esearch

# Settings suitable for operation on single node
settings = {
  :settings => {
    :number_of_shards => 1,
    :number_of_replicas => 0
  }
)

# Connect to a cluster
cluster = Esearch::Cluster.connect(uri, Logger.new($stderr, :debug)) }

# Create index 
index_a = cluster.index('test-a')
index_a.exist? # => false, does not exist jet
index_a.create(settings) 
index_a.exist? # => true

# Create another index
index_b = cluster.index('test-b')
index_b.create(SINGLE_NODE_SETTINGS)

# Wait for cluster initialization
cluster.health(:wait_for_status => :green, :timeout => '10s')

# Access many indexes at once
indices = cluster.indices(%w(test-a test-b))

# Index some data
result_a = index_a.type('type-a').index({'foo' => 'bar'})
result_b = index_b.type('type-b').index({'foo' => 'baz'})

# Refresh indices
indices.refresh

# Query all documents in index a
result = index_a.search({:query => { :match_all => {}}})
result.hits.map(&:source) # => [{'foo' => 'bar'}]

# Query all documents across two indexes
result = indices.search({:query => { :match_all => {}}})
result.hits.map(&:source) # => [{'foo' => 'bar'}, {'foo' => 'baz'}]

# Delete a document
index_a.type('type-a').document(result_a.id).delete

# Query a document by id
document = index_b.type('type-b').document(result_b.id).get
document.source # => {'foo' => 'baz'}