Docker deployment utilities for REDMIC infrastructure.
You can use it to deploy your own services, supporting **Docker Compose** (both v1 and v2) and **Docker Swarm** environments.
## Actions
***deploy**: Perform a service deployment at a Docker environment. Contains 3 stages:
**prepare-deploy*: Check dependencies, copy resources to deployment target host (*compose* files, service configurations...), prepare environment variables and directories, etc.
**do-deploy*: Launch service at deployment target host. Both standard (using `docker compose`) and *Swarm* (using `docker stack deploy`) modes are supported (deprecated versions too), but *Swarm* mode is recommended (even for single-node clusters).
**check-deploy*: Once deployment is done, this stage waits a defined time period for the service to being up and running (or stopped after run successfully). If service status remains stable after several checks, then it is considered successfully deployed.
***create-nets**: Prepare deployment target host environment creating Docker networks which are external to service definition (not created by service deployment itself, defined as *external* in compose files).
***relaunch**: Force a previously deployed service to update, relaunching it with the same service configuration. Available only for *Swarm* mode.
***deploy**:
Perform a service deployment at a Docker environment. Contains several stages:
1.**definitions**:
Set initial configuration values, getting environment values and with local defaults as fallback. Also prints the initial banner.
1.**ssh-config**:
Prepare connection to deployment target host environment. Set connection options, add identity and define functions to run commands at target.
1.**check-env**:
Check dependencies, version requirements and available modes at deployment target host environment.
1.**check-mode**:
Check which deploy mode (`Swarm` or `Compose`) will be used at deployment target host environment.
1.**prepare-env**:
Prepare environment variables, define directories to use and get names of services to deploy.
1.**check-config**:
Validate deployment configuration at *compose* files, using environment variables and deployment mode set.
1.**prepare-deploy**:
Prepare directories and copy resources to deployment target host (*compose* files, service configurations...).
1.**do-deploy**:
Launch service at deployment target host. Both standard (using `docker compose`) and *Swarm* (using `docker stack deploy`) modes are supported (deprecated versions too), but *Swarm* mode is recommended (even for single-node clusters).
1.**check-deploy**:
Once deployment is done, this stage waits a defined time period for the service to being up and running (or stopped after run successfully). If service status remains stable after several checks, then it is considered successfully deployed. This check can be ommited with `OMIT_STATUS_CHECK=1`.
***create-nets**:
Prepare deployment target host environment creating Docker networks which are external to service definition. A network is external when it's not created by service deployment itself, because is defined as *external* in compose files. Contains several stages:
1.**definitions**:
Set initial configuration values, getting environment values and with local defaults as fallback. Also prints the initial banner.
1.**ssh-config**:
Prepare connection to deployment target host environment. Set connection options, add identity and define functions to run commands at target.
1.**check-env**:
Check dependencies, version requirements and available modes at deployment target host environment.
1.**do-create-nets**:
Prepare networks creation command for supported mode and run it at deployment target host environment.
***relaunch**:
Force a previously deployed service to update, relaunching it with the same service configuration. Available only for *Swarm* mode. Contains several stages:
1.**definitions**:
Set initial configuration values, getting environment values and with local defaults as fallback. Also prints the initial banner.
1.**ssh-config**:
Prepare connection to deployment target host environment. Set connection options, add identity and define functions to run commands at target.
1.**check-env**:
Check dependencies, version requirements and available modes at deployment target host environment.
1.**prepare-relaunch**:
Obtain single service name to relaunch, because `SERVICE` value might be a prefix for several service names. Check if service exists at deployment target host environment.
1.**prepare-registry**:
Only when using `USE_IMAGE_DIGEST=1`. Perform login to registry when using credentials, to be able to get updated images.
1.**prepare-digest**:
Only when using `USE_IMAGE_DIGEST=1`. Before relaunching service, get updated image and its current digest data.
1.**do-relaunch**:
Run service relaunch at deployment target host environment.
## Usage
@@ -31,13 +117,19 @@ As you can see, configuration is possible through environment variables and by s
Using environment variables, you can configure:
* Behaviour of this image itself.
* Deployment target host environment (where you are deploying to) for service deployment configuration and deployed service environment variables (the latter only when action is *deploy* and using the `ENV_PREFIX` prefix in your variable names).
* Behaviour of `docker-deploy` itself.
* Deployment target host (where you are deploying to) environment, for:
* service deployment configuration (values used to configure the deployment), but not exposed into the deployed service.
* deployed service environment variables, only when action is *deploy* and you are using the `ENV_PREFIX` (`DD_` by default) prefix in your variable names.
Using script parameters you can set:
* When action is *deploy*, deployment target host environment (where you are deploying to) for service deployment configuration and deployed service environment variables. These parameters overwrite previous environment values, including those defined using the `ENV_PREFIX` prefix.
* When action is *create-nets*, the name of external networks to create.
* When action is *deploy*:
* deployment target host (where you are deploying to) environment, for:
* service deployment configuration (values used to configure the deployment), but not exposed into the deployed service.
* deployed service environment variables, using names without prefix. These parameters will overwrite previous environment values, including those defined using the `ENV_PREFIX` (`DD_` by default) prefix.
* When action is *create-nets*:
* the name of external networks to create.
## Configuration
@@ -50,9 +142,10 @@ You may define these environment variables (**bold** are mandatory):
| **DEPLOY_KEY** | - | Private key used to authenticate, paired with a public key accepted by remote host. |
| **SSH_REMOTE** | - | SSH user and hostname (DNS or IP) of remote host where you are going to deploy. |
| **STACK** | - | Name of Docker stack (*Swarm* mode) or project (*Compose* mode) used to wrap deployed services. |
| *ALLOW_COMPOSE_ENV_FILE_INTERPOLATION* | `0` | Allow passing variable values directly from `COMPOSE_ENV_FILE_NAME` file (`.env` by default), to let *Compose* interpolate variables used into values. By default, values will be single-quoted before checking config and deploying with *Compose*, to avoid getting unwanted variable resolution. Useful only for *Compose* mode. |
| *COMPOSE_ENV_FILE_NAME* | `.env` | Name of variable values definition file. |
| *COMPOSE_FILE* | `compose.yaml` | Name of service definition file. Multiple files are supported, separated by colon (`:`). |
| *DEFAULT_DEPLOY_FILES* | `*compose*.y*ml .env` | Files needed for deployment. Used only if `DEPLOY_DIR_NAME` directory does not exist. |
| *DEFAULT_DEPLOY_FILES* | `*compose*.y*ml ${COMPOSE_ENV_FILE_NAME}` | Files needed for deployment. Used only if `DEPLOY_DIR_NAME` directory does not exist. |
| *DEPLOY_DIR_NAME* | `deploy` | Name of directory containing files needed for deployment. If directory exists, `DEFAULT_DEPLOY_FILES` is ignored and all contents are copied to deployment target host. |
| *DEPLOY_PATH* | `~` | Path in deployment target host where deployment directory (used to hold temporary files) will be created. |
| *ENV_PREFIX* | `DD_` | Prefix used to identify variables to be defined in deployment target host environment and service, available there without this prefix. Change this if default value collides with the beginning of your variable names. |
@@ -61,6 +154,7 @@ You may define these environment variables (**bold** are mandatory):
| *GREP_BIN* | `grep` | Path to *grep* binary in deployment target host. |
| *OMIT_CLEAN_DEPLOY* | `0` | Leave at deployment target host all deployment resources after doing a deploy. Useful when using bind mounts or *Compose* secrets (pointing to static content in deployment resources) or you want to check sent contents. |
| *OMIT_STATUS_CHECK* | `0` | Bypass status check process after deploying services. Useful when you need to be fast. |
| *OMIT_WAITING_TO_CONVERGE* | `0` | Bypass waiting for relaunched service to converge (when all service tasks are running and healthy, if supported). Useful when you need to be fast. Available only for *relaunch* action. |
| *REGISTRY_PASS* | - | Docker registry password, corresponding to a user with read permissions. **Required** for private registry or repository. |
| *REGISTRY_URL* | - | Docker registry address, where Docker must log in to retrieve images. Useful only when using private registry or repository. Default is empty, to use Docker Hub registry. |
| *REGISTRY_USER* | - | Docker registry username, corresponding to a user with read permissions. **Required** for private registry or repository. |
@@ -74,16 +168,19 @@ You may define these environment variables (**bold** are mandatory):
| *STATUS_CHECK_INTERVAL* | `20` | Seconds to wait between check iterations. |
| *STATUS_CHECK_MIN_HITS* | `3` | Minimum number of successful checks to consider deployment as successful. |
| *STATUS_CHECK_RETRIES* | `10` | Maximum number of checks before considering deployment as failed. |
| *SWARM_RESOLVE_IMAGE* | `always` | Allows to edit the behaviour of the registry query to resolve image digests and supported platforms (`always`, `changed` or `never`). |
| *USE_IMAGE_DIGEST* | `0` | Update service image using digest data when relaunching. Available only for *relaunch* action. |
| *SWARM_RESOLVE_IMAGE* | `always` | Allows to edit the behaviour of *deploy* and *relaunch* actions when querying registry, to resolve image digests and supported platforms or not. Supported values are `always`, `changed` or `never`. |
| *USE_IMAGE_DIGEST* | `0` | Update service image using digest data when relaunching. Useful when using images which receive updates under same tag and want to keep them updated with same version on all nodes. Available only for *relaunch* action. |
### Your services
When using *deploy* action, you can configure your own services through variables:
When using *deploy* action, you can configure your own services through variables.
> Note that you must declare them at your compose files too (into `environment` section, for example).
* Define any variable whose name is prefixed by `ENV_PREFIX` prefix:
1. Set variable `docker run ... -e DD_ANY_NAME=value ... deploy`.
2.`ANY_NAME` will be available into service containers with `value` value.
* Pass any variable as deploy script parameter (without `ENV_PREFIX` prefix):
1. Set parameter to deploy script: `docker run ... deploy ANY_NAME=value`.
2.`ANY_NAME` will be available into service containers with `value` value.