Ansible Variables

Overview

Ansible is the configuration management system used in Sunbird. The infrastructure setup, deployment of services and other configurations is handled primarily through ansible.

Updating the Private Repository with Hosts and Variables

Execute the following commands in your local to clone the private repo template and update your private GitHub repository -

git clone https://github.com/project-sunbird/sunbird-devops
cd sunbird-devops
git checkout tags/release-5.1.0 -b release-5.1.0

• Copy the directory sunbird-devops/private_repo/ansible to your private repo local workspace

• Run the key-generate.sh script located under ansible/inventory/dev folder. The script will ask for the ansible vault password. Enter the same vault password that you used in the Jenkins setup stage. The script will generate many private key files and they will be encrypted with the vault password. Deployments will fail if keys are not encrypted.

• Update the files common.yml, hosts, and secrets.yml under Core, KnowledgePlatform and DataPipeline directories. After updating, push them to your private repo branch

• Your private repo structure starting from the root path should be as shown below

  ansible
└── inventory
    └── dev
        ├── Core
        │   ├── common.yml
        │   ├── hosts
        │   └── secrets.yml
        ├── Kubernetes
        │   ├── common.yml   (soft link to Core directory files)
        │   ├── hosts.       (soft link to Core directory files)
        │   └── secrets.yml  (soft link to Core directory files)
        ├── DataPipeline
        │   ├── common.yml
        │   ├── hosts
        │   └── secrets.yml
        └── KnowledgePlatform
            ├── common.yml
            ├── hosts
            └── secrets.yml

Neo4j download and upload to object storage

• Neo4j community artifact needs to be downloaded from neo4j official website http://dist.neo4j.org/neo4j-community-3.3.9-unix.tar.gz (only Neo4j 3.4 and below is supported)

• Upload downloaded artifact to cloud_storage_artifacts_bucketname

• Update neo4j_zip variable with artifact name Eg: neo4j-community-3.3.9-unix.tar.gz

Note:

• The ansible inventory setup is a must before we can start to run jobs from the Provision, ArtifactUpload and Deploy directory in Jenkins. The Build directory on Jenkins does not depend on the ansible variables

• It is highly recommended that you complete the ansible inventory updates before proceeding further

List of Servers with their Ansible Group Names

Module	Servers	Service	Ansible Group Names
Build and Deploy	Server-1	Jenkins Master
Databases	Server-2	Cassandra	cassandra-1, lp-cassandra, dp-cassandra, core-cassandra, report-cassandra, cassandra-node-1, cassandra
		Postgres	postgresql-master-1, postgresql-master, postgres
		Application Elasticsearch	es-1, composite-search-cluster, core-es-1, core-es, es-backup, es
		Neo4j	learning-neo4j-node1, learning-neo4j-cluster
		Mongo	mongo_master, mongo
Knowledge Platform	Server-3	Redis	redis1, redis, redis-ps, redisall, lp-redis, lp-redis-ps, dp-redis, lms-redis, redis-exporter-targets
		Zookeeper	processing-cluster-zookeepers, ingestion-cluster-zookeeper, raw-zookeeper, zookeeper
		Kafka	processing-cluster-kafka, ingestion-cluster-kafka, kafka-1, kafka
Data Pipeline	Server-4	Spark	spark
		Kafka Indexer (Logstash)	kafka-indexer
		InfluxDB	influxdb
		Keycloak	keycloak-1, keycloak
Learning	Server-5	Learning	learning1, learning, learningall
		Graylog	graylog-1, graylog
Druid	Server-6	Druid	druid-postgres, raw-coordinator, raw-overlord, raw-broker, raw-historical, raw-middlemanager, raw-graphite, rollup-coordinator, dp-druid-broker, broker, coordinator, druid-raw
		Logs Elasticsearch	log-es-1, log-es-backup, log-es