> ## Content Index > Fetch the complete content index at: https://www.blog.nestict.com/llms.txt > Use this file to discover other available public pages before exploring further. # How to Fully Restore Docker Volumes From Backup on Ubuntu (Complete Guide With rsync, Permission Fixes & Safety Checks) - URL: https://www.blog.nestict.com/how-to-fully-restore-docker-volumes-from-backup-on-ubuntu-complete-guide-with-rsync-permission-fixes-safety-checks/ - Published: 2025-11-21T17:07:18.000Z - Updated: 2025-11-21T17:07:18.000Z - Author: Nestict Infotech CSR - Tags: Technology Restoring a Docker server after system migration, disk failure, or a panel crash (such as Webuzo or a VPS reinstatement) requires careful handling — especially when your applications depend on persistent data stored in Docker volumes. In this guide, we walk through the **exact steps** used to restore multiple Docker volumes (Vaultwarden, Odoo, Keycloak, Chatwoot, and more) from a backup directory such as: ``` /mnt/backup/docker-backup/var/lib/docker/volumes/ ``` This article is based on a real-world server recovery scenario involving dozens of volumes and containers. --- ## **✔️ Prerequisites** You must have: - Docker installed on the target machine - Root access to the system - Enough disk space A backup directory containing your Docker volumes Example structure: ``` /mnt/backup/docker-backup/var/lib/docker/volumes//_data ``` --- # **1\. Verify the Backup Directory Structure** Before restoring, ensure your backup actually contains the Docker volumes: ```bash ls /mnt/backup/docker-backup/var/lib/docker/volumes/ ``` Typical output: ``` devops_lipanet_chatwoot_data nestict_vaultwarden_data nestict_keycloak_data nestict_ghost_content nestict_odoo-db_odoo-db-data ... ``` If you see folders ending with `_data`, you’re good. --- # **2\. Stop Docker (VERY Important)** To avoid database corruption (especially for PostgreSQL, SQLite, MariaDB, Redis): ```bash systemctl stop docker ``` This freezes all containers and releases volume locks. --- # **3\. Restore ALL Docker Volumes Using rsync** To restore volumes from backup to the live Docker storage: ```bash rsync -avh --progress \ /mnt/backup/docker-backup/var/lib/docker/volumes/ \ /var/lib/docker/volumes/ ``` ### What this does: - `-a` → archive mode (preserves permissions, owners, links) - `-v` → verbose - `-h` → human-readable - `--progress` → show file transfer progress - Trailing `/` → ensures correct directory merging ### Typical rsync output: ``` sending incremental file list ./ metadata.db 4812738e2ea9c598847bc06f322e1879.../_data/jvb.conf ... sent 271.62K bytes received 235 bytes ``` You might notice **very low transferred data size** — this is normal. Rsync only copies *changed* files. --- # **4\. Fix Permissions for Sensitive Containers** Docker containers rely on *specific UIDs/GIDs* to access their volumes. ## **Vaultwarden (Rust / UID 1000)** ```bash chown -R 1000:1000 /var/lib/docker/volumes/nestict_vaultwarden_data/_data ``` ## **Odoo (UID 1000)** ```bash chown -R 1000:1000 /var/lib/docker/volumes/*odoo*/* 2>/dev/null || true ``` ## **PostgreSQL (Keycloak, Chatwoot, Odoo DB) – UID 999** ```bash chown -R 999:999 /var/lib/docker/volumes/*postgres*/* 2>/dev/null || true ``` ## **Chatwoot** No special permissions required unless logs show errors. --- # **5\. Start Docker Again** ```bash systemctl start docker ``` Check running containers: ```bash docker ps ``` --- # **6\. Restart All Containers** ## Method 1 — Restart all running containers: ```bash docker restart $(docker ps -q) ``` ## Method 2 — If using docker-compose: ```bash cd /path/to/compose/project docker compose restart ``` For example (Zammad installation): ```bash cd /home/nestict/git/zammad-docker-compose docker compose restart ``` --- # **7\. Validate Each Service After Restore** Check logs for errors: ### Vaultwarden ```bash docker logs vaultwarden ``` ### Keycloak ```bash docker logs ``` ### Odoo ```bash docker logs ``` ### Chatwoot ```bash docker logs devops_lipanet_chatwoot_app ``` If any container fails to start, logs will show missing permissions, DB migration issues, or wrong file paths. --- # **8\. Common Issues and Fixes** ## **❌ “permission denied” errors after restore** → Wrong UID on volume. Fix using `chown -R :`. ## **❌ Vaultwarden won’t start (SQLite locked)** Delete WAL file: ```bash rm /var/lib/docker/volumes/nestict_vaultwarden_data/_data/db.sqlite3-wal ``` ## **❌ Keycloak fails due to Postgres permissions** Run: ```bash chown -R 999:999 /var/lib/docker/volumes/*postgres*/_data ``` --- # **9\. Best Practices for Docker Backup & Restore** ### ✔️ Always stop docker before restoring volumes ### ✔️ Never restore volumes while containers are running ### ✔️ Use rsync instead of cp for data consistency ### ✔️ Verify permission ownership per service ### ✔️ Test services individually after restore ### ✔️ Keep offsite backups (S3, Backblaze, rsync.net) ### ✔️ Use `docker volume ls` to confirm restored volumes --- # **Conclusion** Restoring Docker volumes safely is critical when handling production systems hosting multiple services like Vaultwarden, Odoo, Keycloak, Zammad, Nextcloud, or Chatwoot. Using `rsync` ensures accurate, atomic file transfer while preserving structure and permissions. With the method shown in this guide, you can recover **an entire multi-container environment** quickly and reliably.