Skip to content

Latest commit

 

History

History
 
 

gcs-to-bq-with-least-privileges

README.md

Cloud Storage to Bigquery with Cloud Dataflow with least privileges

This example creates the infrastructure needed to run a Cloud Dataflow pipeline to import data from GCS to Bigquery. The example will create different service accounts with least privileges on resources. To run the pipeline, users listed in data_eng_principals can impersonate all those service accounts.

The solution will use:

  • internal IPs for GCE and Cloud Dataflow instances
  • Cloud NAT to let resources egress to the Internet, to run system updates and install packages
  • rely on Service Account Impersonation to avoid the use of service account keys
  • Service Accounts with least privilege on each resource
  • (Optional) CMEK encription for GCS bucket, DataFlow instances and BigQuery tables

The example is designed to match real-world use cases with a minimum amount of resources and some compromises listed below. It can be used as a starting point for more complex scenarios.

This is the high level diagram:

GCS to Biquery High-level diagram

Move to real use case consideration

In the example we implemented some compromise to keep the example minimal and easy to read. On a real word use case, you may evaluate the option to:

  • Configure a Shared-VPC
  • Use only Identity Groups to assigne roles
  • Use Authorative IAM role assignement
  • Split resources in different project: Data Landing, Data Transformation, Data Lake, ...
  • Use VPC-SC to mitigate data exfiltration

Managed resources and services

This sample creates several distinct groups of resources:

  • projects
    • Service Project configured for GCS buckets, Dataflow instances and BigQuery tables and orchestration
  • networking
    • VPC network
    • One subnet
    • Firewall rules for SSH access via IAP and open communication within the VPC
  • IAM
    • One service account for uploading data into the GCS landing bucket
    • One service account for Orchestration
    • One service account for Dataflow instances
    • One service account for Bigquery tables
  • GCS
    • One bucket
  • BQ
    • One dataset
    • One table. Tables are defined in Terraform for the porpuse of the example. Probably, in real scenario, would handle Tables creation in a separate Terraform State or using a different tool/pipeline (for example: Dataform).

In this example you can also configure users or group of user to assign them viewer role on the resources created and the ability to imprsonate service accounts to test dataflow pipelines before autometing them with Composer or any other orchestration systems.

Deploy your enviroment

We assume the identiy running the following steps has the following role:

  • resourcemanager.projectCreator in case a new project will be created.
  • owner on the project in case you use an existing project.

Run Terraform init:

$ terraform init

Configure the Terraform variable in your terraform.tfvars file. You need to spefify at least the following variables:

data_eng_principals = ["user:[email protected]"]
project_id      = "datalake-001"
prefix          = "prefix"

You can run now:

$ terraform apply

You should see the output of the Terraform script with resources created and some command pre-created for you to run the example following steps below.

Virtual Private Cloud (VPC) design

As is often the case in real-world configurations, this example accepts as input an existing Shared-VPC via the network_config variable.

If the network_config variable is not provided, one VPC will be created in each project that supports network resources (load, transformation and orchestration).

When network_config variable is configured, the identity running the Terraform script need to have the following roles:

  • roles/compute.xpnAdmin on the host project folder or org
  • roles/resourcemanager.projectIamAdmin on the host project, either with no conditions or with a condition allowing delegated role grants for roles/compute.networkUser, roles/composer.sharedVpcAgent, roles/container.hostServiceAgentUser

Test your environment with Cloud Dataflow

We assume all those steps are run using a user listed on data_eng_principals. You can authenticate as the user using the following command:

$ gcloud init
$ gcloud auth application-default login

For the purpose of the example we will import from GCS to Bigquery a CSV file with the following structure:

name,surname,timestam

We need to create 3 file:

  • A person.csv file containing your data in the form name,surname,timestam. Here an example line `Lorenzo,Caggioni,1637771951'.
  • A person_udf.js containing the UDF javascript file used by the Dataflow template.
  • A person_schema.json file containing the table schema used to import the CSV.

You can find an example of those file in the folder ./data-demo. You can copy the example files in the GCS bucket using the command returned in the terraform output as command_01_gcs. Below an example:

gsutil -i [email protected] cp data-demo/* gs://LANDING_BUCKET

We can now run the Dataflow pipeline using the gcloud returned in the terraform output as command_02_dataflow. Below an example:

gcloud --impersonate-service-account=orch-test@PROJECT.iam.gserviceaccount.com dataflow jobs run test_batch_01 \
    --gcs-location gs://dataflow-templates/latest/GCS_Text_to_BigQuery \
    --project PROJECT \
    --region REGION \
    --disable-public-ips \
    --subnetwork https://www.googleapis.com/compute/v1/projects/PROJECT/regions/REGION/subnetworks/subnet \
    --staging-location gs://PREFIX-df-tmp \
    --service-account-email [email protected] \
    --parameters \
javascriptTextTransformFunctionName=transform,\
JSONPath=gs://PREFIX-data/person_schema.json,\
javascriptTextTransformGcsPath=gs://PREFIX-data/person_udf.js,\
inputFilePattern=gs://PREFIX-data/person.csv,\
outputTable=PROJECT:datalake.person,\
bigQueryLoadingTemporaryDirectory=gs://PREFIX-df-tmp

You can check data imported into Google BigQuery using the command returned in the terraform output as command_03_bq. Below an example:

bq query --use_legacy_sql=false 'SELECT * FROM `PROJECT.datalake.person` LIMIT 1000'

Variables

name description type required default
prefix Unique prefix used for resource names. Not used for project if 'project_create' is null. string ✓
project_id Project id, references existing project if project_create is null. string ✓
cmek_encryption Flag to enable CMEK on GCP resources created. bool false
data_eng_principals Groups with Service Account Token creator role on service accounts in IAM format, eg 'group:[email protected]'. list(string) []
network_config Shared VPC network configurations to use. If null networks will be created in projects with preconfigured values. object({…}) null
project_create Provide values if project creation is needed, uses existing project if null. Parent is in 'folders/nnn' or 'organizations/nnn' format. object({…}) null
region The region where resources will be deployed. string "europe-west1"
vpc_subnet_range Ip range used for the VPC subnet created for the example. string "10.0.0.0/20"

Outputs

name description sensitive
bq_tables Bigquery Tables.
buckets GCS bucket Cloud KMS crypto keys.
command_01_gcs gcloud command to copy data into the created bucket impersonating the service account.
command_02_dataflow Command to run Dataflow template impersonating the service account.
command_03_bq BigQuery command to query imported data.
project_id Project id.
service_accounts Service account.