Change-type: patch Signed-off-by: Scott Lowe <scott@balena.io>
74 KiB
Balena CLI Documentation
The balena CLI (Command-Line Interface) allows you to interact with the balenaCloud and the balena API through a terminal window on Linux, macOS or Windows. You can also write shell scripts around it, or import its Node.js modules to use it programmatically. As an open-source project on GitHub, your contribution is also welcome!
Installation
Check the balena CLI installation instructions on GitHub.
Getting Started
Choosing a shell (command prompt/terminal)
On Windows, the standard Command Prompt (cmd.exe
) and
PowerShell
are supported. We are aware of users also having a good experience with alternative shells,
including:
- MSYS2:
- Install additional packages with the command:
pacman -S git openssh rsync
- Set a Windows environment variable:
MSYS2_PATH_TYPE=inherit
- Note that a bug in the MSYS2 launch script (
msys2_shell.cmd
) makes text-based interactive CLI menus to break. Check this Github issue for a workaround.
- Install additional packages with the command:
- MSYS: select the
msys-rsync
andmsys-openssh
packages too - Git for Windows
- Microsoft's Windows Subsystem for Linux (WSL). In this case, a Linux distribution like Ubuntu is installed via the Microsoft Store, and a balena CLI release for Linux is recommended. See FAQ for using balena CLI with WSL and Docker Desktop for Windows.
On macOS and Linux, the standard terminal window is supported. Optionally, bash
command
auto completion may be enabled by copying the
balena-completion.bash
file to your system's bash_completion
directory: check Docker's command completion
guide for system setup instructions.
Logging in
Several CLI commands require access to your balenaCloud account, for example in order to push a new release to your application. Those commands require creating a CLI login session by running:
$ balena login
Proxy support
HTTP(S) proxies can be configured through any of the following methods, in precedence order (from higher to lower):
-
The
BALENARC_PROXY
environment variable in URL format, with protocol (http
orhttps
), host, port and optionally basic auth. Examples:export BALENARC_PROXY='https://bob:secret@proxy.company.com:12345'
export BALENARC_PROXY='http://localhost:8000'
-
The
proxy
setting in the CLI config file. It may be:-
A string in URL format, e.g.
proxy: 'http://localhost:8000'
-
An object in the format:
proxy: protocol: 'http' host: 'proxy.company.com' port: 12345 proxyAuth: 'bob:secret'
-
-
The
HTTPS_PROXY
and/orHTTP_PROXY
environment variables, in the same URL format asBALENARC_PROXY
.
Note: The
balena ssh
command has additional setup requirements to work behind a proxy. Check the installation instructions, and ensure that the proxy server is configured to allow proxy requests to ssh port 22, using SSL encryption. For example, in the case of the Squid proxy server, it should be configured with the following rules in thesquid.conf
file:
acl SSL_ports port 22
acl Safe_ports port 22
Proxy exclusion
The BALENARC_NO_PROXY
variable may be used to exclude specified destinations from proxying.
- This feature requires balena CLI version 11.30.8 or later. In the case of the npm installation option, it also requires Node.js version 10.16.0 or later.
- To exclude a
balena ssh
target from proxying (IP address or.local
hostname), the--noproxy
option should be specified in addition to theBALENARC_NO_PROXY
variable.
By default (if BALENARC_NO_PROXY
is not defined), all private IPv4
addresses and '*.local'
hostnames are excluded
from proxying. Other hostnames that resolve to private IPv4 addresses are not excluded by
default, because matching takes place before name resolution.
localhost
and 127.0.0.1
are always excluded from proxying, regardless of the value of
BALENARC_NO_PROXY.
The format of the BALENARC_NO_PROXY
environment variable is a comma-separated list of patterns
that are matched against hostnames or IP addresses. For example:
export BALENARC_NO_PROXY='*.local,dev*.mycompany.com,192.168.*'
Matched patterns are excluded from proxying. Wildcard expressions are documented at
matcher. Matching takes place before name
resolution, so a pattern like '192.168.*'
will not match a hostname that resolves to an IP
address like 192.168.1.2
.
Support, FAQ and troubleshooting
If you come across any problems or would like to get in touch:
- Check our FAQ / troubleshooting document.
- Ask us a question through the balenaCloud forum.
- For bug reports or feature requests, have a look at the GitHub issues or create a new one.
Deprecation policy
The balena CLI uses semver versioning, with the concepts of major, minor and patch version releases.
The latest release of the previous major version of the balena CLI will remain compatible with the balenaCloud backend services for one year from the date when the next major version is released. For example, balena CLI v10.17.5, as the latest v10 release, would remain compatible with the balenaCloud backend for one year from the date when v11.0.0 is released.
At the end of this period, the older major version is considered deprecated and some of the functionality that depends on balenaCloud services may stop working at any time. Users are encouraged to regularly update the balena CLI to the latest version.
CLI Command Reference
-
API keys
-
Application
-
Authentication
-
Device
-
Environment Variables
-
Tags
-
Help and Version
-
Keys
-
Logs
-
Network
-
Notes
-
OS
-
Config
-
Preload
-
Push
-
Settings
-
Local
-
Deploy
-
Platform
-
Utilities
API keys
api-key generate <name>
Generate a new balenaCloud API key for the current user, with the given name. The key will be logged to the console.
This key can be used to log into the CLI using 'balena login --token ', or to authenticate requests to the API with an 'Authorization: Bearer ' header.
Examples:
$ balena api-key generate "Jenkins Key"
Arguments
NAME
the API key name
Options
Application
apps
list all your balena applications.
For detailed information on a particular application,
use balena app <name> instead
.
Examples:
$ balena apps
Options
-v, --verbose
No-op since release v12.0.0
app <name>
Display detailed information about a single balena application.
Examples:
$ balena app MyApp
Arguments
NAME
application name or numeric ID
Options
app create <name>
Create a new balena application.
You can specify the application device type with the --type
option.
Otherwise, an interactive dropdown will be shown for you to select from.
You can see a list of supported device types with:
$ balena devices supported
Examples:
$ balena app create MyApp
$ balena app create MyApp --type raspberry-pi
Arguments
NAME
application name
Options
-t, --type TYPE
application device type (Check available types with balena devices supported
)
app rm <name>
Permanently remove a balena application.
The --yes option may be used to avoid interactive confirmation.
Examples:
$ balena app rm MyApp
$ balena app rm MyApp --yes
Arguments
NAME
application name or numeric ID
Options
-y, --yes
answer "yes" to all questions (non interactive use)
app restart <name>
Restart all devices that belongs to a certain application.
Examples:
$ balena app restart MyApp
Arguments
NAME
application name or numeric ID
Options
Authentication
login
Login to your balena account.
This command will prompt you to login using the following login types:
-
Web authorization: open your web browser and prompt to authorize the CLI from the dashboard.
-
Credentials: using email/password and 2FA.
-
Token: using a session token or API key from the preferences page.
Examples:
$ balena login
$ balena login --web
$ balena login --token "..."
$ balena login --credentials
$ balena login --credentials --email johndoe@gmail.com --password secret
Arguments
TOKEN
Options
-w, --web
web-based login
-t, --token
session token or API key
-c, --credentials
credential-based login
-e, --email EMAIL
-u, --user USER
-p, --password PASSWORD
password
logout
Logout from your balena account.
Examples:
$ balena logout
whoami
Get the username and email address of the currently logged in user.
Examples:
$ balena whoami
Device
device identify <uuid>
Identify a device by making the ACT LED blink (Raspberry Pi).
Examples:
$ balena device identify 23c73a1
Arguments
UUID
the uuid of the device to identify
Options
device init
Initialise a device by downloading the OS image of a certain application and writing it to an SD Card.
Note, if the application option is omitted it will be prompted for interactively.
Examples:
$ balena device init
$ balena device init --application MyApp
Options
-a, --application APPLICATION
application name
--app APP
same as '--application'
-y, --yes
answer "yes" to all questions (non interactive use)
-v, --advanced
show advanced configuration options
--os-version OS-VERSION
exact version number, or a valid semver range, or 'latest' (includes pre-releases), or 'default' (excludes pre-releases if at least one stable version is available), or 'recommended' (excludes pre-releases, will fail if only pre-release versions are available), or 'menu' (will show the interactive menu)
-d, --drive DRIVE
the drive to write the image to, eg. /dev/sdb
or /dev/mmcblk0
. Careful with this as you can erase your hard drive. Check balena util available-drives
for available options.
--config CONFIG
path to the config JSON file, see balena os build-config
device <uuid>
Show information about a single device.
Examples:
$ balena device 7cf02a6
Arguments
UUID
the device uuid
Options
device move <uuid>
Move a device to another application.
Note, if the application option is omitted it will be prompted for interactively.
Examples:
$ balena device move 7cf02a6
$ balena device move 7cf02a6 --application MyNewApp
Arguments
UUID
the uuid of the device to move
Options
-a, --application APPLICATION
application name
--app APP
same as '--application'
device reboot <uuid>
Remotely reboot a device.
Examples:
$ balena device reboot 23c73a1
Arguments
UUID
the uuid of the device to reboot
Options
-f, --force
force action if the update lock is set
device register <application>
Register a device to an application.
Examples:
$ balena device register MyApp
$ balena device register MyApp --uuid <uuid>
Arguments
APPLICATION
the name or id of application to register device with
Options
-u, --uuid UUID
custom uuid
device rename <uuid> [newName]
Rename a device.
Note, if the name is omitted, it will be prompted for interactively.
Examples:
$ balena device rename 7cf02a6
$ balena device rename 7cf02a6 MyPi
Arguments
UUID
the uuid of the device to rename
NEWNAME
the new name for the device
Options
device rm <uuid>
Remove a device from balena.
Note this command asks for confirmation interactively.
You can avoid this by passing the --yes
option.
Examples:
$ balena device rm 7cf02a6
$ balena device rm 7cf02a6 --yes
Arguments
UUID
the uuid of the device to remove
Options
-y, --yes
answer "yes" to all questions (non interactive use)
device shutdown <uuid>
Remotely shutdown a device.
Examples:
$ balena device shutdown 23c73a1
Arguments
UUID
the uuid of the device to shutdown
Options
-f, --force
force action if the update lock is set
devices
list all devices that belong to you.
You can filter the devices by application by using the --application
option.
Examples:
$ balena devices
$ balena devices --application MyApp
$ balena devices --app MyApp
$ balena devices -a MyApp
Options
-a, --application APPLICATION
application name
--app APP
same as '--application'
devices supported
List the supported device types (like 'raspberrypi3' or 'intel-nuc').
The --verbose option adds extra columns/fields to the output, including the "STATE" column whose values are one of 'beta', 'released' or 'discontinued'. However, 'discontinued' device types are only listed if the '--discontinued' option is used.
The --json option is recommended when scripting the output of this command, because the JSON format is less likely to change and it better represents data types like lists and empty strings (for example, the ALIASES column contains a list of zero or more values). The 'jq' utility may be helpful in shell scripts (https://stedolan.github.io/jq/manual/).
Examples:
$ balena devices supported
$ balena devices supported --verbose
$ balena devices supported -vj
Options
--discontinued
include "discontinued" device types
-j, --json
produce JSON output instead of tabular output
-v, --verbose
add extra columns in the tabular output (ALIASES, ARCH, STATE)
device os-update <uuid>
Start a Host OS update for a device.
Note this command will ask for confirmation interactively.
This can be avoided by passing the --yes
option.
Requires balenaCloud; will not work with openBalena or standalone balenaOS.
Examples:
$ balena device os-update 23c73a1
$ balena device os-update 23c73a1 --version 2.31.0+rev1.prod
Arguments
UUID
the uuid of the device to update
Options
--version VERSION
a balenaOS version
-y, --yes
answer "yes" to all questions (non interactive use)
device public-url <uuid>
This command will output the current public URL for the specified device. It can also enable or disable the URL, or output the enabled status, using the respective options.
The old command style 'balena device public-url enable ' is deprecated, but still supported.
Examples:
$ balena device public-url 23c73a1
$ balena device public-url 23c73a1 --enable
$ balena device public-url 23c73a1 --disable
$ balena device public-url 23c73a1 --status
Arguments
UUID
the uuid of the device to manage
LEGACYUUID
Options
--enable
enable the public URL
--disable
disable the public URL
--status
determine if public URL is enabled
Environment Variables
envs
List the environment or configuration variables of an application, device or service, as selected by the respective command-line options. (A service is an application container in a "microservices" application.)
The results include application-wide (fleet), device-wide (multiple services on a device) and service-specific variables that apply to the selected application, device or service. It can be thought of as including "inherited" variables; for example, a service inherits device-wide variables, and a device inherits application-wide variables.
The printed output may include DEVICE and/or SERVICE columns to distinguish between application-wide, device-specific and service-specific variables. An asterisk in these columns indicates that the variable applies to "all devices" or "all services".
The --config option is used to list "configuration variables" that control balena platform features, as opposed to custom environment variables defined by the user. The --config and the --service options are mutually exclusive because configuration variables cannot be set for specific services.
The --json option is recommended when scripting the output of this command, because the JSON format is less likely to change and it better represents data types like lists and empty strings. The 'jq' utility may be helpful in shell scripts (https://stedolan.github.io/jq/manual/). When --json is used, an empty JSON array ([]) is printed instead of an error message when no variables exist for the given query. When querying variables for a device, note that the application name may be null in JSON output (or 'N/A' in tabular output) if the application linked to the device is no longer accessible by the current user (for example, in case the current user has been removed from the application by its owner).
Examples:
$ balena envs --application MyApp
$ balena envs --application MyApp --json
$ balena envs --application MyApp --service MyService
$ balena envs --application MyApp --service MyService
$ balena envs --application MyApp --config
$ balena envs --device 7cf02a6
$ balena envs --device 7cf02a6 --json
$ balena envs --device 7cf02a6 --config --json
$ balena envs --device 7cf02a6 --service MyService
Options
--all
No-op since balena CLI v12.0.0.
-a, --application APPLICATION
application name
-c, --config
show configuration variables only
-d, --device DEVICE
device UUID
-j, --json
produce JSON output instead of tabular output
-v, --verbose
produce verbose output
-s, --service SERVICE
service name
env rm ID
Remove a configuration or environment variable from an application, device or service, as selected by command-line options.
Variables are selected by their database ID (as reported by the 'balena envs' command) and one of six database "resource types":
- application (fleet) environment variable
- application (fleet) configuration variable (--config)
- application (fleet) service variable (--service)
- device environment variable (--device)
- device configuration variable (--device --config)
- device service variable (--device --service)
The --device option selects a device-specific variable instead of an application (fleet) variable.
The --config option selects a configuration variable. Configuration variable names typically start with the 'BALENA_' or 'RESIN_' prefixes and are used to configure balena platform features.
The --service option selects a service variable, which is an environment variable that applies to a specifc service (application container) in a microservices (multicontainer) application.
The --service and --config options cannot be used together, but they can be used alongside the --device option to select a device-specific service or configuration variable.
Interactive confirmation is normally asked before the variable is deleted. The --yes option disables this behavior.
Examples:
$ balena env rm 123123
$ balena env rm 234234 --yes
$ balena env rm 345345 --config
$ balena env rm 456456 --service
$ balena env rm 567567 --device
$ balena env rm 678678 --device --config
$ balena env rm 789789 --device --service --yes
Arguments
ID
variable's numeric database ID
Options
-c, --config
select a configuration variable (may be used together with the --device option)
-d, --device
select a device-specific variable instead of an application (fleet) variable
-s, --service
select a service variable (may be used together with the --device option)
-y, --yes
do not prompt for confirmation before deleting the variable
env add NAME [VALUE]
Add an environment or config variable to an application, device or service, as selected by the respective command-line options. Either the --application or the --device option must be provided, and either may be be used alongside the --service option to define a service-specific variable. (A service is an application container in a "microservices" application.) When the --service option is used in conjunction with the --device option, the service variable applies to the selected device only. Otherwise, it applies to all devices of the selected application (i.e., the application's fleet). If the --service option is omitted, the variable applies to all services.
If VALUE is omitted, the CLI will attempt to use the value of the environment
variable of same name in the CLI process' environment. In this case, a warning
message will be printed. Use --quiet
to suppress it.
'BALENA_' or 'RESIN_' are reserved variable name prefixes used to identify "configuration variables". Configuration variables control balena platform features and are treated specially by balenaOS and the balena supervisor running on devices. They are also stored differently in the balenaCloud API database. Configuration variables cannot be set for specific services, therefore the --service option cannot be used when the variable name starts with a reserved prefix. When defining custom application variables, please avoid the reserved prefixes.
Examples:
$ balena env add TERM --application MyApp
$ balena env add EDITOR vim --application MyApp
$ balena env add EDITOR vim --application MyApp --service MyService
$ balena env add EDITOR vim --device 7cf02a6
$ balena env add EDITOR vim --device 7cf02a6 --service MyService
Arguments
NAME
environment or config variable name
VALUE
variable value; if omitted, use value from this process' environment
Options
-a, --application APPLICATION
application name
-d, --device DEVICE
device UUID
-q, --quiet
suppress warning messages
-s, --service SERVICE
service name
env rename ID VALUE
Change the value of a configuration or environment variable for an application, device or service, as selected by command-line options.
Variables are selected by their database ID (as reported by the 'balena envs' command) and one of six database "resource types":
- application (fleet) environment variable
- application (fleet) configuration variable (--config)
- application (fleet) service variable (--service)
- device environment variable (--device)
- device configuration variable (--device --config)
- device service variable (--device --service)
The --device option selects a device-specific variable instead of an application (fleet) variable.
The --config option selects a configuration variable. Configuration variable names typically start with the 'BALENA_' or 'RESIN_' prefixes and are used to configure balena platform features.
The --service option selects a service variable, which is an environment variable that applies to a specifc service (application container) in a microservices (multicontainer) application.
The --service and --config options cannot be used together, but they can be used alongside the --device option to select a device-specific service or configuration variable.
Examples:
$ balena env rename 123123 emacs
$ balena env rename 234234 emacs --service
$ balena env rename 345345 emacs --device
$ balena env rename 456456 emacs --device --service
$ balena env rename 567567 1 --config
$ balena env rename 678678 1 --device --config
Arguments
ID
variable's numeric database ID
VALUE
variable value; if omitted, use value from this process' environment
Options
-c, --config
select a configuration variable (may be used together with the --device option)
-d, --device
select a device-specific variable instead of an application (fleet) variable
-s, --service
select a service variable (may be used together with the --device option)
Tags
tags
List all tags and their values for a particular application, device or release.
Examples:
$ balena tags --application MyApp
$ balena tags --device 7cf02a6
$ balena tags --release 1234
$ balena tags --release b376b0e544e9429483b656490e5b9443b4349bd6
Options
-a, --application APPLICATION
application name
-d, --device DEVICE
device UUID
-r, --release RELEASE
release id
--app APP
same as '--application'
tag rm <tagKey>
Remove a tag from an application, device or release.
Examples:
$ balena tag rm myTagKey --application MyApp
$ balena tag rm myTagKey --device 7cf02a6
$ balena tag rm myTagKey --release 1234
$ balena tag rm myTagKey --release b376b0e544e9429483b656490e5b9443b4349bd6
Arguments
TAGKEY
the key string of the tag
Options
-a, --application APPLICATION
application name
-d, --device DEVICE
device UUID
-r, --release RELEASE
release id
--app APP
same as '--application'
tag set <tagKey> [value]
Set a tag on an application, device or release.
You can optionally provide a value to be associated with the created tag, as an extra argument after the tag key. If a value isn't provided, a tag with an empty value is created.
Examples:
$ balena tag set mySimpleTag --application MyApp
$ balena tag set myCompositeTag myTagValue --application MyApp
$ balena tag set myCompositeTag myTagValue --device 7cf02a6
$ balena tag set myCompositeTag "my tag value with whitespaces" --device 7cf02a6
$ balena tag set myCompositeTag myTagValue --release 1234
$ balena tag set myCompositeTag --release 1234
$ balena tag set myCompositeTag --release b376b0e544e9429483b656490e5b9443b4349bd6
Arguments
TAGKEY
the key string of the tag
VALUE
the optional value associated with the tag
Options
-a, --application APPLICATION
application name
-d, --device DEVICE
device UUID
-r, --release RELEASE
release id
--app APP
same as '--application'
Help and Version
help [command...]
Get detailed help for an specific command.
Examples:
$ balena help apps
$ balena help os download
Options
--verbose, -v
show additional commands
version
Display version information for the balena CLI and/or Node.js.
The --json option is recommended when scripting the output of this command, because the JSON format is less likely to change and it better represents data types like lists and empty strings. The 'jq' utility may be helpful in shell scripts (https://stedolan.github.io/jq/manual/).
Examples:
$ balena version
$ balena version -a
$ balena version -j
Options
-a, --all
include version information for additional components (Node.js)
-j, --json
output version information in JSON format for programmatic use
Keys
keys
List all SSH keys registered in balenaCloud for the logged in user.
Examples:
$ balena keys
Options
key <id>
Display a single SSH key registered in balenaCloud for the logged in user.
Examples:
$ balena key 17
Arguments
ID
balenaCloud ID for the SSH key
Options
key add <name> [path]
Register an SSH in balenaCloud for the logged in user.
If path
is omitted, the command will attempt
to read the SSH key from stdin.
Examples:
$ balena key add Main ~/.ssh/id_rsa.pub
$ cat ~/.ssh/id_rsa.pub | balena key add Main
Arguments
NAME
the SSH key name
PATH
the path to the public key file
Options
key rm <id>
Remove a single SSH key registered in balenaCloud for the logged in user.
The --yes option may be used to avoid interactive confirmation.
Examples:
$ balena key rm 17
$ balena key rm 17 --yes
Arguments
ID
balenaCloud ID for the SSH key
Options
-y, --yes
answer "yes" to all questions (non interactive use)
Logs
logs <device>
Show logs for a specific device.
By default, the command prints all log messages and exits.
To continuously stream output, and see new logs in real time, use the --tail
option.
If an IP or .local address is passed to this command, logs are displayed from a local mode device with that address. Note that --tail is implied when this command is provided a local mode device.
Logs from a single service can be displayed with the --service flag. Just system logs can be shown with the --system flag. Note that these flags can be used together.
Note: --service and --system flags must come after the device parameter, as per examples.
Examples:
$ balena logs 23c73a1
$ balena logs 23c73a1 --tail
$ balena logs 192.168.0.31
$ balena logs 192.168.0.31 --service my-service
$ balena logs 192.168.0.31 --service my-service-1 --service my-service-2
$ balena logs 23c73a1.local --system
$ balena logs 23c73a1.local --system --service my-service
Arguments
DEVICE
device UUID, IP, or .local address
Options
-t, --tail
continuously stream output
-s, --service SERVICE
Reject logs not originating from this service. This can be used in combination with --system or other --service flags.
-S, --system
Only show system logs. This can be used in combination with --service.
Network
scan
Scan for balenaOS devices on your local network.
Examples:
$ balena scan
$ balena scan --timeout 120
$ balena scan --verbose
Options
-v, --verbose
display full info
-t, --timeout TIMEOUT
scan timeout in seconds
ssh <applicationOrDevice> [serviceName]
Start a shell on a local or remote device. If a service name is not provided, a shell will be opened on the host OS.
If an application name is provided, an interactive menu will be presented for the selection of an online device. A shell will then be opened for the host OS or service container of the chosen device.
For local devices, the IP address and .local domain name are supported.
If the device is referenced by IP or .local
address, the connection
is initiated directly to balenaOS on port 22222
via an
openssh-compatible client. Otherwise, any connection initiated remotely
traverses the balenaCloud VPN.
Commands may be piped to the standard input for remote execution (see examples). Note however that remote command execution on service containers (as opposed to the host OS) is not currently possible when a device UUID is used (instead of an IP address) because of a balenaCloud backend limitation.
Note: balena ssh
requires an openssh-compatible client to be correctly
installed in your shell environment. For more information (including Windows
support) please check:
https://github.com/balena-io/balena-cli/blob/master/INSTALL.md#additional-dependencies,
Examples:
$ balena ssh MyApp
$ balena ssh f49cefd
$ balena ssh f49cefd my-service
$ balena ssh f49cefd --port <port>
$ balena ssh 192.168.0.1 --verbose
$ balena ssh f49cefd.local my-service
$ echo "uptime; exit;" | balena ssh f49cefd
$ echo "uptime; exit;" | balena ssh 192.168.0.1 myService
Arguments
APPLICATIONORDEVICE
application name, device uuid, or address of local device
SERVICENAME
service name, if connecting to a container
Options
-p, --port PORT
SSH server port number (default 22222) if the target is an IP address or .local hostname. Otherwise, port number for the balenaCloud gateway (default 22).
-t, --tty
Force pseudo-terminal allocation (bypass TTY autodetection for stdin)
-v, --verbose
Increase verbosity
--noproxy
Bypass global proxy configuration for the ssh connection
tunnel <deviceOrApplication>
Use this command to open local ports which tunnel to listening ports on your balenaOS device.
For example, you could open port 8080 on your local machine to connect to your managed balenaOS device running a web server listening on port 3000.
Port mappings are specified in the format: [:[localIP:]localPort] localIP defaults to 'localhost', and localPort defaults to the specified remotePort value.
You can tunnel multiple ports at any given time.
Note: Port mappings must come after the deviceOrApplication parameter, as per examples.
Examples:
# map remote port 22222 to localhost:22222
$ balena tunnel myApp -p 22222
# map remote port 22222 to localhost:222
$ balena tunnel 2ead211 -p 22222:222
# map remote port 22222 to any address on your host machine, port 22222
$ balena tunnel 1546690 -p 22222:0.0.0.0
# map remote port 22222 to any address on your host machine, port 222
$ balena tunnel myApp -p 22222:0.0.0.0:222
# multiple port tunnels can be specified at any one time
$ balena tunnel myApp -p 8080:3000 -p 8081:9000
Arguments
DEVICEORAPPLICATION
device uuid or application name/id
Options
-p, --port PORT
port mapping in the format [:[localIP:]localPort]
Notes
note <|note>
Set or update a device note. If the note argument is not provided, it will be read from stdin.
To view device notes, use the balena device <uuid>
command.
Examples:
$ balena note "My useful note" --device 7cf02a6
$ cat note.txt | balena note --device 7cf02a6
Arguments
NOTE
note content
Options
-d, --device DEVICE
device UUID
--dev DEV
OS
os versions <type>
Use this command to show the available balenaOS versions for a certain device type.
Check available types with balena devices supported
Example:
$ balena os versions raspberrypi3
os download <type>
Use this command to download an unconfigured os image for a certain device type.
Check available types with balena devices supported
Note: Currently this command only works with balenaCloud, not openBalena. If using openBalena, please download the OS from: https://www.balena.io/os/
If version is not specified the newest stable (non-pre-release) version of OS is downloaded if available, or the newest version otherwise (if all existing versions for the given device type are pre-release).
You can pass --version menu
to pick the OS version from the interactive menu
of all available versions.
Examples:
$ balena os download raspberrypi3 -o ../foo/bar/raspberry-pi.img
$ balena os download raspberrypi3 -o ../foo/bar/raspberry-pi.img --version 1.24.1
$ balena os download raspberrypi3 -o ../foo/bar/raspberry-pi.img --version ^1.20.0
$ balena os download raspberrypi3 -o ../foo/bar/raspberry-pi.img --version latest
$ balena os download raspberrypi3 -o ../foo/bar/raspberry-pi.img --version default
$ balena os download raspberrypi3 -o ../foo/bar/raspberry-pi.img --version menu
Options
--output, -o <output>
output path
--version <version>
exact version number, or a valid semver range, or 'latest' (includes pre-releases), or 'default' (excludes pre-releases if at least one stable version is available), or 'recommended' (excludes pre-releases, will fail if only pre-release versions are available), or 'menu' (will show the interactive menu)
os build-config <image> <device-type>
Use this command to prebuild the OS config once and skip the interactive part of balena os configure
.
Example:
$ balena os build-config ../path/rpi3.img raspberrypi3 --output rpi3-config.json
$ balena os configure ../path/rpi3.img --device 7cf02a6 --config rpi3-config.json
Options
--advanced, -v
show advanced configuration options
--output, -o <output>
the path to the output JSON file
os configure IMAGE
Configure a previously downloaded balenaOS image for a specific device type or balena application.
Configuration settings such as WiFi authentication will be taken from the following sources, in precedence order:
- Command-line options like
--config-wifi-ssid
- A given
config.json
file specified with the--config
option. - User input through interactive prompts (text menus).
The --device-type option may be used to override the application's default device type, in case of an application with mixed device types.
The --system-connection (-c) option can be used to inject NetworkManager connection profiles for additional network interfaces, such as cellular/GSM or additional WiFi or ethernet connections. This option may be passed multiple times in case there are multiple files to inject. See connection profile examples and reference at: https://www.balena.io/docs/reference/OS/network/2.x/ https://developer.gnome.org/NetworkManager/stable/nm-settings.html
The --device-api-key option is deprecated and will be removed in a future release. A suitable key is automatically generated or fetched if this option is omitted.
Note: This command is currently not supported on Windows natively. Windows users are advised to install the Windows Subsystem for Linux (WSL) with Ubuntu, and use the Linux release of the balena CLI: https://docs.microsoft.com/en-us/windows/wsl/about
Examples:
$ balena os configure ../path/rpi3.img --device 7cf02a6
$ balena os configure ../path/rpi3.img --device 7cf02a6 --device-api-key <existingDeviceKey>
$ balena os configure ../path/rpi3.img --app MyApp
$ balena os configure ../path/rpi3.img --app MyApp --version 2.12.7
$ balena os configure ../path/rpi3.img --app MyFinApp --device-type raspberrypi3
$ balena os configure ../path/rpi3.img --app MyFinApp --device-type raspberrypi3 --config myWifiConfig.json
Arguments
IMAGE
path to a balenaOS image file, e.g. "rpi3.img"
Options
-v, --advanced
ask advanced configuration questions (when in interactive mode)
--app APP
same as '--application'
-a, --application APPLICATION
application name
--config CONFIG
path to a pre-generated config.json file to be injected in the OS image
--config-app-update-poll-interval CONFIG-APP-UPDATE-POLL-INTERVAL
interval (in minutes) for the on-device balena supervisor periodic app update check
--config-network CONFIG-NETWORK
device network type (non-interactive configuration)
--config-wifi-key CONFIG-WIFI-KEY
WiFi key (password) (non-interactive configuration)
--config-wifi-ssid CONFIG-WIFI-SSID
WiFi SSID (network name) (non-interactive configuration)
-d, --device DEVICE
device UUID
-k, --device-api-key DEVICE-API-KEY
custom device API key (DEPRECATED and only supported with balenaOS 2.0.3+)
--device-type DEVICE-TYPE
device type slug (e.g. "raspberrypi3") to override the application device type
--initial-device-name INITIAL-DEVICE-NAME
This option will set the device name when the device provisions
--version VERSION
balenaOS version, for example "2.32.0" or "2.44.0+rev1"
-c, --system-connection SYSTEM-CONNECTION
paths to local files to place into the 'system-connections' directory
os initialize <image>
Use this command to initialize a device with previously configured operating system image.
Note: Initializing the device may ask for administrative permissions because we need to access the raw devices directly.
Examples:
$ balena os initialize ../path/rpi.img --type 'raspberry-pi'
Options
--yes, -y
confirm non interactively
--type, -t <type>
device type (Check available types with balena devices supported
)
--drive, -d <drive>
the drive to write the image to, like /dev/sdb
or /dev/mmcblk0
. Careful with this as you can erase your hard drive. Check balena util available-drives
for available options.
Config
config read
Use this command to read the config.json file from the mounted filesystem (e.g. SD card) of a provisioned device"
Examples:
$ balena config read --type raspberry-pi
$ balena config read --type raspberry-pi --drive /dev/disk2
Options
--type, -t <type>
device type (Check available types with balena devices supported
)
--drive, -d <drive>
drive
config write <key> <value>
Use this command to write the config.json file to the mounted filesystem (e.g. SD card) of a provisioned device
Examples:
$ balena config write --type raspberry-pi username johndoe
$ balena config write --type raspberry-pi --drive /dev/disk2 username johndoe
$ balena config write --type raspberry-pi files.network/settings "..."
Options
--type, -t <type>
device type (Check available types with balena devices supported
)
--drive, -d <drive>
drive
config inject <file>
Use this command to inject a config.json file to the mounted filesystem (e.g. SD card or mounted balenaOS image) of a provisioned device"
Examples:
$ balena config inject my/config.json --type raspberry-pi
$ balena config inject my/config.json --type raspberry-pi --drive /dev/disk2
Options
--type, -t <type>
device type (Check available types with balena devices supported
)
--drive, -d <drive>
drive
config reconfigure
Use this command to reconfigure a provisioned device
Examples:
$ balena config reconfigure --type raspberry-pi
$ balena config reconfigure --type raspberry-pi --advanced
$ balena config reconfigure --type raspberry-pi --drive /dev/disk2
Options
--type, -t <type>
device type (Check available types with balena devices supported
)
--drive, -d <drive>
drive
--advanced, -v
show advanced commands
config generate
Use this command to generate a config.json for a device or application.
Calling this command with the exact version number of the targeted image is required.
This is interactive by default, but you can do this automatically without interactivity by specifying an option for each question on the command line, if you know the questions that will be asked for the relevant device type.
In case that you want to configure an image for an application with mixed device types, you can pass the --device-type argument along with --app to specify the target device type.
Examples:
$ balena config generate --device 7cf02a6 --version 2.12.7
$ balena config generate --device 7cf02a6 --version 2.12.7 --generate-device-api-key
$ balena config generate --device 7cf02a6 --version 2.12.7 --device-api-key <existingDeviceKey>
$ balena config generate --device 7cf02a6 --version 2.12.7 --output config.json
$ balena config generate --app MyApp --version 2.12.7
$ balena config generate --app MyApp --version 2.12.7 --device-type fincm3
$ balena config generate --app MyApp --version 2.12.7 --output config.json
$ balena config generate --app MyApp --version 2.12.7 --network wifi --wifiSsid mySsid --wifiKey abcdefgh --appUpdatePollInterval 1
Options
--version <version>
a balenaOS version
--application, -a, --app <application>
application name
--device, -d <device>
device uuid
--deviceApiKey, -k <device-api-key>
custom device key - note that this is only supported on balenaOS 2.0.3+
--deviceType <device-type>
device type slug
--generate-device-api-key
generate a fresh device key for the device
--output, -o <output>
output
--network <network>
the network type to use: ethernet or wifi
--wifiSsid <wifiSsid>
the wifi ssid to use (used only if --network is set to wifi)
--wifiKey <wifiKey>
the wifi key to use (used only if --network is set to wifi)
--appUpdatePollInterval <appUpdatePollInterval>
how frequently (in minutes) to poll for application updates
Preload
preload <image>
Preload a balena application release (app images/containers), and optionally a balenaOS splash screen, in a previously downloaded '.img' balenaOS image file in the local disk (a zip file is only accepted for the Intel Edison device type). After preloading, the balenaOS image file can be flashed to a device's SD card. When the device boots, it will not need to download the application, as it was preloaded.
Warning: "balena preload" requires Docker to be correctly installed in your shell environment. For more information (including Windows support) check: https://github.com/balena-io/balena-cli/blob/master/INSTALL.md
Examples:
$ balena preload balena.img --app 1234 --commit e1f2592fc6ee949e68756d4f4a48e49bff8d72a0 --splash-image image.png
$ balena preload balena.img
Options
--app, -a <appId>
id of the application to preload
--commit, -c <hash>
The commit hash for a specific application release to preload, use "current" to specify the current release (ignored if no appId is given). The current release is usually also the latest, but can be manually pinned using https://github.com/balena-io-projects/staged-releases .
--splash-image, -s <splashImage.png>
path to a png image to replace the splash screen
--dont-check-arch
Disables check for matching architecture in image and application
--pin-device-to-release, -p
Pin the preloaded device to the preloaded release on provision
--add-certificate <certificate.crt>
Add the given certificate (in PEM format) to /etc/ssl/certs in the preloading container. The file name must end with '.crt' and must not be already contained in the preloader's /etc/ssl/certs folder. Can be repeated to add multiple certificates.
--docker, -P <docker>
Path to a local docker socket (e.g. /var/run/docker.sock)
--dockerHost, -h <dockerHost>
Docker daemon hostname or IP address (dev machine or balena device)
--dockerPort <dockerPort>
Docker daemon TCP port number (hint: 2375 for balena devices)
--ca <ca>
Docker host TLS certificate authority file
--cert <cert>
Docker host TLS certificate file
--key <key>
Docker host TLS key file
Push
push <applicationOrDevice>
This command can be used to start a build on the remote balena cloud builders, or a local mode balena device.
When building on the balenaCloud servers, the given source directory will be sent to the remote server. This can be used as a drop-in replacement for the "git push" deployment method.
When building on a local mode device, the given source directory will be built on the device, and the resulting containers will be run on the device. Logs will be streamed back from the device as part of the same invocation. The web dashboard can be used to switch a device to local mode: https://www.balena.io/docs/learn/develop/local-mode/ Note that local mode requires a supervisor version of at least v7.21.0. The logs from only a single service can be shown with the --service flag, and showing only the system logs can be achieved with --system. Note that these flags can be used together.
When pushing to a local device a live session will be started. The project source folder is watched for filesystem events, and changes to files and folders are automatically synchronized to the running containers. The synchronization is only in one direction, from this machine to the device, and changes made on the device itself may be overwritten. This feature requires a device running supervisor version v9.7.0 or greater.
REGISTRY SECRETS
The --registry-secrets option specifies a JSON or YAML file containing private
Docker registry usernames and passwords to be used when pulling base images.
Sample registry-secrets YAML file:
'my-registry-server.com:25000':
username: ann
password: hunter2
'': # Use the empty string to refer to the Docker Hub
username: mike
password: cze14
'eu.gcr.io': # Google Container Registry
username: '_json_key'
password: '{escaped contents of the GCR keyfile.json file}'
For a sample project using registry secrets with the Google Container Registry, check: https://github.com/balena-io-playground/sample-gcr-registry-secrets
If the --registry-secrets option is not specified, and a secrets.yml or secrets.json file exists in the balena directory (usually $HOME/.balena), this file will be used instead.
DOCKERIGNORE AND GITIGNORE FILES
By default, the balena CLI will use a single ".dockerignore" file (if any) at
the project root (--source directory) in order to decide which source files to
exclude from the "build context" (tar stream) sent to balenaCloud, Docker daemon
or balenaEngine. In a microservices (multicontainer) application, the source
directory is the directory that contains the "docker-compose.yml" file.
The --multi-dockerignore (-m) option may be used with microservices (multicontainer)
applications that define a docker-compose.yml file. When this option is used,
each service subdirectory (defined by the build
or build.context
service
properties in the docker-compose.yml file) is filtered separately according to
a .dockerignore file defined in the service subdirectory. If no .dockerignore
file exists in a service subdirectory, then only the default .dockerignore
patterns (see below) apply for that service subdirectory.
When the --multi-dockerignore (-m) option is used, the .dockerignore file (if any) defined at the overall project root will be used to filter files and subdirectories other than service subdirectories. It will not have any effect on service subdirectories, whether or not a service subdirectory defines its own .dockerignore file. Multiple .dockerignore files are not merged or added together, and cannot override or extend other files. This behavior maximises compatibility with the standard docker-compose tool, while still allowing a root .dockerignore file (at the overall project root) to filter files and folders that are outside service subdirectories.
Balena CLI releases older than v12.0.0 also took .gitignore files into account. This behavior is deprecated, but may still be enabled with the --gitignore (-g) option if compatibility is required. This option is mutually exclusive with --multi-dockerignore (-m) and will be removed in the CLI's next major version release (v13).
Default .dockerignore patterns
When --gitignore (-g) is NOT used (i.e. when not in v11 compatibility mode), a
few default/hardcoded dockerignore patterns are "merged" (in memory) with the
patterns found in the applicable .dockerignore files, in the following order:
**/.git
< user's patterns from the applicable '.dockerignore' file, if any >
!**/.balena
!**/.resin
!**/Dockerfile
!**/Dockerfile.*
!**/docker-compose.yml
These patterns always apply, whether or not .dockerignore files exist in the
project. If necessary, the effect of the **/.git
pattern may be modified by
adding counter patterns to the applicable .dockerignore file(s), for example
!mysubmodule/.git
. For documentation on pattern format, see:
- https://docs.docker.com/engine/reference/builder/#dockerignore-file
- https://www.npmjs.com/package/@balena/dockerignore
Examples:
$ balena push myApp
$ balena push myApp --source <source directory>
$ balena push myApp -s <source directory>
$ balena push 10.0.0.1
$ balena push 10.0.0.1 --source <source directory>
$ balena push 10.0.0.1 --service my-service
$ balena push 10.0.0.1 --env MY_ENV_VAR=value --env my-service:SERVICE_VAR=value
$ balena push 10.0.0.1 --nolive
$ balena push 23c73a1.local --system
$ balena push 23c73a1.local --system --service my-service
Options
--source, -s <source>
Source directory to be sent to balenaCloud or balenaOS device (default: current working dir)
--emulated, -e
Force an emulated build to occur on the remote builder
--dockerfile <Dockerfile>
Alternative Dockerfile name/path, relative to the source folder
--nocache, -c
Don't use cached layers of previously built images for this project. This ensures that the latest base image and packages are pulled. Note that build logs may still display the message _"Pulling previous images for caching purposes" (as the cloud builder needs previous images to compute delta updates), but the logs will not display the "Using cache" lines for each build step of a Dockerfile.
--noparent-check
Disable project validation check of 'docker-compose.yml' file in parent folder
--registry-secrets, -R <secrets.yml|.json>
Path to a local YAML or JSON file containing Docker registry passwords used to pull base images. Note that if registry-secrets are not provided on the command line, a secrets configuration file from the balena directory will be used (usually $HOME/.balena/secrets.yml|.json)
--nolive
Don't run a live session on this push. The filesystem will not be monitored, and changes will not be synchronized to any running containers. Note that both this flag and --detached and required to cause the process to end once the initial build has completed.
--detached, -d
When pushing to the cloud, this option will cause the build to start, then return execution back to the shell, with the status and release ID (if applicable).
When pushing to a local mode device, this option will cause the command to not tail application logs when the build has completed.
--service <service>
Reject logs not originating from this service. This can be used in combination with --system and other --service flags. Only valid when pushing to a local mode device.
--system
Only show system logs. This can be used in combination with --service. Only valid when pushing to a local mode device.
--env <env>
When performing a push to device, run the built containers with environment variables provided with this argument. Environment variables can be applied to individual services by adding their service name before the argument, separated by a colon, e.g: --env main:MY_ENV=value Note that if the service name cannot be found in the composition, the entire left hand side of the = character will be treated as the variable name.
--convert-eol, -l
No-op and deprecated since balena CLI v12.0.0
--noconvert-eol
Don't convert line endings from CRLF (Windows format) to LF (Unix format).
--multi-dockerignore, -m
Have each service use its own .dockerignore file. See "balena help push".
--nogitignore, -G
No-op (default behavior) since balena CLI v12.0.0. See "balena help push".
--gitignore, -g
Consider .gitignore files in addition to the .dockerignore file. This reverts to the CLI v11 behavior/implementation (deprecated) if compatibility is required until your project can be adapted.
Settings
settings
Use this command to display current balena CLI settings.
Examples:
$ balena settings
Options
Local
local flash <image>
Flash a balenaOS image to a drive. Image file may be one of: .img|.zip|.gz|.bz2|.xz
If --drive is not specified, then it will interactively show a list of available drives for selection.
Examples:
$ balena local flash path/to/balenaos.img
$ balena local flash path/to/balenaos.img --drive /dev/disk2
$ balena local flash path/to/balenaos.img --drive /dev/disk2 --yes
Arguments
IMAGE
path to OS image
Options
-d, --drive DRIVE
drive to flash
-y, --yes
answer "yes" to all questions (non interactive use)
local configure <target>
Use this command to configure or reconfigure a balenaOS drive or image.
Examples:
$ balena local configure /dev/sdc
$ balena local configure path/to/image.img
Deploy
build [source]
Use this command to build an image or a complete multicontainer project with
the provided docker daemon in your development machine or balena device.
(See also the balena push
command for the option of building images in the
balenaCloud build servers.)
You must provide either an application or a device-type/architecture pair to use the balena Dockerfile pre-processor (e.g. Dockerfile.template -> Dockerfile).
This command will look into the given source directory (or the current working
directory if one isn't specified) for a docker-compose.yml file, and if found,
each service defined in the compose file will be built. If a compose file isn't
found, it will look for a Dockerfile[.template] file (or alternative Dockerfile
specified with the --dockerfile
option), and if no dockerfile is found, it
will try to generate one.
REGISTRY SECRETS
The --registry-secrets option specifies a JSON or YAML file containing private
Docker registry usernames and passwords to be used when pulling base images.
Sample registry-secrets YAML file:
'my-registry-server.com:25000':
username: ann
password: hunter2
'': # Use the empty string to refer to the Docker Hub
username: mike
password: cze14
'eu.gcr.io': # Google Container Registry
username: '_json_key'
password: '{escaped contents of the GCR keyfile.json file}'
For a sample project using registry secrets with the Google Container Registry, check: https://github.com/balena-io-playground/sample-gcr-registry-secrets
If the --registry-secrets option is not specified, and a secrets.yml or secrets.json file exists in the balena directory (usually $HOME/.balena), this file will be used instead.
DOCKERIGNORE AND GITIGNORE FILES
By default, the balena CLI will use a single ".dockerignore" file (if any) at
the project root (--source directory) in order to decide which source files to
exclude from the "build context" (tar stream) sent to balenaCloud, Docker daemon
or balenaEngine. In a microservices (multicontainer) application, the source
directory is the directory that contains the "docker-compose.yml" file.
The --multi-dockerignore (-m) option may be used with microservices (multicontainer)
applications that define a docker-compose.yml file. When this option is used,
each service subdirectory (defined by the build
or build.context
service
properties in the docker-compose.yml file) is filtered separately according to
a .dockerignore file defined in the service subdirectory. If no .dockerignore
file exists in a service subdirectory, then only the default .dockerignore
patterns (see below) apply for that service subdirectory.
When the --multi-dockerignore (-m) option is used, the .dockerignore file (if any) defined at the overall project root will be used to filter files and subdirectories other than service subdirectories. It will not have any effect on service subdirectories, whether or not a service subdirectory defines its own .dockerignore file. Multiple .dockerignore files are not merged or added together, and cannot override or extend other files. This behavior maximises compatibility with the standard docker-compose tool, while still allowing a root .dockerignore file (at the overall project root) to filter files and folders that are outside service subdirectories.
Balena CLI releases older than v12.0.0 also took .gitignore files into account. This behavior is deprecated, but may still be enabled with the --gitignore (-g) option if compatibility is required. This option is mutually exclusive with --multi-dockerignore (-m) and will be removed in the CLI's next major version release (v13).
Default .dockerignore patterns
When --gitignore (-g) is NOT used (i.e. when not in v11 compatibility mode), a
few default/hardcoded dockerignore patterns are "merged" (in memory) with the
patterns found in the applicable .dockerignore files, in the following order:
**/.git
< user's patterns from the applicable '.dockerignore' file, if any >
!**/.balena
!**/.resin
!**/Dockerfile
!**/Dockerfile.*
!**/docker-compose.yml
These patterns always apply, whether or not .dockerignore files exist in the
project. If necessary, the effect of the **/.git
pattern may be modified by
adding counter patterns to the applicable .dockerignore file(s), for example
!mysubmodule/.git
. For documentation on pattern format, see:
- https://docs.docker.com/engine/reference/builder/#dockerignore-file
- https://www.npmjs.com/package/@balena/dockerignore
Examples:
$ balena build
$ balena build ./source/
$ balena build --deviceType raspberrypi3 --arch armv7hf --emulated
$ balena build --application MyApp ./source/
$ balena build --docker /var/run/docker.sock # Linux, Mac
$ balena build --docker //./pipe/docker_engine # Windows
$ balena build --dockerHost my.docker.host --dockerPort 2376 --ca ca.pem --key key.pem --cert cert.pem
Options
--arch, -A <arch>
The architecture to build for
--deviceType, -d <deviceType>
The type of device this build is for
--application, -a <application>
The target balena application this build is for
--projectName, -n <projectName>
Specify an alternate project name; default is the directory name
--emulated, -e
Run an emulated build using Qemu
--dockerfile <Dockerfile>
Alternative Dockerfile name/path, relative to the source folder
--logs
No-op and deprecated since balena CLI v12.0.0. Build logs are now shown by default.
--nologs
Hide the image build log output (produce less verbose output)
--gitignore, -g
Consider .gitignore files in addition to the .dockerignore file. This reverts to the CLI v11 behavior/implementation (deprecated) if compatibility is required until your project can be adapted.
--multi-dockerignore, -m
Have each service use its own .dockerignore file. See "balena help build".
--nogitignore, -G
No-op (default behavior) since balena CLI v12.0.0. See "balena help build".
--noparent-check
Disable project validation check of 'docker-compose.yml' file in parent folder
--registry-secrets, -R <secrets.yml|.json>
Path to a YAML or JSON file with passwords for a private Docker registry
--convert-eol, -l
No-op and deprecated since balena CLI v12.0.0
--noconvert-eol
Don't convert line endings from CRLF (Windows format) to LF (Unix format).
--docker, -P <docker>
Path to a local docker socket (e.g. /var/run/docker.sock)
--dockerHost, -h <dockerHost>
Docker daemon hostname or IP address (dev machine or balena device)
--dockerPort, -p <dockerPort>
Docker daemon TCP port number (hint: 2375 for balena devices)
--ca <ca>
Docker host TLS certificate authority file
--cert <cert>
Docker host TLS certificate file
--key <key>
Docker host TLS key file
--tag, -t <tag>
The alias to the generated image
--buildArg, -B <arg>
Set a build-time variable (eg. "-B 'ARG=value'"). Can be specified multiple times.
--cache-from <image-list>
Comma-separated list (no spaces) of image names for build cache resolution. Implements the same feature as the "docker build --cache-from" option.
--nocache
Don't use docker layer caching when building
--squash
Squash newly built layers into a single new layer
deploy <appName> [image]
Usage: deploy <appName> ([image] | --build [--source build-dir])
Use this command to deploy an image or a complete multicontainer project to an
application, optionally building it first. The source images are searched for
(and optionally built) using the docker daemon in your development machine or
balena device. (See also the balena push
command for the option of building
the image in the balenaCloud build servers.)
Unless an image is specified, this command will look into the current directory
(or the one specified by --source) for a docker-compose.yml file. If one is
found, this command will deploy each service defined in the compose file,
building it first if an image for it doesn't exist. If a compose file isn't
found, the command will look for a Dockerfile[.template] file (or alternative
Dockerfile specified with the -f
option), and if yet that isn't found, it
will try to generate one.
To deploy to an app on which you're a collaborator, use
balena deploy <appOwnerUsername>/<appName>
.
When --build is used, all options supported by balena build
are also supported
by this command.
REGISTRY SECRETS
The --registry-secrets option specifies a JSON or YAML file containing private
Docker registry usernames and passwords to be used when pulling base images.
Sample registry-secrets YAML file:
'my-registry-server.com:25000':
username: ann
password: hunter2
'': # Use the empty string to refer to the Docker Hub
username: mike
password: cze14
'eu.gcr.io': # Google Container Registry
username: '_json_key'
password: '{escaped contents of the GCR keyfile.json file}'
For a sample project using registry secrets with the Google Container Registry, check: https://github.com/balena-io-playground/sample-gcr-registry-secrets
If the --registry-secrets option is not specified, and a secrets.yml or secrets.json file exists in the balena directory (usually $HOME/.balena), this file will be used instead.
DOCKERIGNORE AND GITIGNORE FILES
By default, the balena CLI will use a single ".dockerignore" file (if any) at
the project root (--source directory) in order to decide which source files to
exclude from the "build context" (tar stream) sent to balenaCloud, Docker daemon
or balenaEngine. In a microservices (multicontainer) application, the source
directory is the directory that contains the "docker-compose.yml" file.
The --multi-dockerignore (-m) option may be used with microservices (multicontainer)
applications that define a docker-compose.yml file. When this option is used,
each service subdirectory (defined by the build
or build.context
service
properties in the docker-compose.yml file) is filtered separately according to
a .dockerignore file defined in the service subdirectory. If no .dockerignore
file exists in a service subdirectory, then only the default .dockerignore
patterns (see below) apply for that service subdirectory.
When the --multi-dockerignore (-m) option is used, the .dockerignore file (if any) defined at the overall project root will be used to filter files and subdirectories other than service subdirectories. It will not have any effect on service subdirectories, whether or not a service subdirectory defines its own .dockerignore file. Multiple .dockerignore files are not merged or added together, and cannot override or extend other files. This behavior maximises compatibility with the standard docker-compose tool, while still allowing a root .dockerignore file (at the overall project root) to filter files and folders that are outside service subdirectories.
Balena CLI releases older than v12.0.0 also took .gitignore files into account. This behavior is deprecated, but may still be enabled with the --gitignore (-g) option if compatibility is required. This option is mutually exclusive with --multi-dockerignore (-m) and will be removed in the CLI's next major version release (v13).
Default .dockerignore patterns
When --gitignore (-g) is NOT used (i.e. when not in v11 compatibility mode), a
few default/hardcoded dockerignore patterns are "merged" (in memory) with the
patterns found in the applicable .dockerignore files, in the following order:
**/.git
< user's patterns from the applicable '.dockerignore' file, if any >
!**/.balena
!**/.resin
!**/Dockerfile
!**/Dockerfile.*
!**/docker-compose.yml
These patterns always apply, whether or not .dockerignore files exist in the
project. If necessary, the effect of the **/.git
pattern may be modified by
adding counter patterns to the applicable .dockerignore file(s), for example
!mysubmodule/.git
. For documentation on pattern format, see:
- https://docs.docker.com/engine/reference/builder/#dockerignore-file
- https://www.npmjs.com/package/@balena/dockerignore
Examples:
$ balena deploy myApp
$ balena deploy myApp --build --source myBuildDir/
$ balena deploy myApp myApp/myImage
Options
--source, -s <source>
Specify an alternate source directory; default is the working directory
--build, -b
Force a rebuild before deploy
--nologupload
Don't upload build logs to the dashboard with image (if building)
--projectName, -n <projectName>
Specify an alternate project name; default is the directory name
--emulated, -e
Run an emulated build using Qemu
--dockerfile <Dockerfile>
Alternative Dockerfile name/path, relative to the source folder
--logs
No-op and deprecated since balena CLI v12.0.0. Build logs are now shown by default.
--nologs
Hide the image build log output (produce less verbose output)
--gitignore, -g
Consider .gitignore files in addition to the .dockerignore file. This reverts to the CLI v11 behavior/implementation (deprecated) if compatibility is required until your project can be adapted.
--multi-dockerignore, -m
Have each service use its own .dockerignore file. See "balena help build".
--nogitignore, -G
No-op (default behavior) since balena CLI v12.0.0. See "balena help build".
--noparent-check
Disable project validation check of 'docker-compose.yml' file in parent folder
--registry-secrets, -R <secrets.yml|.json>
Path to a YAML or JSON file with passwords for a private Docker registry
--convert-eol, -l
No-op and deprecated since balena CLI v12.0.0
--noconvert-eol
Don't convert line endings from CRLF (Windows format) to LF (Unix format).
--docker, -P <docker>
Path to a local docker socket (e.g. /var/run/docker.sock)
--dockerHost, -h <dockerHost>
Docker daemon hostname or IP address (dev machine or balena device)
--dockerPort, -p <dockerPort>
Docker daemon TCP port number (hint: 2375 for balena devices)
--ca <ca>
Docker host TLS certificate authority file
--cert <cert>
Docker host TLS certificate file
--key <key>
Docker host TLS key file
--tag, -t <tag>
The alias to the generated image
--buildArg, -B <arg>
Set a build-time variable (eg. "-B 'ARG=value'"). Can be specified multiple times.
--cache-from <image-list>
Comma-separated list (no spaces) of image names for build cache resolution. Implements the same feature as the "docker build --cache-from" option.
--nocache
Don't use docker layer caching when building
--squash
Squash newly built layers into a single new layer
Platform
join [deviceIpOrHostname]
Move a local device to an application on another balena server, causing the device to "join" the new server. The device must be running balenaOS.
For example, you could provision a device against an openBalena installation where you perform end-to-end tests and then move it to balenaCloud when it's ready for production.
To move a device between applications on the same server, use the
balena device move
command instead of balena join
.
If you don't specify a device hostname or IP, this command will automatically scan the local network for balenaOS devices and prompt you to select one from an interactive picker. This requires root privileges. Likewise, if the application flag is not provided then a picker will be shown.
Examples:
$ balena join
$ balena join balena.local
$ balena join balena.local --application MyApp
$ balena join 192.168.1.25
$ balena join 192.168.1.25 --application MyApp
Arguments
DEVICEIPORHOSTNAME
the IP or hostname of device
Options
-a, --application APPLICATION
application name
leave [deviceIpOrHostname]
Remove a local device from its balena application, causing the device to "leave" the server it is provisioned on. This effectively makes the device "unmanaged". The device must be running balenaOS.
The device entry on the server is preserved after running this command, so the device can subsequently re-join the server if needed.
If you don't specify a device hostname or IP, this command will automatically scan the local network for balenaOS devices and prompt you to select one from an interactive picker. This usually requires root privileges.
Examples:
$ balena leave
$ balena leave balena.local
$ balena leave 192.168.1.25
Arguments
DEVICEIPORHOSTNAME
the device IP or hostname
Options
Utilities
util available-drives
List available drives which are usable for writing an OS image to. Does not list system drives.