Cloud Sync Service¶

The cloud sync service replicates mail storage and backups to cloud object storage -- AWS S3 or Azure Blob Storage. It provides an offsite copy of your data for disaster recovery.

What It Does¶

Syncs mail storage (/var/mail/vhosts/) to cloud storage
Syncs backup files to cloud storage
Supports AWS S3 and Azure Blob Storage
Incremental sync -- only uploads changed files
Full resync capability for disaster recovery
Webhook notifications on sync events (started, completed, failed)

How It Works¶

flowchart LR
    subgraph CloudSync["Cloud Sync Service"]
        Scheduler["Sync\nScheduler"]
        SyncEngine["Sync\nEngine"]
        Webhook["Webhook\nSender"]
    end

    Scheduler --> SyncEngine
    SyncEngine -->|"read"| MailDir["/var/mail/vhosts/"]
    SyncEngine -->|"read"| Backups["/backups/"]
    SyncEngine -->|"upload"| S3["AWS S3"]
    SyncEngine -->|"upload"| Azure["Azure Blob"]
    SyncEngine --> Webhook

API Endpoints¶

Text Only

POST /api/sync/start            -- Trigger a manual sync
POST /api/sync/full-resync      -- Full resync (re-upload everything)
GET  /api/sync/status           -- Current sync status
GET  /api/sync/history          -- Sync history
GET  /health                    -- Health check

Configuration¶

Variable	Default	Description
`CLOUD_PROVIDER`	`s3`	Cloud provider (`s3` or `azure`)
`AWS_ACCESS_KEY_ID`	(required for S3)	AWS access key
`AWS_SECRET_ACCESS_KEY`	(required for S3)	AWS secret key
`AWS_S3_BUCKET`	(required for S3)	S3 bucket name
`AWS_S3_REGION`	`us-east-1`	S3 region
`AZURE_STORAGE_ACCOUNT`	(required for Azure)	Azure storage account
`AZURE_STORAGE_KEY`	(required for Azure)	Azure storage key
`AZURE_CONTAINER`	(required for Azure)	Azure blob container
`SYNC_INTERVAL`	`86400`	Seconds between syncs (24 hours)
`SYNC_PATHS`	`/var/mail/vhosts,/backups`	Comma-separated paths to sync
`WEBHOOK_URLS`	(empty)	Webhook URLs for sync notifications

Docker Configuration¶

YAML

cloud_sync:
  build: ./worker/cloud_sync
  container_name: cloud_sync
  volumes:
    - mail_data:/var/mail/vhosts:ro
    - backup_data:/backups:ro

Gotchas¶

Bandwidth and Cost

Initial sync of a large mail store can use significant bandwidth and incur cloud storage costs. The first sync uploads everything; subsequent syncs only upload changes.

Credentials

Cloud storage credentials are sensitive. Use environment variables or Docker secrets -- never hardcode them in config files.

S3 Lifecycle Policies

Configure S3 lifecycle policies to automatically move old backups to Glacier for cost savings. This is managed on the AWS side, not in Mailyte.