# Docker Dashboard: Check Server Health and Free Up Disk Space
URL: https://ctrlops.io/docs/modules/docker/dashboard
Published: 2026-11-05
Updated: 2026-11-11

The Dashboard tab shows at a glance whether Docker on your server is healthy, what needs your attention, and what is using your disk. Free up space with a preview first, restart Docker, and install Docker on a server that does not have it.

The **Dashboard** tab answers one question fast: is Docker on this server OK? If it is not, the tab tells you what is wrong and gives you a button to fix it.

![The Docker Dashboard with the Docker engine card showing Healthy and version 25.0.18 and the Storage card on the left, the Container fleet card with 2 of 3 containers running, the CPU, Memory, and Disk meters, and the Uptime card on the right, and the Refresh, Restart Docker, and Quick Actions buttons at the top](https://cdn.ctrlops.io/docs/docker-dashboard.webp)

The Dashboard shows Docker, its storage, your containers, and the server on one screen.

[Watch: Is Your Docker Server Okay? The Docker Dashboard on One Screen (CtrlOps)](https://www.youtube.com/watch?v=E8rl053HAds)

## Find what you need

| I want to | Instead of typing | Go to |
| --- | --- | --- |
| Check that Docker is healthy | `docker info`, `systemctl status docker` | [Read the Dashboard](#read-the-dashboard) |
| See what needs fixing | Several commands | [Fix what needs attention](#fix-what-needs-attention) |
| See what is using disk | `docker system df` | [Storage card](#read-the-dashboard) |
| Free up disk space | `docker system prune` | [Free up disk space](#free-up-disk-space) |
| Restart Docker | `sudo systemctl restart docker` | [Restart Docker](#restart-docker) |
| Install Docker | `curl -fsSL https://get.docker.com \| sh` | [Install Docker](#install-docker-on-a-server) |

## Read the Dashboard

| Card | What it tells you |
| --- | --- |
| **Docker engine** | Whether Docker runs, for how long, and its version |
| **Storage** | The space used by **Images** and **Volumes**, and how much **Build cache** you can get back. Click a row to open that tab |
| **Container fleet** | How many containers run out of the total, split into **Running**, **Paused**, **Restarting**, and **Stopped** |
| **CPU**, **Memory** | How busy the server is right now |
| **Disk** | How full the disk that holds Docker's data is. Rest on it to see which disk |
| **Uptime** | How long the server has been on, and since when |

The **CPU**, **Memory**, and **Disk** meters use colors: green is fine, amber means 60% or more, and red means 80% or more.

The numbers update by themselves, and **Storage** updates about once a minute. Click **Refresh** to read everything now.

> The Dashboard shows values as they are now. It keeps no history, and it covers this server only. Each server has its own Docker page.

## Fix what needs attention

When something is wrong, a box shows below the cards. It says what is wrong and has a button for the next step. No box means there is nothing to do.

![A box below the Dashboard cards saying Container logs have no size limit, explaining that logs can fill the disk and that the fix restarts the Docker Engine, with the Apply log limits button](https://cdn.ctrlops.io/docs/docker-needs-attention.webp)

Each box says what is wrong and gives you the next step.

| You see | What it means | What to do |
| --- | --- | --- |
| *1 unhealthy container* | A running container fails its health check | Click **Show**, then read its logs |
| *1 container in a restart loop* | It crashes and restarts again and again | Click **Show**, then read its logs |
| *1 container stopped with an error* | It stopped with an error in the last 24 hours | Click **Show**, then read its logs |
| *The Docker disk is 84% full.* | The disk reached 80% | Click **Prune System**, or **Set an alert** to hear about it next time |
| *The certificate for app.example.com expires in 5 days.* | A certificate is close to expiring, has expired, or could not be issued | Click **Show domains** and see [Domains](https://ctrlops.io/docs/modules/docker/domains#read-the-domain-list) |
| *Certificates do not renew on this server.* | Renewal has stopped | Click **Show domains**, then **Start renewal** |
| *Docker Compose is not installed.* | Stacks and Compose cannot run | Click **Install Compose plugin** |
| *Container logs have no size limit.* | Logs can fill the disk | Click **Apply log limits** |
| *Docker from snap has limits.* | Stacks, Compose, Build, and bind mounts outside your home folder are off | Install Docker another way if you need them |
| *Docker is too old.* | The server runs a version below 20.10 | Upgrade Docker on the server |

**Show** opens the [Containers](https://ctrlops.io/docs/modules/docker/containers) tab with only those containers listed. To find the cause, see [Find out why a container failed](https://ctrlops.io/docs/modules/docker/logs#find-out-why-a-container-failed).

> **Apply log limits** restarts Docker, so read [Restart Docker](#restart-docker) first. It sets 10 MB and 3 files for each **new** container. If it fails, CtrlOps puts the old settings back.

## Free up disk space

**Prune System** is the safe place to start. You see what it found before anything is removed, and it keeps anything a container uses.

### Open Prune System

Click **Quick Actions** at the top right of the Dashboard, then the **Prune System** card. If the disk is nearly full, the same button is also in a box below the cards.

![The Docker Dashboard with the cursor on the green Quick Actions button at the top right, next to Refresh and Restart Docker](https://cdn.ctrlops.io/docs/docker-quick-action-button.webp)

Quick Actions is at the top right of the Dashboard.

### Choose what to remove

CtrlOps lists what it found, with the number of items and their size. Only the groups with something in them show, and the arrow on a group lists its items. Three groups are ticked for you, and three are not, because they cost you something.

| Group | Ticked at first | Worth knowing |
| --- | --- | --- |
| **Stopped containers** | Yes | You lose any data stored inside them |
| **Unused networks** | Yes |  |
| **Untagged images** | Yes | Old image layers nothing points to |
| **Images with a digest only** | No | A compose file can name an image this way |
| **Stopped stacks** | No | Keep them if you want to start them again |
| **Build cache** | No | Your next builds can take longer |

![The Prune System dialog for one server, saying CtrlOps removes the selected items and Docker keeps any item in use, with the Stopped containers group ticked at 1 item, the Type prune to confirm field, and the Cancel and Prune buttons](https://cdn.ctrlops.io/docs/docker-prune-system.webp)

This server has only one stopped container, so only that group shows. Prune works once you type prune.

### Confirm

Type `prune` and click **Prune**. If a pull or a build is running, CtrlOps waits until Docker is idle, and **Run now** skips the wait. You can also click **Run in background** and keep working.

### Read the result

The result shows what went and how much you got back, for example *Removed 12 items. About 4.2 GB freed.* Anything that came into use is listed as *Kept*, with the reason.

### Remove unused volumes

Volumes hold your data, so **Prune System** never touches them. To remove them, click **Quick Actions**, then **Prune Volumes**. CtrlOps lists the volumes no container uses. Anonymous volumes are ticked for you, and named ones are not. Pick what you want gone, and type `delete 2` to confirm. See [Remove unused volumes](https://ctrlops.io/docs/modules/docker/volumes#remove-unused-volumes).

> You lose the data in a deleted volume, and you cannot undo it. A database volume can look unused while its container is deleted. If you are not sure, check it first in the **Volumes** tab.

## Use Quick Actions

**Quick Actions** at the top of the Dashboard opens the common jobs on one page. The back arrow next to the title takes you back to the Dashboard.

| Card | What it does |
| --- | --- |
| **Create Container** | Opens the [Create Container](https://ctrlops.io/docs/modules/docker/containers#run-a-container-from-an-image) form |
| **Pull Image** | [Downloads an image](https://ctrlops.io/docs/modules/docker/images#pull-an-image) from a registry to this server |
| **Restart Docker Engine** | Same as [Restart Docker](#restart-docker). Needs root or sudo |
| **Compose Up** | Lists the stopped stacks, so you can start one. It is off if the server has no stacks |
| **Prune System** | [Removes unused items](#free-up-disk-space) |
| **Prune Volumes** | [Removes unused volumes](#remove-unused-volumes) |

![The Quick Actions page of a server, with the back arrow next to the title and six cards, each with a short note: Create Container, Pull Image, Restart Docker Engine, Compose Up, Prune System, and Prune Volumes](https://cdn.ctrlops.io/docs/docker-quick-actions.webp)

Quick Actions puts the common jobs in one place.

## Restart Docker

Click **Restart Docker** at the top of the Dashboard, and check the dialog before you confirm.

Every container stops while Docker restarts. Containers with a restart policy come back by themselves. The dialog tells you how many will not and names them, for example *2 running containers have no restart policy. They stay stopped after the restart.*

To avoid that, set a [restart policy](https://ctrlops.io/docs/modules/docker/containers#change-the-name-restart-policy-or-limits) on those containers first.

![The Restart the Docker Engine dialog, saying 2 running containers have no restart policy and stay stopped after the restart, naming the test and grafana containers, with the Cancel button and the cursor on Restart](https://cdn.ctrlops.io/docs/docker-restart-dialog.webp)

The dialog names the containers that will not come back by themselves.

## Install Docker on a server

If the server has no Docker, the tab shows *Docker Engine not installed*. You need a login with root or sudo rights.

### Click Install Docker

![The Docker Engine not installed screen in CtrlOps, saying the host is reachable over SSH but the Docker Engine isn't installed yet, with the cursor on the green Install Docker button](https://cdn.ctrlops.io/docs/if-not-then-you-can-install.webp)

On a server without Docker, the tab offers to install it.

### Read what will change

The dialog lists the OS, the Docker version, and these changes:

- It installs the Docker packages.
- It starts `docker.service` when the server starts.
- It adds log limits to `/etc/docker/daemon.json`: 10 MB and 3 files for each container.

On a server that uses pacman, it also updates all packages on the server. **Show commands** shows the exact commands.

![The Install Docker window in CtrlOps listing the OS Ubuntu 24.04.1 LTS, Docker version 28.5.0, a time of about 2 to 5 minutes, and the changes it makes on the server, with Show commands, Cancel, and the cursor on the green Install button](https://cdn.ctrlops.io/docs/clarification-before-install-docker.webp)

Nothing changes on the server until you click Install.

### Click Install

Each step shows with its output: connect, detect the OS, prepare the packages, install Docker, enable the service, and verify. It takes about 2 to 5 minutes. When it is ready, the Dashboard opens.

The install runs on the server, so it keeps going if you lose the connection. You can cancel until the packages start installing.

![The Installing Docker Engine screen at 58 percent, with the six steps listed, Connect over SSH, Detect operating system, and Add Docker apt repository done, Install Docker Engine running, and the live install log below the steps](https://cdn.ctrlops.io/docs/docker-installing.webp)

Each step shows as it runs, with the live output below.

CtrlOps installs on servers that use apt, dnf, yum, zypper, apk, or pacman. It adds Docker's own repository on apt and dnf, and uses the package from your OS on the others.

### If Docker is stopped

The tab shows *Docker Engine stopped.* Click **Start Docker**. If your login has no sudo, the tab shows a command for an administrator to run.

## Tips

- **Check the Disk meter before a big pull or build.** A full disk is the most common reason a build fails.
- **Set an alert before the disk fills up.** When the disk is nearly full, click **Set an alert** in the box below the Dashboard cards to open [Alerts](https://ctrlops.io/docs/modules/alerts).
- **Tick Build cache only when you need the space.** Your next builds take longer without it.

## Troubleshooting

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

### CtrlOps cannot install Docker on this server.

The server has no supported package manager. Use the **Docker install guide** link to install Docker yourself, then connect to the server again.

### To install Docker, you need root or sudo.

Your login cannot install packages. Ask an administrator to install Docker, or connect with an account that has sudo.

### Another program uses the package manager.

The server is busy with something else, often automatic updates. Wait until it stops, then click **Retry**.

### Packages on this server conflict with Docker, for example podman or runc.

The output names them. CtrlOps does not remove packages, so remove them on the server, then click **Retry**.

### The server could not download a file.

The server cannot reach the internet. Check its connection, then click **Retry**. If the message is *A downloaded file is not correct.*, click **Retry**, and send us the report if it happens again.

### The Docker Engine does not answer after the install.

The output below the message shows the cause. Copy the report and send it to us if it is not clear.

### The Docker Engine did not start in 60 seconds.

The last lines of the Docker log show below the message, and they usually name the cause. A full disk and a broken `/etc/docker/daemon.json` are common reasons.

### Could not read the server values.

CtrlOps could not read CPU, memory, and disk from this server, but Docker can still work. Click **Retry**. *Could not read the disk use.* is the same for the **Storage** card, which can be slow on a server with many volumes.

### Nothing to remove.

The server has no stopped containers, unused networks, or untagged images. If the disk is still full, check **Images** and **Volumes** on the **Storage** card to see what takes the space.

### Prune finished, but it kept some items

Something started to use them after the preview, or a backup of a volume was running. The result lists each one as *Kept*, with the reason. Run the prune again later if you still want them gone.

## Frequently Asked Questions

### Does Prune System delete my volumes?

No. **Prune System** never touches volumes. Volumes have their own action, **Prune Volumes**, where you pick each volume and type a confirmation.

### Is it safe to prune while containers are running?

Yes. CtrlOps shows you what it will remove before it starts, and it keeps anything a container uses. It also checks each item again just before it removes it.

### Will Restart Docker stop my containers?

Yes, for a short time, because every container stops when Docker restarts. Containers with a restart policy come back by themselves. Running containers with no restart policy stay stopped, and the dialog lists how many before you confirm.

### What does the Disk meter measure?

The disk that holds Docker's data, which can be a different disk from the rest of the server. Rest on the meter to see the folder and the mount.

### What does CtrlOps change when it installs Docker?

It installs the Docker packages, starts `docker.service` when the server starts, and adds log limits of 10 MB and 3 files for each container. The dialog lists these before it starts, and **Show commands** shows the exact commands.

### Can I install Docker without CtrlOps?

Yes. If Docker is already on the server, CtrlOps reads it and uses it. You only need **Install Docker** on a server that does not have it.
