Scaling Elasticsearch

create_alias.rb

req_body_for_patient_a = {
  routing: 'PatientA',
  filter: {
    term: { patient_name: 'PatientA' }
  }
}

req_body_for_patient_b = {
  routing: 'PatientB',
  filter: {
    term: { patient_name: 'PatientB' }
  }
}

raw_put 'patients/_alias/PatientA', req_body_for_patient_a
raw_put 'patients/_alias/PatientB', req_body_for_patient_b

client.indices.exists_alias? name: 'PatientA' # => true
client.indices.exists_alias? name: 'PatientB' # => true

index_docs.rb

# It's a good idea to disable refresh when indexing documents to improve performance
client.indices.put_settings index: 'PatientA', body: { refresh_interval: -1 }
client.indices.put_settings index: 'PatientB', body: { refresh_interval: -1 }

# Now index (store) the documents
patient_a_doc1 = {
  patient_name: 'PatientA',
  content: 'The quick brown fox'
}
patient_a_doc2 = {
  patient_name: 'PatientA',
  content: 'It was a dark and stormy night'
}

patient_b_doc1 = {
  patient_name: 'PatientB',
  content: 'Two roads diverged in a yellow wood'
}
patient_b_doc2 = {
  patient_name: 'PatientB',
  content: 'Lorem ipsum dolor sit amet'
}

bulk_index_body = [
  { index: { _index: 'PatientA', _type: 'patient_type', data: patient_a_doc1 } },
  { index: { _index: 'PatientA', _type: 'patient_type', data: patient_a_doc2 } },
  { index: { _index: 'PatientB', _type: 'patient_type', data: patient_b_doc1 } },
  { index: { _index: 'PatientB', _type: 'patient_type', data: patient_b_doc2 } }
]
client.bulk body: bulk_index_body

# Don't forget to enable refresh
client.indices.put_settings index: 'PatientA', body: { refresh_interval: '30s' }
client.indices.put_settings index: 'PatientB', body: { refresh_interval: '30s' }

primary_index.rb

def raw_put(path, body)
  conn = Faraday.new url: 'http://localhost:9200'
  conn.put(path) do |req|
    req.body = body.to_json
    req.headers['Content-Type'] = 'application/json'
  end
end

# Create primary index
req_body = {
  settings: {
    index: {
      number_of_shards:   5, # In production, this number would be much larger
      number_of_replicas: 1
    }
  },
  mappings: {
    patient_type: {
      properties: {
        patient_name: { index: 'not_analyzed', type: 'string' },
        content: { index: 'not_analyzed', type: 'string' }
      }
    }
  }
}
raw_put 'patients', req_body
client.indices.exists? index: 'patients' # => true

search.rb

pry(main)> client.search index: 'patients', q: 'content:The quick brown'
=> {"took"=>4,
 "timed_out"=>false,
 "_shards"=>{"total"=>5, "successful"=>5, "failed"=>0},
 "hits"=>
  {"total"=>1,
   "max_score"=>0.26442188,
   "hits"=>
    [{"_index"=>"patients",
      "_type"=>"patient_type",
      "_id"=>"AVkBDPk-1ECoVlT3taI2",
      "_score"=>0.26442188,
      "_routing"=>"PatientA",
      "_source"=>{"patient_name"=>"PatientA", "content"=>"The quick brown fox"}}]}}
pry(main)> client.search index: 'PatientA', q: 'content:The quick brown'
=> {"took"=>2,
 "timed_out"=>false,
 "_shards"=>{"total"=>1, "successful"=>1, "failed"=>0},
 "hits"=>
  {"total"=>1,
   "max_score"=>0.26442188,
   "hits"=>
    [{"_index"=>"patients",
      "_type"=>"patient_type",
      "_id"=>"AVkBDPk-1ECoVlT3taI2",
      "_score"=>0.26442188,
      "_routing"=>"PatientA",
      "_source"=>{"patient_name"=>"PatientA", "content"=>"The quick brown fox"}}]}}

setup.rb

require 'json'
require 'elasticsearch'

client = Elasticsearch::Client.new hosts: 'http://localhost:9200'

# Test that Elasticsearch is up
client.info

setup_results.rb

pry(main)> client.info
=> {"name"=>"Dredmund Druid",
 "cluster_name"=>"elasticsearch_mkim",
 "cluster_uuid"=>"f19ZvZWLTsaJeUM9ww1nQw",
 "version"=>
  {"number"=>"2.4.2",
   "build_hash"=>"161c65a337d4b422ac0c805f284565cf2014bb84",
   "build_timestamp"=>"2016-11-17T11:51:03Z",
   "build_snapshot"=>false,
   "lucene_version"=>"5.5.2"},
 "tagline"=>"You Know, for Search"}