Data Analytics Integration

This application syncs OpenShift user activity data with external analytics systems.

Usage

Using Woopra with default IDs

oc new-app -n openshift-infra -f ansible/roles/oso_analytics/files/user-analytics.yaml \
-p WOOPRA_ENABLED="true" \
-p WOOPRA_ENDPOINT="http://www.woopra.com/track/ce" \
-p WOOPRA_DOMAIN="YOURDOMAIN" \
-p USER_KEY_STRATEGY="[name/uid/annotation]" \
-p USER_KEY_ANNOTATION="[desired annotation]" \
-p LOG_LEVEL=5
-p CLUSTER_NAME="<int/stg/prod/..>"

The CLUSTER_NAME parameter is used to distinguish different environments (such as INT, STG, or a local test cluster). Its value by default is "kubernetes".

Note the 2 new flags USER_KEY_STRATEGY and USER_KEY_ANNOTATION. Key strategy refers to the keying method used for users in Woopra:

• name will key users by their user.Name
• uid will key users by their user.UID
• annotation will key users by an annotation specified in USER_KEY_ANNOTATION (this flag is only required if using strategy=annotation)

The annotation previously used in Devpreview was openshift.io/online-managed-id

Prometheus metrics enablement

The following flags enable Prometheus metrics gathering:

...
-p METRICS_PORT="8080" \
-p METRICS_COLLECT_RUNTIME="true" \
-p METRICS_COLLECT_WOOPRA="true" \
-p METRICS_COLLECT_QUEUE="true" \

The metrics are available at METRICS_PORT/metrics

METRICS_COLLECT_RUNTIME enables various runtime, Go, and process metrics provided by Prometheus

METRICS_COLLECT_WOOPRA enables Woopra latency metrics at /analytics_woopra_latency_seconds

METRICS_COLLECT_QUEUE enables analytics on the controller's internal events processing queue available at /analytics_queue_size_events, which returns the current size of the queue, and /analytics_events_handled, which returns the total number of events processed since the start of the controller.

Local development and testing

Using a Local Endpoint

oc new-app -n openshift-infra -f templates/user-analytics.yaml \
-p WOOPRA_ENABLED="false" \
-p LOCAL_ENDPOINT_ENABLED="true" \
-p LOG_LEVEL=5

Note! This will log to glog.V(5). A user must be logged in for an analytic to be counted.

Generate analytics by logging in and creating a project w/ basic app.

$ oc login
Authentication required for https://10.240.0.2:8443 (openshift)
Username: foo
Password:
Login successful.

You don't have any projects. You can try to create a new project, by running

    oc new-project <projectname>

$ oc new-project test
Now using project "test" on server "https://10.240.0.2:8443".

You can add applications to this project with the 'new-app' command. For example, try:

    oc new-app centos/ruby-22-centos7~https://github.com/openshift/ruby-ex.git

to build a new example application in Ruby.
$ oc new-app centos/ruby-22-centos7~https://github.com/openshift/ruby-ex.git

Building

Build and test with make:

• make will run vendor dependencies and then install
• make build will compile the binary for the application
• make test will compile and run the unit tests
• make test-integration will compile and run the integration tests against an OpenShift master

Analytic Events

The following events are observed by a controller and sent to an analytics provider via Basic Authenticated GET w/ encoded URL.

User

user_created

user_deleted

pod_failed

ReplicationController

replicationcontroller_created

PersistentVolumeClaim

persistentvolumeclaim_created

persistentvolumeclaim_bound

deploymentconfig_failed

build_failed

Pod

pod_created

ReplicationRontroller

replicationrontroller_created

PersistentVolumeClaim

persistentvolumeclaim_created

Secret

secret_created

Service

service_created

Namespace

namespace_created

Deployment

deployment_created

Route

route_created

Build

build_created

RoleBinding

rolebinding_created

Template

template_created

ImageStream

imagestream_created