Set up Files with Portabase
Portabase turns Files backups into a productized workflow instead of a pile of forgotten scripts. Rather than scattering cron entries, credentials, and restore notes across different servers, you get one place to schedule jobs, inspect execution history, define retention, control storage destinations, and keep the whole team aligned on what is actually protected. That matters most in real incidents, where a backup process needs to be understandable by more than the person who originally wrote it.
File sources are documented in the Portabase docs. Mount the directory in the agent container, then add a Files source from the dashboard, the CLI or databases.json.
Portabase builds tar.gz archives, incremental restic snapshots or rclone mirrors, and adds scheduling, storage routing, retention and restore orchestration on top. Requires Agent 1.24.0+. Archive and Snapshots sources can be restored; Sync keeps a mirror of the latest state. Best for teams that want their files protected with the same schedules, storages and alerts as their databases.
- 1Install the Portabase dashboard with Docker or Kubernetes.
- 2Deploy a Portabase agent close to your Files environment.
- 3Register the Files database, then choose the schedule, storage destinations, and notifications.
Key features for Files
Three methods, one source type
Archive uploads a full tar.gz per run, Snapshots uploads only what changed with restic deduplication, and Sync keeps a browsable mirror on each storage.
Incremental and encrypted snapshots
Snapshots are deduplicated and always encrypted by restic with a key derived from the agent key, so large directories back up quickly without exposing their content.
Exclude patterns
Leave caches, build output or temporary files out with simple patterns. Excluded paths are kept untouched on restore.
Safe restores
Archives are checked before anything is changed, and a run that cannot read every file fails instead of producing a partial backup.
How Portabase handles Files backups
Operationally, Portabase sits between your team and your Files backup routines to add coordination rather than hide how the database works. The dashboard stores metadata, schedules, execution status, and storage configuration. Agents live close to the target environment and execute backup work where database connectivity already exists. That separation is especially valuable when you need to protect databases inside private networks, customer deployments, or tightly controlled infrastructure where inbound access is not acceptable.
The result is a workflow that scales more cleanly than server-local scripts. You can start with one Files database, then extend the same operating model to more environments, business units, or customer stacks. Retention rules, alerts, and storage destinations stay visible and centralized, which narrows the gap between "we probably have backups" and "we know exactly how backup coverage works."
Compared with managed cloud backup products, Portabase is attractive when you want the operational layer without surrendering control over storage, topology, or recovery habits. The platform stays open source, the storage stays yours, and the workflow stays understandable to your own operators. That combination is often what makes backup practice durable as infrastructure, compliance requirements, or team ownership changes over time.
Why self-hosted Files backup workflows matter
Files often hold the most sensitive data
Documents, uploads and exports stay on infrastructure you control, from the agent to the storage you choose.
One place for files and databases
An application is rarely just a database. Back up its files in the same project, with the same schedules and alerts.
No more cron and rsync scripts
Replace ad hoc scripts with a visible, auditable workflow that reports every run and failure.
Files backup FAQ
Which method should I choose?
Archive for small directories, Snapshots for large directories that change little, and Sync when you want a browsable copy of the latest state. Sync is a mirror, not a history: pair it with Archive or Snapshots.
Can I restore a directory from Portabase?
Yes, for Archive and Snapshots sources. The directory is made identical to the backup, except excluded paths. Sync sources cannot be restored from Portabase.
Does this work with Docker and Kubernetes?
Yes. Mount the directory in the agent container with a bind mount, or with extraVolumes in the Helm chart. The Portabase CLI generates the mount for you.
Where is the setup guide?
Use the Files page in the Portabase docs for the configuration fields, exclude rules and restore behavior.