# Docker Volumes: Keep, Back Up, and Restore Your Data
URL: https://ctrlops.io/docs/modules/docker/volumes
Published: 2026-11-05
Updated: 2026-11-14

The Volumes tab lists the storage your containers use, with its size and who uses it. Create and delete volumes, open their files, schedule backups to S3 or to the same server, and restore a volume from any backup, even on another server.

A **volume** is where your data lives, so it stays when a container is removed. The **Volumes** tab shows every volume on this server, and how to keep that data safe.

> **New to Docker?**
> 
> A container loses the files inside it when you delete it. Put data you want to keep, such as a database or uploads, in a *volume*. A volume is storage that Docker manages, and it stays until you delete it yourself.

![The Docker Volumes tab with the All 8, In use 4, and Unused 4 filters, the Search volumes box and the Remove unused, Create volume, and Restore from a destination buttons, and a list of volumes such as postgres-data and redis-data with their driver, size, the container that uses them, mount point, and a folder icon and bin icon on each row](https://cdn.ctrlops.io/docs/docker-volumes-list.webp)

The In use column names the container that uses each volume. A dash means no container does.

[Watch: Docker Volumes Without the Terminal - Find, Inspect & Clean Up Your Data (CtrlOps)](https://www.youtube.com/watch?v=dZon9ixGsuo)

## Find what you need

| I want to | Instead of typing | Go to |
| --- | --- | --- |
| See my volumes and their size | `docker volume ls`, `docker system df -v` | [Read the list](#read-the-volume-list) |
| See who uses a volume | `docker volume inspect` | [See the details](#see-the-details-of-a-volume) |
| Back up a volume | A `tar` command and a cron line | [Back up a volume](#back-up-a-volume) |
| Get my data back | A `tar` command in a helper container | [Restore a volume](#restore-a-volume) |
| Move a volume to another server | `tar`, `scp`, then `tar` again | [Move a volume](#move-a-volume-to-another-server) |
| Look at the files | `sudo ls /var/lib/docker/volumes/...` | [Open its files](#open-the-files-in-a-volume) |
| Make a new volume | `docker volume create` | [Create a volume](#create-a-volume) |
| Delete a volume | `docker volume rm` | [Delete a volume](#delete-a-volume) |
| Free up disk | `docker volume prune` | [Remove unused](#remove-unused-volumes) |

## Read the volume list

Each row shows the name, driver, size, the containers that use it, and where it is on the server. **Search volumes…** finds a volume by its name, driver, or stack. Use **In use** and **Unused** to filter the list.

Two marks can show next to a name:

| Mark | What it means |
| --- | --- |
| **anonymous** | Docker made it with a random name, because the container did not name one. It is harder to tell what is inside, and it cannot be backed up until you give it a name |
| **stack** | A Compose stack made it. If you delete it, the stack makes a new empty one at its next deploy |

Sizes can take a moment to show, because Docker has to measure each volume.

## See the details of a volume

Click a row to open its details: the **Summary**, **Used by** with the path in each container, **Labels**, and **Options**. **Used by** is the place to check before you delete a volume. A stopped container counts as using it too.

## Back up a volume

> **Backups start from the container**
> 
> The Volumes tab has no backup button. You set up a backup from a container that uses the volume, because that is where CtrlOps knows which containers to stop.

### Open the container

In the [Containers](https://ctrlops.io/docs/modules/docker/containers) tab, click a container that uses the volume, then open its **Volumes** tab.

![The Volumes tab of the postgres container with one mount, postgres-data on /var/lib/postgresql with mode rw and type volume, the Set up link in the Backup column, the bin icon, and the Add volume button](https://cdn.ctrlops.io/docs/docker-volumes-tab.webp)

Set up in the Backup column starts a backup of that volume.

### Click Set up

Each mount has a **backup** column. Click **Set up** to open the backup panel.

![The backup panel open under the postgres-data mount, with Volume backup chosen over S3 destination and the folder on the server where the backups go, the Scheduler set to every day, the Backup prefix daily-, Keep last backups 7, the Stop container and Enabled switches on, the next run time, the container it stops, and the Create button](https://cdn.ctrlops.io/docs/docker-volume-backup-panel.webp)

The backup panel opens under the mount. It shows the next run and which containers stop.

### Choose where the backups go

| Choice | Where they go | Use it when |
| --- | --- | --- |
| **S3 destination** | Your S3 bucket | Almost always. Your data is safe if the server dies |
| **Volume backup** | This same server | You want quick local copies. If you lose this server, you also lose these backups |

For S3, pick a saved destination, or click **+ Add destination** and click **Test connection**. The fields are the same as in the [Backup tab](https://ctrlops.io/docs/modules/backup#add-your-destination), and destinations are shared between the two.

### Choose when and how many

- **Scheduler:** from every minute to every year, or **custom** with a cron expression. Once a day is enough for most data. Times use the time zone of the server, and the panel shows the next run.
- **Keep last backups:** how many copies to keep, from 1 to 1000. Older ones go only after a new backup succeeds.
- **Stop container:** stops the containers that use the volume during the backup, then starts them again.
- **Backup prefix** (optional): added to the start of each file name, such as `daily-`.

> Turn on **Stop container** for a database, because a copy of a running database can be broken. CtrlOps turns it on for you on a database volume. Plan the schedule for a quiet time, because the app is down during the backup.

### Create and test it

Click **Create**, then click **Start backup** to make the first one now. Do not wait for the schedule to find out if it works.

CtrlOps may ask to install rclone or cron on the server first. Backups run on the server, so they keep running while CtrlOps is closed.

Once a volume has a backup, its cell shows the schedule and a colored dot. Click it to see **Last run**, **Restore points**, and **Runs**. To get a message when a backup fails, turn on the **Backup failed or late** alert in the [Alerts](https://ctrlops.io/docs/modules/alerts) tab.

### Back up the volumes of a new stack

When you create a container or [deploy a Compose stack](https://ctrlops.io/docs/modules/docker/containers#deploy-a-compose-stack), the **Container stack** table has a **backup** column. Click **Set up** to open **Volumes to back up**, then tick **Enable backup for this volume**. Bind folders are set up from the Volumes tab of the container instead.

## Restore a volume

### Open the backup panel

Open the container, go to its **Volumes** tab, and click the backup cell of the volume. If the container is gone, see [Restore when the container is gone](#restore-when-the-container-is-gone).

### Pick a restore point

**Restore points** lists every backup with its time and size. Click **Restore…** on the one you want. You can also **Download** it to your computer.

### Choose where it goes

| Option | What happens | Use it when |
| --- | --- | --- |
| **Restore to** the volume | The current data is replaced. Containers stop during the restore and start again after it | You want to go back to that backup |
| **Restore to a new volume** | The backup goes into a new volume. Nothing else changes | You want to look at old data first, or copy one file back |

![The Restore volume dialog with Restore to the volume and Restore to a new volume, the list of containers CtrlOps stops and starts again, the note that CtrlOps keeps a copy of the current data first, and the field to type the volume name](https://cdn.ctrlops.io/docs/docker-volume-restore-dialog.webp)

The dialog names the containers it will stop before you confirm.

### Confirm

Type the volume name and click **Restore**. For a restore over the current volume, each step shows as it runs: get the backup, check it, stop the containers, copy the current data, empty the volume, restore the backup, and start the containers.

> **You have a safety net**
> 
> Before CtrlOps replaces a volume, it keeps a copy of the current data, which shows under **Restore points** as *Before restore*. If the restore fails, CtrlOps puts the current data back.

A new volume is not attached to anything. To use it, [add it to a container](https://ctrlops.io/docs/modules/docker/containers#add-or-remove-a-mount) or name it in a compose file.

### Restore when the container is gone

Click **Restore from a destination…** at the top of the Volumes tab. Pick the **Destination**, the **Server**, the **Volume or folder**, and the **Restore point**. If the volume no longer exists, CtrlOps makes it.

This works for backups in S3 destinations saved on this device.

### Move a volume to another server

1. On the first server, [back up the volume](#back-up-a-volume) to an **S3 destination**, and click **Start backup**.
2. Connect to the second server and open its **Volumes** tab.
3. Click **Restore from a destination…**, pick the first server, the volume, and the backup.
4. Start your container on the new server with that volume.

## Open the files in a volume

Click the folder icon (**Open in File Manager**) on the row to go to the folder of the volume in the [File Manager](https://ctrlops.io/docs/modules/file-manager). This works for volumes that use the `local` driver.

> Be careful when you change files here while a container uses the volume, because the app does not know you changed them.

## Create a volume

You rarely need this. When you [create a container](https://ctrlops.io/docs/modules/docker/containers#run-a-container-from-an-image) and type a new name under **Volume mounts**, such as `app-data:/data`, Docker makes the volume for you.

To make one first, click **Create volume**, type a **Name**, and leave the **Driver** on `local` unless you use a volume plugin. **Driver options** and **Labels** are optional. Do not put secrets in **Driver options**, because anyone who can run Docker can read them.

## Delete a volume

Click the bin icon (**Delete**) on the row and type the name of the volume to confirm.

> You lose the data in the volume, and you cannot undo this. Backups in your bucket stay, so you can still restore them.

If a container uses the volume, **Delete** is off and names the container. Delete or change that container first.

## Remove unused volumes

Click **Remove unused** to see the volumes no container uses. Each shows its size, or *size unknown* if Docker has not measured it. Anonymous volumes are ticked for you, and named volumes are not, because a name usually means someone wanted to keep that data.

![The Docker Volumes tab with the cursor on the Remove unused button at the top, next to Create volume and Restore from a destination](https://cdn.ctrlops.io/docs/docker-click-remove-unused-in-volume.webp)

Remove unused is at the top of the Volumes tab.

Tick what you want gone, and type `delete` and the number you ticked, such as `delete 2`, to confirm. Before you tick a named volume, ask yourself if its container was just deleted and you plan to bring it back.

![The Prune Volumes dialog saying CtrlOps deletes the selected volumes and you lose the data in them, listing four unused volumes such as couchdb-data and n8n_n8n-data, each marked named, not ticked, and with size unknown, the Type delete 0 to confirm field, and the Cancel and Delete buttons](https://cdn.ctrlops.io/docs/docker-prune-volumes.webp)

Named volumes start unticked, so nothing goes by accident. The number to type follows what you tick.

## Tips

- **Name your volumes.** `billing-db-data` tells you what is inside, and only named volumes can be backed up.
- **Back up to S3, not to the same server.** A backup on the same server is gone when the server is gone.
- **Test a restore once, to a new volume.** A backup you have never restored is a backup you only hope works.

## Troubleshooting

Most failures show **Show details** with the **Likely cause**, **Retry** where it helps, and **Copy report**. Send that report with any support question.

### Give the volume a name first. A recreate gives an anonymous volume a new name.

An anonymous volume cannot be backed up, because its name changes when the container is recreated. Change the mount to a named volume, then recreate the container.

### tmpfs data is in memory only. CtrlOps cannot make a backup of it.

A `tmpfs` mount keeps its files in memory, and they are gone when the container stops. Use a volume for data you want to keep.

### A backup stopped 1 container and did not start it.

A backup with **Stop container** ended before it could start the container again, for example when the server restarted. Click **Start it**.

### The schedule of this backup is not in the crontab.

Someone or something removed the cron line. Click **Repair** to put it back. If the message is *The crontab has a backup schedule, but the backup is not on the server.*, click **Remove**.

### Another device made this backup.

The S3 keys are saved on the computer that made it. To change the backup here, add the same destination with its keys on this computer.

### There is not enough free disk space.

The message says how much CtrlOps needs. A restore needs room for the backup, and for the copy of the current data. Use [Prune System](https://ctrlops.io/docs/modules/docker/dashboard#free-up-disk-space) to free some, then try again.

### The restore failed. CtrlOps put the current data back.

Your data is as it was before the restore. Open **Show details** to read why it failed. If the message says CtrlOps could not put the data back, it names the folder that holds the copy.

### You need sudo to open this folder.

The folder of a volume belongs to root. Connect with a login that has sudo rights, and type the password when CtrlOps asks. Volumes on other drivers cannot open in the File Manager.

### Docker did not send the volume sizes in 60 seconds.

Measuring sizes is slow on a server with many volumes or files. Click **Refresh** to read them again. The rest of the list still works.

### Docker does not have this volume driver.

Install the volume plugin on the server, or choose the `local` driver.

## Frequently Asked Questions

### Does deleting a container delete its volume?

A named volume stays on the server when you delete its container. An anonymous volume goes only if you tick **Also remove anonymous volumes**. A volume is removed only when you delete it yourself or remove it as unused.

### Where do I set up a volume backup?

Not in the Volumes tab. Open a container that uses the volume, go to its **Volumes** tab, and click **Set up** in the backup column. You can also set up backups in the **Container stack** table when you create a container or deploy a Compose stack.

### Does CtrlOps have to be open for a backup to run?

No. A scheduled backup runs on the server through cron, so it runs while CtrlOps and your computer are off.

### Is it safe to back up a running database?

A copy of a running database can be broken, so turn on **Stop container**. CtrlOps turns it on by default for a database volume. The container stops during the backup, and CtrlOps starts it again afterwards.

### How do I move a volume to another server?

Back it up to S3 on the first server. Connect to the second server, open the **Volumes** tab, click **Restore from a destination…**, and pick the backup. CtrlOps makes the volume if it is not there.

### How are volume backups different from the Backup tab?

Volume backups are for Docker volumes and bind folders of a container, and they can stop the container during the backup. The [Backup tab](https://ctrlops.io/docs/modules/backup) is for any folder on the server. Both use the same destinations, and volume backups also show in the Backup tab with a **Docker** tag.
