Skip to content

System Checks failed in Prerequisites (clock drift, username not found, Docker not running, docker0)

TroubleshootingDeployrr v6.0+CurrentInstallDockerHost system

You run Prerequisites > System Checks and it stops with a red error. The menu row then reads System and Docker Checks - NOT DONE. Later steps, such as Domain Checks, need it to show DONE. This page lists the five checks, the message each failure prints, and how to fix it.

System Checks runs five numbered steps in the terminal, in this order. It stops at the first failure. The failure line starts with a red ✗.

Step What you see while it runs Failure message
1/5 Verifying system time synchronization... System clock drift detected (N seconds off), then a box titled Time Synchronization Error
2/5 Verifying Linux username... Username YOUR_USERNAME not found
3/5 Checking Docker installation... See Docker is missing or not running
4/5 Verifying Docker daemon... Docker installed but not running or Docker installation failed
5/5 Checking docker0 network IP... Could not determine IP of docker0 interface

After any failure you also see:

Check the documentation at https://docs.deployrr.app.
If needed, you may generate a troubleshooting log from the Settings menu.

Then you see Going back to the menu in 60 seconds. Or, press OK to go back now.... Deployrr returns to the menu after 60 seconds, or sooner when you press Enter.

When everything passes, the last line is:

All System and Docker checks passed

System Checks makes sure the clock is right, your Linux user exists, Docker runs, and Docker has its network address. A failed check means one of these is wrong on the host. Fix it on the host, then run the step again. See Run System Checks again.

If you see Requirements Not Met instead of the five steps, finish Disclaimers and Username first. The username must be a Linux user other than root. See Requirements Not Met.

All commands below run on the Docker host, in a terminal. Commands marked as general Linux advice are not part of Deployrr. Menus and package names can differ by distribution.

You see this line, then a box titled Time Synchronization Error:

System clock drift detected (42 seconds off)

The number is an example. The box says your clock is off by that many seconds. It warns that Let’s Encrypt and Authelia will fail. It tells you to fix the clock on the host (Proxmox or the operating system) using NTP or timedatectl.

Deployrr compares your server’s clock with the time from an internet server. A gap of more than 10 seconds fails the check. The check uses the real time, so a wrong timezone alone does not cause it.

The steps below are general Linux advice.

  1. Look at the clock status:

    Terminal window
    timedatectl status
  2. Find the lines System clock synchronized and NTP service. If either says no or inactive, the clock is not syncing.

  3. Turn on time syncing:

    Terminal window
    sudo timedatectl set-ntp true
  4. Wait about 30 seconds. Check again:

    Terminal window
    timedatectl status

    System clock synchronized: yes means it worked.

  5. If set-ntp fails with a message that the NTP service is not installed, install a time service. Pick one. This is general Debian and Ubuntu advice.

    Terminal window
    sudo apt install systemd-timesyncd

    or:

    Terminal window
    sudo apt install chrony
  6. Run System Checks again.

This is general Proxmox advice. An LXC container uses the clock of the Proxmox host. You can’t set it from inside the container. Fix the time on the Proxmox host, then run System Checks again.

A VM keeps its own clock. If the time keeps drifting, check the time settings of the Proxmox host as well.

“Skipping time synchronization check” is not an error

Section titled ““Skipping time synchronization check” is not an error”

If the server can’t reach the time source, you see a yellow warning instead of a failure:

Network API unreachable for time check. Skipping time synchronization check.

System Checks skips the clock check and goes on to step 2. You may also see:

Time parsing failed. Assuming time is fine.

It also goes on. If you see either warning, check that the server has internet access. See No internet on the server or in containers after setup.

You see:

Username YOUR_USERNAME not found

The name saved in the Username step is not a Linux user on this server. The name may have a typo, the user may have been deleted, or you are on a different server than before.

  1. Check whether the user exists. Replace YOUR_USERNAME with the name from the message. This is a value you fill in.

    Terminal window
    id YOUR_USERNAME

    If the user exists, you see its uid, gid and groups. If not, you see no such user.

  2. If the name has a typo, go back to Prerequisites > Username and enter the correct one. See Prerequisites: Username.

  3. If the user does not exist, run Username again. Deployrr asks if you want to create the user and add it to sudoers.

  4. Run System Checks again.

If the user was renamed, or you want Deployrr to use a different user on purpose, read Use a different Linux user after setup before you change anything.

To see a user created by hand, watch Docker on Proxmox LXC: creating a non-root user at 10:20. For the written steps, read UDMS part 05: create a non-root Linux user (manual setup).

Username only saves a name. Your Docker folder, your apps and a few other steps keep the old user until you update them. The steps now have their own page: Use a different Linux user after setup.

Fix Docker is missing or not running (steps 3 and 4)

Section titled “Fix Docker is missing or not running (steps 3 and 4)”

Deployrr checks Docker when it starts. It counts Docker as installed only if Docker answers. This matters for what happens next.

Docker answered at start. You see:

Docker already installed
Would you like to check and update your system? (Y/n):

Type y to update all system packages now. Press Enter, or any other key, to skip. The capital Y in the prompt is not the default: Enter skips. You then see Skipping system updates. If you update, Deployrr checks the Docker service and starts it if needed. You see Docker service is running or Starting Docker service....

Docker did not answer at start. You see:

Docker not installed. Installing now...

This also appears when Docker is installed but stopped. Deployrr then installs Docker:

  • On Arch, you see Installing Docker from Arch repos....
  • On every other system, Deployrr uses the Debian and Ubuntu install. You see Installing Docker using the convenience script.... If that fails, you see Convenience script failed. Trying fallback installation..., then Installing Docker packages with held package override....
  • On systems that are not based on Debian or Arch, the install is likely to fail. Install Docker yourself first, start it, then restart Deployrr. See the Docker documentation.

After the install, you see Starting Docker service....

To see System Checks install Docker, watch Deployrr 5 walkthrough: System Checks at 11:17 (shows Deployrr v5; menus differ in v6).

Step 4 shows one of these:

Docker is installed and running
Docker installed but not running
Docker installation failed
  • Docker installed but not running: the Docker program is on the server, but the Docker service did not start or does not answer.
  • Docker installation failed: the Docker program is still not on the server after the install. The install may have failed because of no internet, a package manager problem, or an unsupported distribution.
  1. Check the service:

    Terminal window
    sudo systemctl status docker
  2. If it says inactive or failed, start it:

    Terminal window
    sudo systemctl start docker
  3. Check that Docker answers:

    Terminal window
    sudo docker info

    You should see a long list of details, not an error.

  4. If the service will not start, read the reason in its log. This is general Linux advice:

    Terminal window
    sudo journalctl -u docker --no-pager -n 50
  5. Once sudo docker info works, close Deployrr and start it again. Step 3 uses the Docker state from when Deployrr started.

  6. Run System Checks again.

  1. Check that the server has internet access. See No internet on the server or in containers after setup.
  2. Check your distribution. Deployrr runs the Arch install on Arch. On every other system it runs the Debian and Ubuntu install, which only works on Debian-based systems.
  3. Scroll up in the terminal and read the errors printed during the install.
  4. Fix those errors. Or install Docker by hand from the Docker documentation and make sure sudo docker info works.
  5. Run System Checks again.

To see Docker installed by hand with the convenience script, watch Docker on Proxmox LXC: Docker installation at 18:22. To check that Docker runs, read UDMS part 12: verify that Docker is running (manual setup).

Containers fail to start after Docker is reinstalled

Section titled “Containers fail to start after Docker is reinstalled”

If Docker was installed again on a Proxmox LXC container, your containers may fail with failed to create shim task. See Docker fails with “failed to create shim task” in a Proxmox LXC container.

You only need this if you pinned a Docker package with apt-mark hold, as described in the shim task page.

How System Checks treats held packages on Debian and Ubuntu:

  • You answer y to the update. Deployrr shows no hold messages. A normal apt upgrade skips held packages, so your hold usually stays. This part is general apt behaviour.
  • Deployrr installs Docker (you saw Docker not installed. Installing now...). You see Checking for held packages.... For each one it shows Unholding PACKAGE..., where PACKAGE is the package name, such as containerd.io. It then installs the newest version. It does not hold the package again.

The packages it releases are docker-ce, docker-ce-cli, containerd.io, docker-compose-plugin, docker-ce-rootless-extras, docker-buildx-plugin and docker-model-plugin.

To check and re-apply a hold:

  1. Check which packages are held:

    Terminal window
    apt-mark showhold

    If containerd.io is not in the list, the hold is gone.

  2. Check the installed version:

    Terminal window
    containerd --version
  3. If you need an older version, follow fix 1 in the shim task page. It installs the version you want and holds it.

  4. If you already have the right version, hold it again:

    Terminal window
    sudo apt-mark hold containerd.io
  5. Restart Docker:

    Terminal window
    sudo systemctl restart docker

Fix “Could not determine IP of docker0 interface” (step 5)

Section titled “Fix “Could not determine IP of docker0 interface” (step 5)”

You see:

Could not determine IP of docker0 interface

Docker is running, but Deployrr can’t read the address of Docker’s default network (docker0). Docker may have started without that network, for example because of a custom Docker setup.

The steps below are general Docker and Linux advice.

  1. Check that the network exists:

    Terminal window
    ip addr show docker0

    A working network shows a line like inet 172.17.0.1/16. The address on your server can differ.

  2. Check what Docker reports for it:

    Terminal window
    sudo docker network inspect bridge

    Look for Gateway under IPAM.

  3. If docker0 is missing, restart Docker:

    Terminal window
    sudo systemctl restart docker

    Then run ip addr show docker0 again.

  4. If it is still missing, look at your Docker settings file for a "bridge": "none" line. That line turns the default network off.

    Terminal window
    sudo cat /etc/docker/daemon.json
  5. After you change the file, restart Docker with sudo systemctl restart docker.

  6. Run System Checks again.

On success, you see a line like docker0 network IP: 172.17.0.1. If your Docker folder already has a .env file, Deployrr saves the address there and shows DOCKER0_IP saved to .env.

  1. From the main menu, select Prerequisites.

  2. Select System Checks.

  3. If the step already shows DONE, you see:

    It appears that System Checks were already completed.
    If you want to redo it, press Y/y to resume or any other key to go back:

    Press y to run it again. Any other key goes back to the menu.

  4. Read each step as it runs.

When you press y, the status changes to NOT DONE. If the rerun fails, it stays NOT DONE, even if the step passed before. Fix the cause and run it again.

The steps end with:

All System and Docker checks passed

In the Prerequisites menu, the row now reads System and Docker Checks - DONE. In Hybrid mode you can continue with Domain Checks. If that step fails, see Domain Checks failed. In Local mode, Domain Checks is not needed.