Skip to content

Latest commit

 

History

History

ocp4-logging

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
.applier
.applier
 
 
.openshift
.openshift
 
 
README.md
README.md
 
 
galaxy-requirements.yml
galaxy-requirements.yml
 
 
requirements.yml
requirements.yml
 
 

README.md

Deploying Openshift v4.X Cluster-Logging Stack

Overview

This quickstart helps you deploy the cluster-logging stack for Openshift 4.X. This is not deployed by default and is accomplished through the subscribing to 2 operators via OperatorHub and then creating the appropriate custom resource.

This example deploys a 1 Elasticsearch and 1 Kibana EFK stack. This can be modified by updating the ELASTICSEARCH_NODE_COUNT and KIBANA_NODE_COUNT parameters in the Cluster Logging Params.

🚨If deploying a multi-node Elasticsearch cluster, make sure to also update the ELASTICSEARCH_REDUNDANCY_POLICY parameter to SingleRedundancy🚨

If you're interested in the other parameters for this logging stack, you can see them here

OCP Objects

This quickstart will create the following:

  • A namespace called openshift-logging with the appropriate labels
  • One OperatorGroup: openshift-logging-operatorgroup
  • Two CatalogSourceConfig objects created in the openshift-marketplace project: elasticsearch-operator and cluster-logging-operator
  • Two Subscription objects: cluster-logging in the openshift-logging project and elasticsearch-operator in the openshift-operators project
  • One ClusterLogging custom resource named instance in the openshift-logging project

Prerequisites

  • Openshift 4.X cluster
  • The appropriate permissions to deploy/create these projects
  • Openshift-Applier v2.1.2 or higher (this is accomplished via the ansible-galaxy step below)

Deployment

  1. ansible-galaxy install -r requirements.yml -p galaxy
  2. ansible-playbook -i .applier roles/openshift-applier/playbooks/openshift-cluster-seed.yml -e 'exclude_tags=cluster-logging-cr'
  3. ansible-playbook -i .applier roles/openshift-applier/playbooks/openshift-cluster-seed.yml -e 'include_tags=cluster-logging-cr'

This deployment needs to be run in two stages due to the fact that the ClusterLogging CustomResourceDefinition needs to exist prior to trying to deploy this CustomResource. The second step will deploy all components and operators from OperatorHub. The third step will then ensure that the cluster-logging stack is then deployed (which includes ElasticSearch, Kibana and Fluentd).

🚨 In between Step 2 and 3, you should verify that the appropriate operators are up and running before trying to deploy the ClusterLogging custom resource. You can do this by running oc get pods in both the openshift-operators and openshift-logging projects 🚨