Skip to main content

Cloud Archive Guide

This page is intentionally aligned with current FolderRewind implementation, not just generic rclone usage.

It focuses on three real product flows:

  • Config Settings -> Cloud tab
  • Config-level "Sync this config from cloud" dialog
  • History page per-item cloud upload/download

Starting in v1.7.0, FolderRewind cloud archive is a full workflow:

  • post-backup auto cloud upload (optional)
  • manual per-history upload/download in History page
  • config-level cloud sync with pre-analysis

1) Core model​

Cloud archive in FolderRewind is a two-layer design:

  1. FolderRewind builds a restorable local backup chain
  2. rclone executes cloud transfer commands

Cloud is an extra redundancy layer, not your primary backup engine. Always validate local backup/restore first.

2) Remote data layout​

Recommended remote base path format:

remote:FolderRewind

Effective cloud structure is:

{RemoteBasePath}/{ConfigName}/{FolderName}/
├─ [Full][...].7z / [Smart][...].7z / ...
└─ _metadata/
	 ├─ state.json
	 └─ records/
			└─ {ArchiveFileName}.json

{RemoteBasePath}/history.json

Notes:

  • archive and metadata are transferred separately
  • history.json is used for cross-device history analysis/import
  • metadata may be partial while archive transfer still succeeds

3) rclone initial setup (OneDrive example)​

Download and install​

  • Download: https://rclone.org/downloads/
  • Typical package: Windows amd64
  • Place rclone.exe in a stable folder (for example C:\Tools\rclone\)
  • Add that folder to PATH (recommended)

Validate:

rclone version

Run rclone config​

rclone config

Typical flow:

  1. n for new remote
  2. Name it (example: fr_onedrive)
  3. Select OneDrive as storage
  4. Keep Client ID/Secret default for standard usage
  5. Keep advanced config default first
  6. Complete OAuth browser sign-in
  7. Select drive type and target drive
  8. Save and quit

Validate remote:

rclone lsd fr_onedrive:

Suggested remote base path to use in FolderRewind:

fr_onedrive:FolderRewind

4) Global Settings prerequisites​

Path: Settings -> Tools and Dependencies -> Environment

  • rclone path: global fallback runtime for cloud operations
  • default cloud remote base path: default for new configs and cloud import/export dialogs

Configure these first, then tune each config.

5) Config Settings -> Cloud tab (actual UI behavior)​

Path: target config -> Config Settings -> Cloud

Environment section​

  • Executable path: config-level override; default can fall back to global rclone path
  • Working directory (optional)
  • Remote base path

Cloud sync section (manual)​

  • button: Sync this config from cloud
  • available in rclone mode

Auto upload switch​

Switch: Enable upload after backup

This controls only post-backup automatic upload queueing.

Manual sync/history cloud actions are governed primarily by mode support (rclone vs legacy custom command).

Advanced upload section (visible when auto upload enabled)​

Templates​

  • Upload current archive (recommended)
  • Upload backup directory
  • Custom

Argument template and variables​

Supported variables:

{ArchiveFilePath}
{ArchiveFileName}
{BackupSubDir}
{MetadataDir}
{ConfigName}
{ConfigId}
{FolderName}
{SourcePath}
{DestinationPath}
{BackupMode}
{Comment}
{Timestamp}
{RemoteBasePath}

Sync history snapshot after upload​

Switch: Sync history after upload

When enabled, FolderRewind uploads latest history.json after successful backup upload.

Timeout/retry/last run status​

  • timeout (seconds)
  • retry count
  • last run time / exit code / error summary

6) Config-level sync dialog details​

Entry: Cloud tab -> Sync this config from cloud

Flow is "analyze first, then sync".

Analysis stage​

FolderRewind downloads remote history.json and resolves safe mappings:

  1. exact match by ConfigId
  2. fallback to remote path prefix ({RemoteBasePath}/{ConfigName})
  3. map folder by:
    • exact FolderPath first
    • unique FolderName fallback
    • ambiguous name matches are skipped

UI shows:

  • matched remote entries
  • importable entries
  • unmapped entries
  • ambiguous entries

Sync scopes​

  • History only: import mapped history entries (merge mode, duplicates skipped)
  • History + backups: import history, then pull archives + metadata for current config folders and recover local history

7) History page cloud integration​

Path: History page

Each history row exposes:

  • Upload to cloud
  • Download from cloud

Status and visual cues​

  • cloud icon appears when cloud copy metadata is available
  • status examples:
    • cloud copy available
    • cloud only copy available
  • cloud-only entries have dedicated timeline color

Button enable rules​

  • Upload enabled when local archive exists
  • Download enabled when cloud copy path exists
  • Legacy custom-command mode disables history-page upload/download (rclone-only)

What upload/download actually does​

Upload:

  1. upload archive first
  2. then upload metadata state/record files
  3. if metadata is partial, archive success is kept and warning is shown

Download:

  1. download archive to local backup storage
  2. then download metadata files
  3. metadata partial warning may still appear

Restore from cloud-only entry​

Recommended sequence:

  1. click download from cloud first
  2. restore after local archive exists

8) Best practices​

  • use "upload current archive" template for most scenarios
  • enable history snapshot sync for multi-device workflows
  • run one end-to-end drill after setup:
    • backup
    • cloud upload
    • cloud download
    • restore verification

9) Troubleshooting​

History cloud buttons disabled​

  • check if config is in legacy custom-command mode
  • upload disabled means local archive is missing
  • download disabled means cloud copy path metadata is missing

Analysis importable count is zero​

  • remote history may not map to current config/folders
  • folder mapping ambiguity may block import
  • remote base path may be wrong

Metadata partial warning​

  • archive transfer succeeded
  • metadata state/record transfer is partial
  • run another sync pass to complete metadata consistency