BYOC Onboarding

BYOC onboarding

note

This page applies to BYOC deployments. To compare the plans, see Deployment models.

You create and manage your BYOC deployment in the SubImage management portal, manage.subimage.io. The portal is where you connect your cloud account, create your deployment, and follow its status.

Before you start

  • A dedicated cloud account. Use an AWS account or a Google Cloud project that holds only the SubImage deployment.
  • A region. The portal lists the regions available for your plan.
  • A hostname. Pick the hostname your users will open, for example subimage.example.com. You need to create a DNS record for it.
  • Compute quota. A deployment needs more compute than a new cloud account allows by default. On AWS, ask for at least 128 vCPUs of the "Running On-Demand Standard instances" quota in your region before you start. On Google Cloud, check the CPU quota of the region.
  • An administrator who can create IAM roles in the account, or IAM service accounts in the project.

1. Subscribe to a plan

Sign in to manage.subimage.io and subscribe to the BYOC or BYOC with Private Graph plan. SubImage approves the subscription.

2. Connect your cloud account

In the portal, add your AWS account or Google Cloud project. The portal gives you an account configuration and a setup method. Choose one of the methods below. Each method creates the same access, described in Access and security.

AWS

  • Terraform. Download subimage_byoc_aws_onboarding.tf. Set config_identity to the AccountConfigIdentityId value of the CloudFormation link that the portal shows. Then run terraform init && terraform apply with credentials for the account.
  • CloudFormation. Open the quick-create link that the portal shows. Before you create the stack, set the IsBYOCPrivateAccount and EnablePrivateArtifactRegistry parameters to true.

Both methods create the AccountConfigSetup CloudFormation stack.

Google Cloud

First, create the identity pool that Google Cloud adds to a project when a GKE cluster first exists in it. A temporary GKE Autopilot cluster is enough:

gcloud container clusters create-auto temp-pool --project=<project> --region=us-central1
gcloud container clusters delete temp-pool --project=<project> --region=us-central1 --quiet

Then enable the Service Usage API:

gcloud services enable serviceusage.googleapis.com --project=<project>

Then choose a method:

  • Setup script. Run the command that the portal shows, in Cloud Shell or with the gcloud CLI.
  • Terraform. Download subimage_byoc_gcp_onboarding.tf. Set project_id to your project and account_config_id to the Account Config ID that the portal shows. Then run terraform init && terraform apply.

Wait until the portal shows the account as ready.

3. Create your deployment

In the portal, create a deployment on the connected account. Fill in these fields:

Field What to enter
Region The region for the deployment.
Deployment ID A short lowercase name for the deployment, for example acme. You cannot change it later.
Organization display name Your company name, shown on the sign-in page.
Tenant hostname The hostname you picked, without https:// or a path. You cannot change it later.
Graph memory and Graph storage BYOC with Private Graph only. See Operations.

SubImage then creates the infrastructure and installs the application. The portal shows the progress.

4. Point your hostname to the deployment

When the deployment is ready, the portal shows the DNS record to create for your hostname. Create it at your DNS provider. SubImage then issues the TLS certificate for the hostname.

5. Sign in and connect your data

Open your hostname and sign in. Then connect your sources: see Getting started and the module setup guides.

On this page
Before you start1. Subscribe to a plan2. Connect your cloud accountAWSGoogle Cloud3. Create your deployment4. Point your hostname to the deployment5. Sign in and connect your data