terraform init again to install new dependency

4. Run terraform apply again

   terraform apply -var-file=demo.tfvars

5. Confirm argocd namespace exists cluster namespaces

6. Check ArgoCD services

   kubectl -n argocd get all

7. Log in to UI

   1. Get argo password 1.

      kubectl -n argocd get secret argocd-initial-admin-secret -o jsonpath="{.data.password}" | base64 -d

      2. Do not include the last character % that's a newline

   2. Get argo-server pod

      kubectl -n argocd get pods -l app.kubernetes.io/name=argocd-server

   3. Port forward to pod 1.

      ARGO_POD=argocd-server-?????????-?????

      2. kubectl -n argocd port-forward pod/$ARGO_POD 8080:8080

   4. NOPE: In browser go to localhost:8080

   5. Login with username admin and password from above

Step 4: Add app to ArgoCD

1. Go to forked helm charts repo

2. Edit apps/values.yaml

   1. Set deployWatchman to true
   2. Commit and push changes

3. Confirm watchman resources exist

   kubectl -n moov get all

4. Test watchman api

   1. Get pod

      kubectl -n moov get pods -l app.kubernetes.io/instance=my-watchman

   2. Port forward to pod 1.

      MOOV_POD=my-watchman-?????????-?????

      2. kubectl -n moov port-forward pod/$MOOV_POD 8081:8080

   3. Test endpoint http://localhost:8081/downloads

      curl -uri http://localhost:8081/downloads

Step 5: Enable ingress for Watchman API

1. Install AWS Load Balancer controller

   1. Uncomment alb_ingress_controller module in main.tf

   2. Run

      terraform init

      again to install new dependency

   3. Run terraform apply again

      terraform apply -var-file=demo.tfvars

2. In helm charts repo, set deployWatchmanIngress variable to true

3. Commit and push changes for helm charts repo

4. Aws load balancer provisioning can take several minutes

   1. Can view load balancer in aws console

5. Get dns entry for watchman api from argocd

   1. Open watchman-gateway app
   2. Select ingress
   3. Go to live manifest
   4. Find dns address

6. Test api from public address 1.

   MOOV_DNS=??????.us-east-1.elb.amazonaws.com

   2. curl -uri http://$MOOV_DNS:8081/downloads

Step 6: destroy

1. Delete argo apps

   1. Disable variables in helm charts repo
      1. deployWatchman
      2. deployWatchmanIngress
   2. This cleans up ELB nicely

2. Run terraform destroy 1.

   terraform destroy -var-file=demo.tfvars

   2. Confirm by typing yes
   3. This will take around 4 minutes