Dockerfile | ||
LICENSE | ||
psu | ||
README.md |
Portainer Stack Utils
Bash script to deploy/update/undeploy stacks in a Portainer instance from a docker-compose yaml file. Based on previous work by @vladbabii on docker-how-to/portainer-bash-scripts.
Supported Portainer API
Script was created for the latest Portainer API, which at the time of writing is 1.9.2.
Requirements
How to use
The provided psu
script allows to deploy/update/undeploy Portainer stacks. Settings can be passed through envvars and/or flags. Both envvars and flags can be mixed but flags will always overwrite envvar values. When deploying a stack, if it doesn't exist a new one is created, otherwise it's updated (unless strict mode is active).
With envvars
This is particularly useful for CI/CD pipelines.
ACTION
("deploy" or "undeploy", required): Whether to deploy or undeploy the stackPORTAINER_USER
(string, required): UsernamePORTAINER_PASSWORD
(string, required): PasswordPORTAINER_URL
(string, required): URL to PortainerPORTAINER_STACK_NAME
(string, required): Stack nameDOCKER_COMPOSE_FILE
(string, required if action=deploy): Path to doker-compose filePORTAINER_PRUNE
("true" or "false", optional): Whether to prune unused containers or not. Defaults to"false"
.PORTAINER_ENDPOINT
(int, optional): Which endpoint to use. Defaults to1
.HTTPIE_VERIFY_SSL
("yes" or "no", optional): Whether to verify SSL certificate or not. Defaults to"yes"
.VERBOSE_MODE
("true" or "false", optional): Whether to activate verbose output mode or not. Defaults to"false"
. See verbose mode below.DEBUG_MODE
("true" or "false", optional): Whether to activate debug output mode or not. Defaults to"false"
. See debug mode below.STRICT_MODE
("true" or "false", optional): Whether to activate strict mode or not. Defaults to"false"
. See strict mode below.
Examples
export ACTION="deploy"
export PORTAINER_USER="admin"
export PORTAINER_PASSWORD="password"
export PORTAINER_URL="http://portainer.local"
export PORTAINER_STACK_NAME="mystack"
export DOCKER_COMPOSE_FILE="/path/to/docker-compose.yml"
./psu
export ACTION="undeploy"
export PORTAINER_USER="admin"
export PORTAINER_PASSWORD="password"
export PORTAINER_URL="http://portainer.local"
export PORTAINER_STACK_NAME="mystack"
./psu
With flags
This is more suitable for standalone script usage.
-a
("deploy" or "undeploy", required): Whether to deploy or undeploy the stack-u
(string, required): Username-p
(string, required): Password-l
(string, required): URL to Portainer-n
(string, required): Stack name-c
(string, required if action=deploy): Path to doker-compose file-r
("true" or "false", optional): Whether to prune unused containers or not. Defaults to"false"
.-e
(int, optional): Which endpoint to use. Defaults to1
.-s
("yes" or "no", optional): Whether to verify SSL certificate or not. Defaults to"yes"
.-v
("true" or "false", optional): Whether to activate verbose output mode or not. Defaults to"false"
. See verbose mode below.-d
("true" or "false", optional): Whether to activate debug output mode or not. Defaults to"false"
. See debug mode below.-t
("true" or "false", optional): Whether to activate strict mode or not. Defaults to"false"
. See strict mode below.
Examples
./psu -a deploy -u admin -p password -l http://portainer.local -n mystack -c /path/to/docker-compose.yml
./psu -a undeploy -u admin -p password -l http://portainer.local -n mystack
Verbose mode
In verbose mode the script prints execution steps.
Getting auth token...
Getting stack mystack...
Stack mystack not found.
Getting Docker info...
Getting swarm cluster (if any)...
Swarm cluster found.
Preparing stack JSON...
Creating stack mystack...
Verbose mode can be enabled through VERBOSE_MODE envvar or -v flag.
Debug mode
In debug mode the script prints as much information as possible to help diagnosing a malfunction.
WARNING: Debug mode will print configuration values (with Portainer credentials) and Portainer API responses (with sensitive information like authentication token and stacks environment variables). Avoid using debug mode in CI/CD pipelines, as pipeline logs are usually recorded.
Debug mode can be enabled through DEBUG_MODE envvar or -d flag.
Strict mode
In strict mode the script never updates an existent stack, and instead exits with an error. Following the same principle, it never removes an unexistent stack.
Strict mode can be enabled through STRICT_MODE envvar or -t flag.
License
Source code contained by this project is licensed under the GNU General Public License version 3. See LICENSE file for reference.