Skip to main content
Version: v2.4.1

Data Recovery

ReVISit stores study and participant data in Firebase, which is backed by Google Cloud services. Over time, you may need to restore an accidentally deleted file, or review the history of changes to a stored object. This page describes how to recover data and view history of changes to a file using Firebase and Google Cloud Storage.

warning

By default, Object Versioning is disabled for Firebase storage buckets. Without it, there is still a one-week soft delete recovery window in which data can be restored.

reVISit data stored with Firebase lives in Google Cloud services:

  • Study and participant records are stored in Cloud Firestore.
  • Files (for example, audio and screen recordings) are stored in Cloud Storage for Firebase, which uses a Google Cloud Storage bucket.

Viewing Version History for a Single Stored Object​

Object version history is managed through Google Cloud Storage, not directly in the Firebase UI.

Every noncurrent object version is stored and billed until removed (manually or with lifecycle rules). For some studies in reVISit, such as studies that store large amounts of provenance data, this may significantly increase costs.

  1. Enable Object Versioning on your bucket:
    • Open the Google Cloud Console.
    • Go to Cloud Storage -> Buckets -> your Firebase bucket.
    • Open the Protection tab and enable Object versioning.
  2. View versions for one object:
    • In the bucket's Objects tab, change Show to Live and noncurrent objects.
    • Click the object name.
    • Open the Version history tab to see each generation.
  3. (Optional) Restore an older version from that same version history view.

How Long Version Histories Last​

  • By default (without Object Versioning), deleted files are recoverable for one week (7 days) via the bucket's soft delete policy.
  • With Object Versioning enabled, noncurrent versions are retained until you delete them.
  • There is no default limit on the number of versions kept.
  • If you want automatic cleanup, add Object Lifecycle rules (for example, delete noncurrent versions after N days).
  • Noncurrent versions are billed as stored Cloud Storage data, so turning on version history will increase storage costs.