Install Statistics Engine

To install the statistics engine please follow the steps described.

Download the Docker image

The Onegini Statistics Engine is delivered as a Docker image. There are two ways to receive a Docker image:

Docker Index

The first and preferred way is to use the Onegini index. This is the central container repository of Onegini. To download the containers you first need to login with the Docker client using your login credentials. 

docker login https://release.onegini.com

After you logged in on the Onegini Index you can pull the latest images.

docker pull release.onegini.com/onegini/statistics-engine:<version>

NOTE: If you did not not receive any login credentials, please contact Onegini support.

Import from tar

The second way is to import the image from a tarball. This options is mostly used when the docker host has no internet connection.

docker load < /path/to/statistics-engine-<version>.tar

Show images

To verify that the docker images are loaded, use the docker images command.

docker images

ETCD configuration

The Onegini Statistics Engine uses ETCD to retrieve its configuration.

Specify ETCD cluster location

Therefore the location of the ETCD cluster should be specified at login. This can be done by setting a set of environment variables via the docker-compose file.

Parameter Description Default value
ETCD_PREFIX Specify the path within ETCD the configuration can be found
ETCD_HOST The ip or hostname of the ETCD installation localhost
ETCD_PORT The port to ETCD installation is listening to 4001
ETCD_PROTOCOL The protocol used to connect to the ETCD installation http

Mandatory settings set in ETCD

The Onegini Statistics Engine requires some settings to be present in ETCD to be able to start.

Parameter Description
/statistics-engine/common/property-encryption-password Password to decrypt encrypted property values see property encryption

Before starting the Onegini Statistics Engine you might want to review the default database connection settings and adapt them if required, see database setup.

Deploy application

The application can be deployed using docker-compose. At least 512m of memory is required by the application, this should be assigned using the JAVA_OPTS environment variable. To enable the use of the spark console, port 4040 should be exposed.

Example docker-compose file

statisticsengine:
  image: snapshot.onegini.com/onegini/statistics-engine:snapshot
  environment:
    - JAVA_OPTS=-Xmx512m -Xms512m
  restart: always
  user: onegini
  ports:
    - 4040:4040