Environment Variables

Here is a list of predefined environment variables visible during runtime (including hook processing, cron jobs and ssh commands):

Note

For complete list of available environment variables, deploy sample app: node-env.

DATABASE_URL

Contains an access URL for a database service. If more than one type of database is present, DATABASE_URL will not be available. Instead, use the Database Specific URL variables below.

Example:

mysql://u93Mm8XmGXQ9R:[email protected]:3306/d0a60c0be931f4982bbef153f993237bc

MEMCACHE_URL

Location of and credentials for the bound Memcached service, if there is (only) one.

MYSQL_URL

Location of and credentials for the bound MySQL service, if there is (only) one.

POSTGRESQL_URL

Location of and credentials for the bound PostgreSQL service, if there is (only) one.

REDIS_URL

Location of and credentials for the bound Redis service, if there is (only) one.

MONGODB_URL

Location of and credentials for the bound MongoDB service, if there is (only) one.

RABBITMQ_URL

Location of and credentials for the bound RabbitMQ service, if there is (only) one.

HOME

Identifies the working directory assigned to a particular user on login. In a Stackato application container, this is generally set to /app/app/ by default.

HTTP_PROXY

A variable recognized by many web applications to direct them to a proxy HTTP server.

PATH

A list of directories, separated by ":", which are to be searched for the names of executable files to be interpreted as commands.

PIP_OPTS

Custom/alternate PIP repo location. See running your own package index for more info.

Example:

env:
  PIP_OPTS: "--extra-index-url=http://company.com/inhouse-pypi-mirror"

PORT

Stackato alternative for VCAP_APP_PORT.

PROCESSES_WEB

This variable contains the default start command that would be used when stackato.yml doesn't override it. It is provided so that users can specify a wrapper around the default command, e.g.

processes:

web: newrelic_wrapper $PROCESSES_WEB

Note that PROCESSES_WEB may be undefined when Stackato can't determine the default command (e.g. because the app uses a non-standard main application file).

PYPM_OPTS

Custom/alternate PyPM repo location. Repo mirroring is sort of undocumented feature. Example:

env:
    PYPM_OPTS: "-R http://pypm-free.activestate.com/2.7/linux-x86_64/"

STACKATO_APP_ENV

Note

Internal use, subject to change.

Contains a list of all environment variables set with stackato env-add or stackato.yml.

STACKATO_APP_NAME

Contains the application name as specified during application push (or in stackato.yml or manifest.yml). Available during staging as well as in the application instance.

STACKATO_APP_NAME_UPCASE

Contains the same value as STACKATO_APP_NAME transformed to uppercase, with dashes replaced by underscores. For example if STACKATO_APP_NAME is "php-info", then STACKATO_APP_NAME_UPCASE will be "PHP_INFO".

This makes it possible to access the environment variables for harbor and filesystem services.

STACKATO_APP_ROOT

This is the "root" directory from the Stackato point of view. It contains app specific HOME directory (app/), the log file directory (logs/) and various scripts.

The HOME environment variable will actually point to the app directory, which looks mostly like the directory uploaded by the client. This is where stackato.yml and all the application files live.

STACKATO_DOCUMENT_ROOT

This contains the root directory where the user can access. The document-root must always be specified relative to $HOME (/app/app).

STACKATO_FILESYSTEM

If the app uses a single filesystem service, then the local mount point is stored in this variable.

If there is more than one filesystem service, STACKATO_FILESYSTEM is not available. Instead, a custom environment variable STACKATO_FILESYSTEM_* will be created based on the name of each filesystem service (with hyphens replaced by underscores).

For example, if your stackato.yml file configures the following services:

services:
  my-data: filesystem
  plugins: filesystem

Two environment variables would be created: STACKATO_FILESYSTEM_MY_DATA and STACKATO_FILESYSTEM_PLUGINS.

STACKATO_GROUP

Can be set in the local shell to specify the group for the stackato client. When set, the stackato group command is ignored until the variable is explicitly unset. If empty, the client sets the group to 'none'. Can be overridden with the --group option.

STACKATO_SERVICES

Contains connection details and credentials for services bound to the application. For filesystem services, it contains the local mount point. See STACKATO_SERVICES.

STACKATO_TARGET

Can be set in the local shell to specify the API endpoint target for the stackato client. When set, the stackato target command is ignored until the variable is explicitly unset. Can be overridden with the --target option.

STACKATO_UWSGI

Set by the Perl and Python frameworks only. It contains the default uswgi start command to run the application via uwsgi. It is provided in case the user wants to add additional uwsgi options in stackato.yml:

processes:
  web: $STACKATO_UWSGI --another-uwsgi-option

VCAP_APP_HOST

This variable contains the IP address of the host that the application is running on.

VCAP_APP_PORT

This variable contains the port that the application will be exposed on.

VCAP_APPLICATION

This variable contails all relevant application details for the Stackato Application. (Instance ID, App Name, App Uris, Users/Groups etc.)

VCAP_SERVICES

Contains connection details, credentials, and meta data for services bound to the application. See VCAP_SERVICES.