Skip to content

Latest commit

 

History

History
343 lines (287 loc) · 27.2 KB

README.md

File metadata and controls

343 lines (287 loc) · 27.2 KB
Raw

Secured Data Warehouse Blueprint

FAQ | Troubleshooting Guide.

This repository contains Terraform configuration modules that allow Google Cloud customers to quickly deploy a secured BigQuery data warehouse, following the Secure a BigQuery data warehouse that stores confidential data guide. The blueprint allows customers to use Google Cloud's core strengths in data analytics, and to overcome typical challenges that include:

  • Limited knowledge/experience with best practices for creating, deploying, and operating in Google Cloud.
  • Security/risk concerns and restrictions from their internal security, risk, and compliance teams.
  • Regulatory and compliance approval from external auditors.

The Terraform configurations in this repository provide customers with an opinionated architecture that incorporates and documents best practices for a performant and scalable design, combined with security by default for control, logging and evidence generation. It can be simply deployed by customers through a Terraform workflow.

Disclaimer

When using this blueprint, it is important to understand how you manage separation of duties. We recommend you remove all primitive owner roles in the projects used as inputs for the Data Warehouse module. The secured data warehouse itself does not need any primitive owner roles for correct operations.

When using this blueprint in the example mode or when using this blueprint to create the new projects with default configurations for the deployment, we automatically remove the owner role as it has too broad access.

However, if you choose to use this blueprint with pre-existing projects in your organization, we will not proactively remove any pre-existing owner role assignments, as we won’t know your intent for or dependency on these role assignments in your pre-existing workloads. The pre-existing presence of these roles does expand the attack and risk surface of the resulting deployment. Therefore, we highly recommend you review your use of owner roles in these pre-existing cases and see if you can eliminate them to improve your resulting security posture. Only you can determine the appropriate trade-off to meet your business requirements.

You can check the current situation of your project with either of the following methods:

See the terraform-example-foundation for additional good practices.

Usage

Basic usage of this module is as follows:

module "secured_data_warehouse" {
  source  = "terraform-google-modules/secured-data-warehouse/google"
  version = "~> 0.1"

  org_id                           = ORG_ID
  data_governance_project_id       = DATA_GOVERNANCE_PROJECT_ID
  confidential_data_project_id     = CONFIDENTIAL_DATA_PROJECT_ID
  non_confidential_data_project_id = NON_CONFIDENTIAL_DATA_PROJECT_ID
  data_ingestion_project_id        = DATA_INGESTION_PROJECT_ID
  sdx_project_number               = EXTERNAL_TEMPLATE_PROJECT_NUMBER
  terraform_service_account        = TERRAFORM_SERVICE_ACCOUNT
  access_context_manager_policy_id = ACCESS_CONTEXT_MANAGER_POLICY_ID
  bucket_name                      = DATA_INGESTION_BUCKET_NAME
  pubsub_resource_location         = PUBSUB_RESOURCE_LOCATION
  location                         = LOCATION
  trusted_locations                = TRUSTED_LOCATIONS
  dataset_id                       = DATASET_ID
  confidential_dataset_id          = CONFIDENTIAL_DATASET_ID
  cmek_keyring_name                = CMEK_KEYRING_NAME
  perimeter_additional_members     = PERIMETER_ADDITIONAL_MEMBERS
  data_engineer_group              = DATA_ENGINEER_GROUP
  data_analyst_group               = DATA_ANALYST_GROUP
  security_analyst_group           = SECURITY_ANALYST_GROUP
  network_administrator_group      = NETWORK_ADMINISTRATOR_GROUP
  security_administrator_group     = SECURITY_ADMINISTRATOR_GROUP
  delete_contents_on_destroy       = false
}

Note: There are three inputs related to GCP Locations in the module:

  • pubsub_resource_location: is used to define which GCP location will be used to Restrict Pub/Sub resource locations. This policy offers a way to ensure that messages published to a topic are never persisted outside of a Google Cloud regions you specify, regardless of where the publish requests originate. Zones or multi-region locations are not supported.
  • location: is used to define which GCP region will be used for all other resources created: Cloud Storage buckets, BigQuery datasets, and Cloud KMS key rings. Multi-region locations are supported.
  • trusted_locations: is a list of locations that are used to set an Organization Policy that restricts the GCP locations that can be used in the projects of the Secured Data Warehouse. Both pubsub_resource_location and location must respect this restriction.

Inputs

Name Description Type Default Required
access_context_manager_policy_id The id of the default Access Context Manager policy. Can be obtained by running gcloud access-context-manager policies list --organization YOUR-ORGANIZATION_ID --format="value(name)". string "" no
additional_restricted_services The list of additional Google services to be protected by the VPC-SC service perimeters. list(string) [] no
bucket_class The storage class for the bucket being provisioned. string "STANDARD" no
bucket_lifecycle_rules List of lifecycle rules to configure. Format is the same as described in provider documentation https://www.terraform.io/docs/providers/google/r/storage_bucket.html#lifecycle_rule except condition.matches_storage_class should be a comma delimited string. 
set(object({
    action    = any
    condition = any
  }))
 
[
  {
    "action": {
      "type": "Delete"
    },
    "condition": {
      "age": 30,
      "matches_storage_class": [
        "STANDARD"
      ],
      "with_state": "ANY"
    }
  }
]
 no
bucket_name The name of the bucket being provisioned. string n/a yes
cmek_keyring_name The Keyring prefix name for the KMS Customer Managed Encryption Keys being provisioned. string n/a yes
confidential_data_dataflow_deployer_identities List of members in the standard GCP form: user:{email}, serviceAccount:{email} that will deploy Dataflow jobs in the Confidential Data project. These identities will be added to the VPC-SC secure data exchange egress rules. list(string) [] no
confidential_data_egress_policies A list of all egress policies for the Confidential Data perimeter, each list object has a from and to value that describes egress_from and egress_to. See also secure data exchange and the VPC-SC module. 
list(object({
    from = any
    to   = any
  }))
 [] no
confidential_data_perimeter Existing confidential data perimeter to be used instead of the auto-created perimeter. The service account provided in the variable terraform_service_account must be in an access level member list for this perimeter before this perimeter can be used in this module. string "" no
confidential_data_project_id Project where the confidential datasets and tables are created. string n/a yes
confidential_dataset_id Unique ID for the confidential dataset being provisioned. string "secured_dataset" no
data_analyst_group Google Cloud IAM group that analyzes the data in the warehouse. string n/a yes
data_engineer_group Google Cloud IAM group that sets up and maintains the data pipeline and warehouse. string n/a yes
data_governance_egress_policies A list of all egress policies for the Data Governance perimeter, each list object has a from and to value that describes egress_from and egress_to. See also secure data exchange and the VPC-SC module. 
list(object({
    from = any
    to   = any
  }))
 [] no
data_governance_perimeter Existing data governance perimeter to be used instead of the auto-created perimeter. The service account provided in the variable terraform_service_account must be in an access level member list for this perimeter before this perimeter can be used in this module. string "" no
data_governance_project_id The ID of the project in which the data governance resources will be created. string n/a yes
data_ingestion_dataflow_deployer_identities List of members in the standard GCP form: user:{email}, serviceAccount:{email} that will deploy Dataflow jobs in the Data Ingestion project. These identities will be added to the VPC-SC secure data exchange egress rules. list(string) [] no
data_ingestion_egress_policies A list of all egress policies for the Data Ingestion perimeter, each list object has a from and to value that describes egress_from and egress_to. See also secure data exchange and the VPC-SC module. 
list(object({
    from = any
    to   = any
  }))
 [] no
data_ingestion_perimeter Existing data ingestion perimeter to be used instead of the auto-created perimeter. The service account provided in the variable terraform_service_account must be in an access level member list for this perimeter before this perimeter can be used in this module. string "" no
data_ingestion_project_id The ID of the project in which the data ingestion resources will be created string n/a yes
dataset_default_table_expiration_ms TTL of tables using the dataset in MS. The default value is null. number null no
dataset_description Dataset description. string "Data-ingestion dataset" no
dataset_id Unique ID for the dataset being provisioned. string n/a yes
dataset_name Friendly name for the dataset being provisioned. string "Data-ingestion dataset" no
delete_contents_on_destroy (Optional) If set to true, delete all the tables in the dataset when destroying the resource; otherwise, destroying the resource will fail if tables are present. bool false no
enable_bigquery_read_roles_in_data_ingestion (Optional) If set to true, it will grant to the dataflow controller service account created in the data ingestion project the necessary roles to read from a bigquery table. bool false no
key_rotation_period_seconds Rotation period for keys. The default value is 30 days. string "2592000s" no
kms_key_protection_level The protection level to use when creating a key. Possible values: ["SOFTWARE", "HSM"] string "HSM" no
labels (Optional) Labels attached to Data Warehouse resources. map(string) {} no
location The location for the KMS Customer Managed Encryption Keys, Cloud Storage Buckets, and Bigquery datasets. This location can be a multi-region. string "us-east4" no
network_administrator_group Google Cloud IAM group that reviews network configuration. Typically, this includes members of the networking team. string n/a yes
non_confidential_data_project_id The ID of the project in which the Bigquery will be created. string n/a yes
org_id GCP Organization ID. string n/a yes
perimeter_additional_members The list additional members to be added on perimeter access. Prefix user: (user:[email protected]) or serviceAccount: (serviceAccount:[email protected]) is required. list(string) [] no
pubsub_resource_location The location in which the messages published to Pub/Sub will be persisted. This location cannot be a multi-region. string "us-east4" no
remove_owner_role (Optional) If set to true, remove all owner roles in all projects in case it has been found in some project. bool false no
sdx_project_number The Project Number to configure Secure data exchange with egress rule for dataflow templates. Required if using a dataflow job template from a private storage bucket outside of the perimeter. string "" no
security_administrator_group Google Cloud IAM group that administers security configurations in the organization(org policies, KMS, VPC service perimeter). string n/a yes
security_analyst_group Google Cloud IAM group that monitors and responds to security incidents. string n/a yes
terraform_service_account The email address of the service account that will run the Terraform code. string n/a yes
trusted_locations This is a list of trusted regions where location-based GCP resources can be created. list(string) 
[
  "us-locations"
]
 no
trusted_subnetworks The URI of the subnetworks where resources are going to be deployed. list(string) [] no

Outputs

Name Description
blueprint_type Type of blueprint this module represents.
cmek_bigquery_crypto_key The Customer Managed Crypto Key for the BigQuery service.
cmek_confidential_bigquery_crypto_key The Customer Managed Crypto Key for the confidential BigQuery service.
cmek_data_ingestion_crypto_key The Customer Managed Crypto Key for the data ingestion crypto boundary.
cmek_reidentification_crypto_key The Customer Managed Crypto Key for the Confidential crypto boundary.
confidential_access_level_name Access context manager access level name.
confidential_data_dataflow_bucket_name The name of the bucket created for dataflow in the confidential data pipeline.
confidential_dataflow_controller_service_account_email The confidential Dataflow controller service account email. See https://cloud.google.com/dataflow/docs/concepts/security-and-permissions#specifying_a_user-managed_controller_service_account.
confidential_service_perimeter_name Access context manager service perimeter name.
data_governance_access_level_name Access context manager access level name.
data_governance_service_perimeter_name Access context manager service perimeter name.
data_ingestion_access_level_name Access context manager access level name.
data_ingestion_bigquery_dataset The bigquery dataset created for data ingestion pipeline.
data_ingestion_bucket_name The name of the bucket created for the data ingestion pipeline.
data_ingestion_dataflow_bucket_name The name of the bucket created for dataflow in the data ingestion pipeline.
data_ingestion_service_perimeter_name Access context manager service perimeter name.
data_ingestion_topic_name The topic created for data ingestion pipeline.
dataflow_controller_service_account_email The Dataflow controller service account email. See https://cloud.google.com/dataflow/docs/concepts/security-and-permissions#specifying_a_user-managed_controller_service_account.
pubsub_writer_service_account_email The PubSub writer service account email. Should be used to write data to the PubSub topics the data ingestion pipeline reads from.
scheduler_service_account_email The Cloud Scheduler service account email, no roles granted.
storage_writer_service_account_email The Storage writer service account email. Should be used to write data to the buckets the data ingestion pipeline reads from.

Requirements

These sections describe requirements for using this module.

Note: Please see the Disclaimer regarding project owners before creating projects.

Software

Install the following dependencies:

Security Groups

Provide the following groups for separation of duty. Each group is granted roles to perform their tasks. Then, add users to the appropriate groups as needed.

  • Data Engineer group: Google Cloud IAM group that sets up and maintains the data pipeline and warehouse.
  • Data Analyst group: Google Cloud IAM group that analyzes the data in the warehouse.
  • Security Analyst group: Google Cloud IAM group that monitors and responds to security incidents.
  • Network Administrator group: Google Cloud IAM group that reviews network configuration. Typically, this includes members of the networking team.
  • Security Administrator group: Google Cloud IAM group that administers security configurations in the organization(org policies, KMS, VPC service perimeter).

Groups can be created in the Google Workspace Admin Console, in the Google Cloud Console, and using gcloud identity groups create.

Service Account

To provision the resources of this module, create a privileged service account, where the service account key cannot be created. In addition, consider using Cloud Monitoring to alert on this service account's activity. Grant the following roles to the service account.

  • Organization level
    • Access Context Manager Admin: roles/accesscontextmanager.policyAdmin
    • Organization Policy Administrator: roles/orgpolicy.policyAdmin
    • Organization Administrator: roles/resourcemanager.organizationAdmin
  • Project level:
    • Data ingestion project
      • App Engine Creator:roles/appengine.appCreator
      • Cloud Scheduler Admin:roles/cloudscheduler.admin
      • Compute Network Admin:roles/compute.networkAdmin
      • Compute Security Admin:roles/compute.securityAdmin
      • Dataflow Developer:roles/dataflow.developer
      • DNS Administrator:roles/dns.admin
      • Project IAM Admin:roles/resourcemanager.projectIamAdmin
      • Pub/Sub Admin:roles/pubsub.admin
      • Service Account Admin:roles/iam.serviceAccountAdmin
      • Service Account Token Creator:roles/iam.serviceAccountTokenCreator
      • Service Usage Admin: roles/serviceusage.serviceUsageAdmin
      • Storage Admin:roles/storage.admin
    • Data governance project
      • Cloud KMS Admin:roles/cloudkms.admin
      • Cloud KMS CryptoKey Encrypter:roles/cloudkms.cryptoKeyEncrypter
      • DLP De-identify Templates Editor:roles/dlp.deidentifyTemplatesEditor
      • DLP Inspect Templates Editor:roles/dlp.inspectTemplatesEditor
      • DLP User:roles/dlp.user
      • Data Catalog Admin:roles/datacatalog.admin
      • Project IAM Admin:roles/resourcemanager.projectIamAdmin
      • Secret Manager Admin: roles/secretmanager.admin
      • Service Account Admin:roles/iam.serviceAccountAdmin
      • Service Account Token Creator:roles/iam.serviceAccountTokenCreator
      • Service Usage Admin: roles/serviceusage.serviceUsageAdmin
      • Storage Admin:roles/storage.admin
    • Non Confidential project
      • BigQuery Admin:roles/bigquery.admin
      • Project IAM Admin:roles/resourcemanager.projectIamAdmin
      • Service Account Admin:roles/iam.serviceAccountAdmin
      • Service Account Token Creator:roles/iam.serviceAccountTokenCreator
      • Service Usage Admin: roles/serviceusage.serviceUsageAdmin
      • Storage Admin:roles/storage.admin
    • Confidential project
      • BigQuery Admin:roles/bigquery.admin
      • Compute Network Admin:roles/compute.networkAdmin
      • Compute Security Admin:roles/compute.securityAdmin
      • DNS Administrator:roles/dns.admin
      • Dataflow Developer:roles/dataflow.developer
      • Project IAM Admin:roles/resourcemanager.projectIamAdmin
      • Service Account Admin:roles/iam.serviceAccountAdmin
      • Service Account Token Creator:roles/iam.serviceAccountTokenCreator
      • Service Usage Admin: roles/serviceusage.serviceUsageAdmin
      • Storage Admin:roles/storage.admin

You can use the Project Factory module and the IAM module in combination to provision a service account with the necessary roles applied.

The user using this service account must have the necessary roles to impersonate the service account.

APIs

Create four projects with the following APIs enabled to host the resources of this module:

Data ingestion project

  • Access Context Manager API: accesscontextmanager.googleapis.com
  • App Engine Admin API:appengine.googleapis.com
  • Artifact Registry API:artifactregistry.googleapis.com
  • BigQuery API:bigquery.googleapis.com
  • Cloud Billing API:cloudbilling.googleapis.com
  • Cloud Build API:cloudbuild.googleapis.com
  • Cloud Key Management Service (KMS) API:cloudkms.googleapis.com
  • Cloud Resource Manager API:cloudresourcemanager.googleapis.com
  • Cloud Scheduler API:cloudscheduler.googleapis.com
  • Compute Engine API:compute.googleapis.com
  • Google Cloud Data Catalog API:datacatalog.googleapis.com
  • Dataflow API:dataflow.googleapis.com
  • Cloud Data Loss Prevention (DLP) API:dlp.googleapis.com
  • Cloud DNS API:dns.googleapis.com
  • Identity and Access Management (IAM) API:iam.googleapis.com
  • Cloud Pub/Sub API:pubsub.googleapis.com
  • Service Usage API:serviceusage.googleapis.com
  • Google Cloud Storage JSON API:storage-api.googleapis.com

Data governance project

  • Access Context Manager API: accesscontextmanager.googleapis.com
  • Cloud Billing API:cloudbilling.googleapis.com
  • Cloud Key Management Service (KMS) API:cloudkms.googleapis.com
  • Cloud Resource Manager API:cloudresourcemanager.googleapis.com
  • Google Cloud Data Catalog API:datacatalog.googleapis.com
  • Cloud Data Loss Prevention (DLP) API:dlp.googleapis.com
  • Identity and Access Management (IAM) API:iam.googleapis.com
  • Service Usage API:serviceusage.googleapis.com
  • Google Cloud Storage JSON API:storage-api.googleapis.com
  • Secrect Manager API: secretmanager.googleapis.com

Non-confidential data project

  • Access Context Manager API: accesscontextmanager.googleapis.com
  • BigQuery API:bigquery.googleapis.com
  • Cloud Billing API:cloudbilling.googleapis.com
  • Cloud Key Management Service (KMS) API:cloudkms.googleapis.com
  • Cloud Resource Manager API:cloudresourcemanager.googleapis.com
  • Identity and Access Management (IAM) API:iam.googleapis.com
  • Service Usage API:serviceusage.googleapis.com
  • Google Cloud Storage JSON API:storage-api.googleapis.com

Confidential data project

  • Access Context Manager API: accesscontextmanager.googleapis.com
  • Artifact Registry API:artifactregistry.googleapis.com
  • BigQuery API:bigquery.googleapis.com
  • Cloud Billing API:cloudbilling.googleapis.com
  • Cloud Build API:cloudbuild.googleapis.com
  • Cloud Key Management Service (KMS) API:cloudkms.googleapis.com
  • Cloud Resource Manager API:cloudresourcemanager.googleapis.com
  • Compute Engine API:compute.googleapis.com
  • Google Cloud Data Catalog API:datacatalog.googleapis.com
  • Dataflow API:dataflow.googleapis.com
  • Cloud Data Loss Prevention (DLP) API:dlp.googleapis.com
  • Cloud DNS API:dns.googleapis.com
  • Identity and Access Management (IAM) API:iam.googleapis.com
  • Service Usage API:serviceusage.googleapis.com
  • Google Cloud Storage JSON API:storage-api.googleapis.com

The following APIs must be enabled in the project where the service account was created

  • Access Context Manager API: accesscontextmanager.googleapis.com
  • App Engine Admin API: appengine.googleapis.com
  • Cloud Billing API:cloudbilling.googleapis.com
  • Cloud Key Management Service (KMS) API:cloudkms.googleapis.com
  • Cloud Pub/Sub API: pubsub.googleapis.com
  • Cloud Resource Manager API:cloudresourcemanager.googleapis.com
  • Compute Engine API:compute.googleapis.com
  • Dataflow API:dataflow.googleapis.com
  • Identity and Access Management (IAM) API:iam.googleapis.com

You can use the Project Factory module to provision the projects with the necessary APIs enabled.

Security Disclosures

Please see our security disclosure process.

Contributing

Refer to the contribution guidelines for information on contributing to this module.

This is not an officially supported Google product