This project implements a Docker container for Firefox.
The GUI of the application is accessed through a modern web browser (no installation or configuration needed on the client side) or via any VNC client.
&textColor=rgba(121,121,121,1))
Mozilla Firefox is a free and open-source web browser developed by Mozilla Foundation and its subsidiary, Mozilla Corporation.
- Quick Start
- Usage
- Docker Compose File
- Docker Image Versioning
- Docker Image Update
- User/Group IDs
- Accessing the GUI
- Security
- Reverse Proxy
- Shell Access
- Allowing the membarrier System Call
- Sound Support
- Setting Firefox Preferences Via Environment Variables
- Troubleshooting
- Support or Contact
NOTE: The Docker command provided in this quick start is given as an example and parameters should be adjusted to your need.
Launch the Firefox docker container with the following command:
docker run -d \
--name=firefox \
-p 5800:5800 \
-v /docker/appdata/firefox:/config:rw \
jlesage/firefoxWhere:
/docker/appdata/firefox: This is where the application stores its configuration, states, log and any files needing persistency.
Browse to http://your-host-ip:5800 to access the Firefox GUI.
docker run [-d] \
--name=firefox \
[-e <VARIABLE_NAME>=<VALUE>]... \
[-v <HOST_DIR>:<CONTAINER_DIR>[:PERMISSIONS]]... \
[-p <HOST_PORT>:<CONTAINER_PORT>]... \
jlesage/firefox| Parameter | Description |
|---|---|
| -d | Run the container in the background. If not set, the container runs in the foreground. |
| -e | Pass an environment variable to the container. See the Environment Variables section for more details. |
| -v | Set a volume mapping (allows to share a folder/file between the host and the container). See the Data Volumes section for more details. |
| -p | Set a network port mapping (exposes an internal container port to the host). See the Ports section for more details. |
To customize some properties of the container, the following environment
variables can be passed via the -e parameter (one for each variable). Value
of this parameter has the format <VARIABLE_NAME>=<VALUE>.
| Variable | Description | Default |
|---|---|---|
USER_ID |
ID of the user the application runs as. See User/Group IDs to better understand when this should be set. | 1000 |
GROUP_ID |
ID of the group the application runs as. See User/Group IDs to better understand when this should be set. | 1000 |
SUP_GROUP_IDS |
Comma-separated list of supplementary group IDs of the application. | (no value) |
UMASK |
Mask that controls how file permissions are set for newly created files. The value of the mask is in octal notation. By default, the default umask value is 0022, meaning that newly created files are readable by everyone, but only writable by the owner. See the online umask calculator at http://wintelguy.com/umask-calc.pl. |
0022 |
LANG |
Set the locale, which defines the application's language, if supported. Format of the locale is language[_territory][.codeset], where language is an ISO 639 language code, territory is an ISO 3166 country code and codeset is a character set, like UTF-8. For example, Australian English using the UTF-8 encoding is en_AU.UTF-8. |
en_US.UTF-8 |
TZ |
TimeZone used by the container. Timezone can also be set by mapping /etc/localtime between the host and the container. |
Etc/UTC |
KEEP_APP_RUNNING |
When set to 1, the application will be automatically restarted when it crashes or terminates. |
0 |
APP_NICENESS |
Priority at which the application should run. A niceness value of -20 is the highest priority and 19 is the lowest priority. The default niceness value is 0. NOTE: A negative niceness (priority increase) requires additional permissions. In this case, the container should be run with the docker option --cap-add=SYS_NICE. |
0 |
INSTALL_PACKAGES |
Space-separated list of packages to install during the startup of the container. Packages are installed from the repository of the Linux distribution this container is based on. ATTENTION: Container functionality can be affected when installing a package that overrides existing container files (e.g. binaries). | (no value) |
CONTAINER_DEBUG |
Set to 1 to enable debug logging. |
0 |
DISPLAY_WIDTH |
Width (in pixels) of the application's window. | 1920 |
DISPLAY_HEIGHT |
Height (in pixels) of the application's window. | 1080 |
DARK_MODE |
When set to 1, dark mode is enabled for the application. |
0 |
SECURE_CONNECTION |
When set to 1, an encrypted connection is used to access the application's GUI (either via a web browser or VNC client). See the Security section for more details. |
0 |
SECURE_CONNECTION_VNC_METHOD |
Method used to perform the secure VNC connection. Possible values are SSL or TLS. See the Security section for more details. |
SSL |
SECURE_CONNECTION_CERTS_CHECK_INTERVAL |
Interval, in seconds, at which the system verifies if web or VNC certificates have changed. When a change is detected, the affected services are automatically restarted. A value of 0 disables the check. |
60 |
WEB_LISTENING_PORT |
Port used by the web server to serve the UI of the application. This port is used internally by the container and it is usually not required to be changed. By default, a containe
8000
r is created with the default bridge network, meaning that, to be accessible, each internal container port must be mapped to an external port (using the -p or --publish argument). However, if the container is created with another network type, changing the port used by the container might be useful to prevent conflict with other services/containers. NOTE: a value of -1 disables listening, meaning that the application's UI won't be accessible over HTTP/HTTPs. |
5800 |
VNC_LISTENING_PORT |
Port used by the VNC server to serve the UI of the application. This port is used internally by the container and it is usually not required to be changed. By default, a container is created with the default bridge network, meaning that, to be accessible, each internal container port must be mapped to an external port (using the -p or --publish argument). However, if the container is created with another network type, changing the port used by the container might be useful to prevent conflict with other services/containers. NOTE: a value of -1 disables listening, meaning that the application's UI won't be accessible over VNC. |
5900 |
VNC_PASSWORD |
Password needed to connect to the application's GUI. See the VNC Password section for more details. | (no value) |
ENABLE_CJK_FONT |
When set to 1, open-source computer font WenQuanYi Zen Hei is installed. This font contains a large range of Chinese/Japanese/Korean characters. |
0 |
FF_OPEN_URL |
The URL to open when Firefox starts. | (no value) |
FF_KIOSK |
Set to 1 to enable kiosk mode. This mode launches Firefox in a very restricted and limited mode best suitable for public areas or customer-facing displays. |
0 |
Many tools used to manage Docker containers extract environment variables defined by the Docker image and use them to create/deploy the container. For example, this is done by:
- The Docker application on Synology NAS
- The Container Station on QNAP NAS
- Portainer
- etc.
While this can be useful for the user to adjust the value of environment variables to fit its needs, it can also be confusing and dangerous to keep all of them.
A good practice is to set/keep only the variables that are needed for the container to behave as desired in a specific setup. If the value of variable is kept to its default value, it means that it can be removed. Keep in mind that all variables are optional, meaning that none of them is required for the container to start.
Removing environment variables that are not needed provides some advantages:
- Prevents keeping variables that are no longer used by the container. Over time, with image updates, some variables might be removed.
- Allows the Docker image to change/fix a default value. Again, with image updates, the default value of a variable might be changed to fix an issue, or to better support a new feature.
- Prevents changes to a variable that might affect the correct function of
the container. Some undocumented variables, like
PATHorENV, are required to be exposed, but are not meant to be changed by users. However, container management tools still show these variables to users. - There is a bug with the Container Station on QNAP and the Docker application on Synology, where an environment variable without value might not be allowed. This behavior is wrong: it's absolutely fine to have a variable without value. In fact, this container does have variables without value by default. Thus, removing unneeded variables is a good way to prevent deployment issue on these devices.
The following table describes data volumes used by the container. The mappings
are set via the -v parameter. Each mapping is specified with the following
format: <HOST_DIR>:<CONTAINER_DIR>[:PERMISSIONS].
| Container path | Permissions | Description |
|---|---|---|
/config |
rw | This is where the application stores its configuration, states, log and any files needing persistency. |
Here is the list of ports used by the container.
When using the default bridge network, ports can be mapped to the host via the
-p parameter (one per port mapping). Each mapping is defined with the
following format: <HOST_PORT>:<CONTAINER_PORT>. The port number used inside
the container might not be changeable, but you are free to use any port on the
host side.
See the Docker Container Networking documentation for more details.
| Port | Protocol | Mapping to host | Description |
|---|---|---|---|
| 5800 | TCP | Optional | Port to access the application's GUI via the web interface. Mapping to the host is optional if access through the web interface is not wanted. For a container not using the default bridge network, the port can be changed with the WEB_LISTENING_PORT environment variable. |
| 5900 | TCP | Optional | Port to access the application's GUI via the VNC protocol. Mapping to the host is optional if access through the VNC protocol is not wanted. For a container not using the default bridge network, the port can be changed with the VNC_LISTENING_PORT environment variable. |
As can be seen, environment variables, volume and port mappings are all specified while creating the container.
The following steps describe the method used to add, remove or update parameter(s) of an existing container. The general idea is to destroy and re-create the container:
- Stop the container (if it is running):
docker stop firefox- Remove the container:
docker rm firefox- Create/start the container using the
docker runcommand, by adjusting parameters as needed.
NOTE:
Since all application's data is saved under the /config container
folder, destroying and re-creating a container is not a problem: nothing is
lost and the application comes back with the same state (as long as the
mapping of the /config folder remains the same).
Here is an example of a docker-compose.yml file that can be used with
Docker Compose.
Make sure to adjust according to your needs. Note that only mandatory network ports are part of the example.
version: '3'
services:
firefox:
image: jlesage/firefox
ports:
- "5800:5800"
volumes:
- "/docker/appdata/firefox:/config:rw"Each release of a Docker image is versioned. Prior to october 2022, the semantic versioning was used as the versioning scheme.
Since then, versioning scheme changed to
calendar versioning. The format used is YY.MM.SEQUENCE,
where:
YYis the zero-padded year (relative to year 2000).MMis the zero-padded month.SEQUENCEis the incremental release number within the month (first release is 1, second is 2, etc).
Because features are added, issues are fixed, or simply because a new version of the containerized application is integrated, the Docker image is regularly updated. Different methods can be used to update the Docker image.
The system used to run the container may have a built-in way to update containers. If so, this could be your primary way to update Docker images.
An other way is to have the image be automatically updated with Watchtower. Watchtower is a container-based solution for automating Docker image updates. This is a "set and forget" type of solution: once a new image is available, Watchtower will seamlessly perform the necessary steps to update the container.
Finally, the Docker image can be manually updated with these steps:
- Fetch the latest image:
docker pull jlesage/firefox- Stop the container:
docker stop firefox- Remove the container:
docker rm firefox- Create and start the container using the
docker runcommand, with the the same parameters that were used when it was deployed initially.
For owners of a Synology NAS, the following steps can be used to update a container image.
- Open the Docker application.
- Click on Registry in the left pane.
- In the search bar, type the name of the container (
jlesage/firefox). - Select the image, click Download and then choose the
latesttag. - Wait for the download to complete. A notification will appear once done.
- Click on Container in the left pane.
- Select your Firefox container.
- Stop it by clicking Action->Stop.
- Clear the container by clicking Action->Reset (or Action->Clear if you don't have the latest Docker application). This removes the container while keeping its configuration.
- Start the container again by clicking Action->Start. NOTE: The container may temporarily disappear from the list while it is re-created.
For unRAID, a container image can be updated by following these steps:
- Select the Docker tab.
- Click the Check for Updates button at the bottom of the page.
- Click the update ready link of the container to be updated.
When using data volumes (-v flags), permissions issues can occur between the
host and the container. For example, the user within the container may not
exist on the host. This could prevent the host from properly accessing files
and folders on the shared volume.
To avoid any problem, you can specify the user the application should run as.
This is done by passing the user ID and group ID to the container via the
USER_ID and GROUP_ID environment variables.
To find the right IDs to use, issue the following command on the host, with the user owning the data volume on the host:
id <username>
Which gives an output like this one:
uid=1000(myuser) gid=1000(myuser) groups=1000(myuser),4(adm),24(cdrom),27(sudo),46(plugdev),113(lpadmin)
The value of uid (user ID) and gid (group ID) are the ones that you should
be given the container.
Assuming that container's ports are mapped to the same host's ports, the graphical interface of the application can be accessed via:
- A web browser:
http://<HOST IP ADDR>:5800
- Any VNC client:
<HOST IP ADDR>:5900
By default, access to the application's GUI is done over an unencrypted connection (HTTP or VNC).
Secure connection can be enabled via the SECURE_CONNECTION environment
variable. See the Environment Variables section for
more details on how to set an environment variable.
When enabled, application's GUI is performed over an HTTPs connection when accessed with a browser. All HTTP accesses are automatically redirected to HTTPs.
When using a VNC client, the VNC connection is performed over SSL. Note that few VNC clients support this method. SSVNC is one of them.
SSVNC is a VNC viewer that adds encryption security to VNC connections.
While the Linux version of SSVNC works well, the Windows version has some
issues. At the time of writing, the latest version 1.0.30 is not functional,
as a connection fails with the following error:
ReadExact: Socket error while reading
However, for your convenience, an unofficial and working version is provided here:
https://github.com/jlesage/docker-baseimage-gui/raw/master/tools/ssvnc_windows_only-1.0.30-r1.zip
The only difference with the official package is that the bundled version of
stunnel has been upgraded to version 5.49, which fixes the connection
problems.
Here are the certificate files needed by the container. By default, when they are missing, self-signed certificates are generated and used. All files have PEM encoded, x509 certificates.
| Container Path | Purpose | Content |
|---|---|---|
/config/certs/vnc-server.pem |
VNC connection encryption. | VNC server's private key and certificate, bundled with any root and intermediate certificates. |
/config/certs/web-privkey.pem |
HTTPs connection encryption. | Web server's private key. |
/config/certs/web-fullchain.pem |
HTTPs connection encryption. | Web server's certificate, bundled with any root and intermediate certificates. |
NOTE: To prevent any certificate validity warnings/errors from the browser or VNC client, make sure to supply your own valid certificates.
NOTE: Certificate files are monitored and relevant daemons are automatically restarted when changes are detected.
To restrict access to your application, a password can be specified. This can be done via two methods:
- By using the
VNC_PASSWORDenvironment variable. - By creating a
.vncpass_clearfile at the root of the/configvolume. This file should contain the password in clear-text. During the container startup, content of the file is obfuscated and moved to.vncpass.
The level of security provided by the VNC password depends on two things:
- The type of communication channel (encrypted/unencrypted).
- How secure the access to the host is.
When using a VNC password, it is highly desirable to enable the secure connection to prevent sending the password in clear over an unencrypted channel.
ATTENTION: Password is limited to 8 characters. This limitation comes from the Remote Framebuffer Protocol RFC (see section 7.2.2). Any characters beyond the limit are ignored.
The following sections contain NGINX configurations that need to be added in order to reverse proxy to this container.
A reverse proxy server can route HTTP requests based on the hostname or the URL path.
In this scenario, each hostname is routed to a different application/container.
For example, let's say the reverse proxy server is running on the same machine
as this container. The server would proxy all HTTP requests sent to
firefox.domain.tld to the container at 127.0.0.1:5800.
Here are the relevant configuration elements that would be added to the NGINX configuration:
map $http_upgrade $connection_upgrade {
default upgrade;
'' close;
}
upstream docker-firefox {
# If the reverse proxy server is not running on the same machine as the
# Docker container, use the IP of the Docker host here.
# Make sure to adjust the port according to how port 5800 of the
# container has been mapped on the host.
server 127.0.0.1:5800;
}
server {
[...]
server_name firefox.domain.tld;
location / {
proxy_pass http://docker-firefox;
}
location /websockify {
proxy_pass http://docker-firefox;
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection $connection_upgrade;
proxy_read_timeout 86400;
}
}
In this scenario, the hostname is the same, but different URL paths are used to route to different applications/containers.
For example, let's say the reverse proxy server is running on the same machine
as this container. The server would proxy all HTTP requests for
server.domain.tld/firefox to the container at 127.0.0.1:5800.
Here are the relevant configuration elements that would be added to the NGINX configuration:
map $http_upgrade $connection_upgrade {
default upgrade;
'' close;
}
upstream docker-firefox {
# If the reverse proxy server is not running on the same machine as the
# Docker container, use the IP of the Docker host here.
# Make sure to adjust the port according to how port 5800 of the
# container has been mapped on the host.
server 127.0.0.1:5800;
}
server {
[...]
location = /firefox {return 301 $scheme://$http_host/firefox/;}
location /firefox/ {
proxy_pass http://docker-firefox/;
location /firefox/websockify {
proxy_pass http://docker-firefox/websockify/;
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection $connection_upgrade;
proxy_read_timeout 86400;
}
}
}
To get shell access to the running container, execute the following command:
docker exec -ti CONTAINER shWhere CONTAINER is the ID or the name of the container used during its
creation.
To properly work, recent versions of Firefox need the
membarrier system call. Without it, tabs would frequently crash.
Docker uses seccomp profile to restrict system calls available to the
container. Before Docker version 20.10.0, the membarrier system call was
not allowed in the default profile. If you run a such version, you can use one
of the following solutions, from the most to the least secure, to provide the
container permission to use this sytem call:
- Run the container with a custom seccomp profile allowing the
membarriersystem call. The latest official seccomp profile can be used. Download the file and then add the following parameter when creating the container:--security-opt seccomp=/path/to/seccomp_profile.json. - Run the container without the default seccomp profile (thus allowing all
system calls). Use the following parameter when creating the container:
--security-opt seccomp=unconfined. - Run the container in privileged mode. This effectively disables usage of
seccomp. Add the
--privilegedparameter when creating the container.
For Firefox to be able to use the audio device available on
the host, /dev/snd must be exposed to the container by adding the
--device /dev/snd parameter to the docker run command.
Firefox preferences can be set via environment variables passed to the container. During the startup, a script process all these variables and modify the preference file accordingly.
The name of the environment variable must start with FF_PREF_, followed by a
string of your choice. For example, FF_PREF_MY_PREF is a valid name.
The content of the variable should be in the format NAME=VAL, where NAME is
the name of the preference (as found in the about:config page) and VAL is
its value. A value can be one of the following types:
- string
- integer
- boolean
It is important to note that a value of type string should be surrounded by
double quotes. Other types don't need them.
For example, to set the network.proxy.http preference, one would pass the
environment variable to the container by adding the following argument to the
docker run command:
-e "FF_PREF_HTTP_PROXY=network.proxy.http=\"proxy.example.com\""
If a preference needs to be removed, its value should be set to UNSET. For
example:
-e "FF_PREF_HTTP_PROXY=network.proxy.http=UNSET"
NOTE: This is an advanced usage and it is recommended to set preferences via Firefox directly.
If Firefox is crashing frequently, make sure that:
- The
membarriersystem call is not blocked by Docker. See the Allowing the membarrier System Call for more details. - Make sure the kernel of your Linux distribution is up-to-date.
Having troubles with the container or have questions? Please create a new issue.
For other great Dockerized applications, see https://jlesage.github.io/docker-apps.