Solr Ansible role

This ansible role installs a Solr server in a debian environment.

Solr Ansible role

Getting Started

These instructions will get you a copy of the role for your ansible playbook. Once launched, it will install a Solr server in a Debian system.

This role is tested on:

Debian
- stretch (java 8 and 11)
- buster (java 11)
Ubuntu
- focal (java 8 and 14)

Prerequisities

Ansible 2.8.8 version installed. Inventory destination should be a Debian environment.

For testing purposes, Molecule with Docker as driver.

Installing

Create or add to your roles dependency file (e.g requirements.yml):

- src: idealista.solr_role
  version: x.x.x
  name: solr

Install the role with ansible-galaxy command:

ansible-galaxy install -p roles -r requirements.yml -f

Use in a playbook:

---
- hosts: someserver
  roles:
    - { role: solr }

Playbook example below showing how to provision from scratch a SolrCloud cluster with two nodes plus create an example (and empty) collection called sample_techproducts_configs, using idealista java, zookeeper and solr roles:

Note: Assuming that 'solrcloud' group has two nodes (solrcloud1 and solrcloud2) as is declared in molecule.yml, collection will have two shards, one replica and one shard per node as is declared in group vars file called solrcloud.yml and configuration files are stored under directory called sample_techproducts_configs under template directory.

⚠️ Use the example below just as a reference, requires inventory host groups solr and zookeeper to be correctly defined

---

- hosts: zookeeper
  roles:
    - role: zookeeper
  pre_tasks:
    - name: installing required libs
      apt:
        pkg: "{{ item }}"
        state: present
      with_items:
        - net-tools
        - netcat

- hosts: solrcloud
  roles:
    - role: solr_role

Usage

Look to the defaults properties file to see the possible configuration properties.

Add JVM Agent to your installation

This role supports JVM agents (such as Newrelic, Datadog, etc.) to be used inside your installation. You can view an example with JVM Agents in the molecule tests section --> example_setup_with_agent.

Its very simple, must follow this steps (in this case we will add config for Newrelic's agent):

Just add the config in your group_vars.

solr_agents_required_libs:
  - unzip
  - apt-transport-https
solr_agents_config:
  - name: "newrelic"
    download_url: "http://download.newrelic.com/newrelic/java-agent/newrelic-agent/current/newrelic-java.zip"
    vm_opts:
      - '-javaagent:{{ solr_installation_dir }}/newrelic/newrelic.jar'
    configuration_files:
      - "newrelic.yml"
    params: {
      application_name: "application_sample_name",
      license_key: "your_license_key"
    }

Optional: Place the configuration files in the templates folder using this order "templates/{{ agent_name }}/{{ file names specified in solr_agents_config.configuration_files }}.j2. In this case we have the newrelic.yml.j2 in templates/agents/newrelic/newrelic.yml.j2.

Prometheus Exporter

If you want to scrape metrics from Solr using Prometheus, you will need to configure a exporter. We have a Prometheus Solr Exporter role that will make configuration easier for you, just keep in mind that the variables solr_version and prometheus_solr_exporter_version must have the same value.

Testing

$ pipenv sync
$ pipenv shell

# This will execute tests but doesn't destroy created environment (because of --destroy=never)
$ molecule test --destroy=never -s setup_with_collections

Solr Admin UI should be accessible from docker container host at URL:

http://localhost:8983/solr/#/ (node: solrcloud1)

http://localhost:8984/solr/#/ (node: solrcloud2)