Skip to content

Latest commit

 

History

History
153 lines (112 loc) · 6.17 KB

README.md

File metadata and controls

153 lines (112 loc) · 6.17 KB
OpenConext

Build status License

OpenConext User Lifecycle

Deprovision users within the OpenConext platform. The User Lifecycle application is where the last login information of OpenConext suite users is stored. From this application you can trigger the deprovisioning of users that are no longer considered active users.

Configuring deprovision clients

A deprovision client is an OpenConext suite app that implements the deprovisioning API. And can therefor be used by OpenConext User Lifecycle to deprovision users from the platform. To configure a client, please update the config/legacy/parameters.yml file. For each client provide an entry in the open_conext_user_lifecycle_clients configuration section. An example can be found below.

open_conext_user_lifecycle_clients:
    openconext_engineblock:
        url: 'https://engine.example.com/path/to/api/'
        username: 'my-user-name'
        password: 'secret'
        verify_ssl: false
    teams:
        url: 'https://teams.example.com/api'
        username: 'deprovision'
        password: 'secret'

For more information about setting up the clients, see the /config/legacy/parameters.yml.dist file.

Deprovisioning users

Deprovisioning users can be done on a user basis, providing the user collab person id. Or automatically after a period of inactiviy. This period can be configured in the /config/legacy/parameters.yml. Both options use the userlifecycle deprovision console command.

Single user

The userlifecycle deprovision takes an user argument and several other options.

The user argument should be the one and only argument of the command.

Options

Name Shortcut Description
--dry-run none Enables dry run mode, simulates a deprovision action, returning the output a regular run would, but without actually deprovisioning the user.
--json none Only outputs JSON. Must be used in combination with the --no-interaction option.
--pretty none Pretty-print JSON output.
--no-interaction -n Prevents the confirmation question.

Example usage

$ userlifecycle deprovision urn:collab:person:surf.nl:janis_joplin
Continue with deprovisioning of "urn:collab:person:surf.nl:janis_joplin"? (y/n)
# Will start deprovisioning after a positive answer to the confirmation.
$ userlifecycle deprovision urn:collab:org:surf.nl:janis_joplin --dry-run
# Asks confirmation, will not deprovision actual user data
$ userlifecycle deprovision urn:collab:org:surf.nl:janis_joplin --no-interaction --json
# Starts deprovisioning right away, will only output the JSON returned from the services.

Batch deprovisioning

When the user argument is omitted, the deprovision command will start deprovisioning the users that have exceeded the inactivity period set in the inactivity_period parameter in parameters.yml. This parameter must be an integer value representing the months of inactivity before a user must be deprovisioned.

By default 37 months used as the inactivity period.

Options

The same options can be used as described in the Single user section above.

Example usage

$ userlifecycle deprovision
Continue with deprovisioning? (y/n)
# Will start deprovisioning after a positive answer to the confirmation.
$ userlifecycle deprovision --dry-run --no-interaction
Continue with deprovisioning? (y/n)
# Will start a dry run without asking for confirmation.

Gather information about a user

To read user information you can use the information console command.

The information command takes one argument which is the collabPersonId.

Options

Name Shortcut Description
--json none Only outputs JSON.

Example usage

$ userlifecycle information urn:collab:example.org:user_id

API

An API can be toggled, exposing the deprovision command (in read mode). Use the following feature toggle to enable/disable the API.

In config/legacy/parameters.yml

# By default the API is disabled
deprovision_api_settings_enabled: true

Only user information can be read from the endpoint. The API by default is configured with basic authentication, using a configurable username and password.

In config/legacy/parameters.yml

# To enable the API
deprovision_api_settings_enabled: true
deprovision_api_settings_username: userlifecycle
deprovision_api_settings_password: secret

Please note that the username and password should always be provided even when the API is disabled.

The API can be called in the following manner for a given user's collabPersonId:

GET /api/deprovision/urn:collab:person:example.org:jdoe

and will return the deprovision information in JSON format.

There are some rules on how the user data should be structured. User Lifecycle will only accept properly formatted user data. The contract can be found in the docs/deprovision-information.md.

Logging

Production logging

Logging is configured slightly different for the UserLifecycle project. On other OpenConext apps logging on production is done in syslog using the fingers crossed strategy. Fingers crossed means that no detailed log trails are produced in syslog unless a certain log level is reached. Say the application logs an error. Fingers crossed will then also log any previous log messages along the error. Giving the log-auditor all the context it needs.

A great log solution, but this did not fit for UserLifecyle. Here we log data we always want to see in syslog. And using the fingers crossed strategy here was not practical. So the regular stream log strategy is used, logging everything surpassing the configured log level (notice).

For developers

See the /docs folder for more details information about the application.