Inception process for the GCP Climate Analytics & Machine Learning platform. This repo creates basic GCP resources and a pipeline trigger, using a shell script and Terraform code. The rest of the platform is created from the downstream analytics repo triggering the pipeline.
- terraform >= 0.12
- bash
- GCP user account with Organization Admin privileges
Configure _start_vars.sh (see also _start_vars.sh.example):
INCEPTION_PROJECT : Inception project id
STATE_BUCKET : "${INCEPTION_PROJECT}-terraform-state" - Name of inception process terraform state bucket
REGION : Inception project region, eg. europe-west2
BILLING_ACCOUNT : Organisation billing account id
Configure _tf_vars.sh (see also _tf_vars.sh.example):
TF_VAR_org_id : $(gcloud organizations list | tail -1 | awk '{ print $2 }') - GCP Organisation id
GOOGLE_PROJECT : "${INCEPTION_PROJECT}" - Inception project id
TF_VAR_orchestration_project_number : $(gcloud projects list | awk -v p="${INCEPTION_PROJECT}" '($1 == p) { print $3 }') - Inception project number
TF_VAR_analytics_project : Analytics project id
TF_VAR_region : "${REGION}" - Inception project region
TF_VAR_billing_account : "${BILLING_ACCOUNT}" - Organisation billing account id
TF_VAR_owner : Github account name, eg. thundercomb
TF_VAR_owner_email : Project owner email
TF_VAR_kubeflow_host : Kubeflow host id (Only available after Kubeflow pipeline creation)
TF_VAR_inception_ip: $(curl ifconfig.me)
Run the start script.
bash start.sh