Google Container Registry (GCR)

Scan GCR registries.

Scan images on Google Container Registry

Prerequisites:
  • GCR access is governed by Google’s storage permissions. For Prisma Cloud to scan GCR, your service account must have the GCP IAM
    Storage View
    role.
  • You must grant Prisma Cloud access to your registry with a service account JSON key file. Your JSON token blob will look something like this:
    { "type": "service_account", "project_id": "my_project_id", "private_key_id": "XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX", "private_key": "-----BEGIN PRIVATE KEY-----\nXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX==\n-----END PRIVATE KEY-----\n", "client_email": "XXXXXXXXXXXXXXX@XXXXXXXXXXXXXX.iam.gserviceaccount.com", "client_id": "XXXXXXXXXXXXXXXXXXXXXXXXX", "auth_uri": "https://accounts.google.com/o/oauth2/auth", "token_uri": "https://oauth2.googleapis.com/token", "auth_provider_x509_cert_url": "https://www.googleapis.com/oauth2/v1/certs", "client_x509_cert_url": "https://www.googleapis.com/robot/v1/metadata/x509/XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX.iam.gserviceaccount.com" }
  1. Open Console, then go to
    Defend > Vulnerabilities > Registry
    .
  2. Click
    Add registry
    .
  3. In the
    Version
    field, select
    Google Container Registry
    .
  4. Enter the registry address in the
    Registry
    field (e.g. gcr.io).
  5. Enter the repository name followed by /* in the
    Repository
    field (e.g. company-sandbox/*).
  6. In the
    Credential
    field, enter the credentials required to access the registry. If the credentials have already been created in the Prisma Cloud credential store, select it. If not, click
    Add
    to create new credentials.
    1. Select the
      GCP
      credential type and credential level, then paste the JSON token blob from your service account into the
      Service Account
      field. Leave the
      API Token
      field blank.
      For GCP organizations with hundreds of projects, scanning GCR using organization level credentials might affect the scanning performance due to long query time from GCP. Therefore, the best approach to reduce scan time and to avoid potential timeouts, is to divide the projects within your organization into multiple GCP folders. Then, create a service account and credential for each one of them, and use these credentials for GCR scanning.
    2. Save your credentials.
  7. In
    OS type
    , specify whether the repo holds
    Linux
    or
    Windows
    images.
  8. In
    Scanners scope
    , specify the collections of defenders to use for the scan.
    Console selects the available Defenders from the scope to execute the scan job according to the
    Number of scanners
    setting. For more information, see deployment patterns.
  9. In
    Number of scanners
    , enter the number of Defenders across which scan jobs can be distributed.
  10. Set
    Cap
    to the number of most recent images to scan.
    Leaving
    Cap
    set to
    5
    will scan the 5 most recent images. Setting this field to
    0
    will scan all images.
  11. Click
    Add
    .
  12. Click the
    Save
    button.

Results

Verify that the images in the repository are being scanned.
  1. Go to
    Monitor > Vulnerabilities > Images > Registries
    .
    A progress indicator at the top right of the window shows the status of the current scan. As the scan of each image is completed, its findings are added to the results table.
  2. To get details about the vulnerabilities in an image, click on it.
    To force a specific repository to be scanned again, select
    Scan
    from the top right of the results table, then click on the specific registry to rescan.