Issue

The data extraction phase is scheduled to commence in early April and will conclude by late June , running concurrently with the five migration waves to ensure all necessary data is captured prior to final decommissioning.

• Perform Google Vault data exports for the elements deleted by active and archived users from December 8, 2023 and store them into the same GCP bucket containing the data exported from Solvay on the separation date. (Mail, Drive, Chat, Anywhere retention rules apply)
• Deleted user data :
  • audit The data shared with active users and accessed within the last 18 months.
  • Transfer the ownership of these documents to the first user that has access to the document in question.
• Talarian AwesomeTable tool auditing : Extract an inventory of Awesome Table. (Linked source spreadsheets, Owners of each spreadsheet, Configuration metadata (Connectors, Apps, filters, templates, etc.))

To perform all of those activities Syensqo will rely on Revevol who will use Google Application Manager (GAM).

Details

• GAM is a windows tool, cannot run on linux server.
• Do not need to connect the windows server to the active directory, while the server will remains in GCP boundaries and identity management.
• GAM will Download from API → Server Disk → GCP (keep the data until will leave GCP)
• Strategy to extract the data out of GCP to be defined.
• Operating model of data being extracted to be clarified (with security team)
• GCP projects → Syensqo Projects and API Lists

Recommendation

Go for Option 1 reusing one of the existing GCP project running GAM today for daily operation with a new dedicated machine completely delegated to Dawn and Revevol. Delete the data and or the machine a the end of the migration in June or July 2026.

Options Considered

Option 1: GCP same project : reuse existing project with an additional machine to not impact the run activities, while the data export will be huge. (today the delegation are managed manually)

• • No extension of the scope, same place.
  • Hosting will grant permission to Dawn to provide delegation and create machine.
  • Dawn will delegate the machine creation to revevol to become T0 (inform Rizwan Parwaiz)
  • Revevol creates the necessary stuff on the machine to perform the SOW activities.
  • Dawn to ensure post migration , Revevol delete the machine, or the disk being cleaned-up. (do we need this machine with GAM later?)
  • Post Vault extraction Vault operating model for 5y to be reviewed with security.
  • Existing projects for 2 technical team Indian and french team
    • Gcp-sqo-gam-in-p
    • Gcp-sqo-gam-fr-p ← (the one to be reused)

Option 2: GCP but with a new project completely managed by REVEVOL : replicate current GCP project flows etc, Dawn having the access, and sharing appropriate permission with Revevol only. But do not make sense to segregate the project while we can have a dedicated machine in one of the existing project.

Option 3: AWS : too long, and extra security concerns

Option 4: AZURE : too long, and extra security concerns, Azure is not yet configured for T0

Option 5: Revevol : if we can handle on our GCP , better to avoid.

See also

The following section describes relevant documentation: