Contents

# Configures the Elastic MapReduce cluster that is launched to run parse and
# combine jobs. The list of EC2 instance types can be found at
# http://aws.amazon.com/ec2/instance-types/#instance-details

# Using spot instances is recommended to reduce costs. However if the spot
# price rises above your bid price the cluster may be terminated. Elasticrawl
# tries to reduce the effect of this by parsing each Commmon Crawl segment
# in a separate job flow step.

# The master node manages the cluster.
master_instance_group:
  instance_type: m1.medium
  use_spot_instances: true
  bid_price: 0.120

# Core nodes run map and reduce tasks and store data using HDFS.
core_instance_group:
  instance_type: m1.medium
  instance_count: 2
  use_spot_instances: true
  bid_price: 0.120

# Task nodes are optional and only run map and reduce tasks.
task_instance_group:
  instance_type: m1.small
  instance_count: 0
  use_spot_instances: true
  bid_price: 0.080

# Array of bootstrap scripts that will be applied when the cluster nodes are
# initialized. The example installs the Ganglia distributed monitoring system.
bootstrap_scripts: #['s3://elasticmapreduce/bootstrap-actions/install-ganglia']

# Specifying an EC2 key pair allows SSH access to the master node. This also
# allows accessing the Hadoop Web UI over an SSH tunnel.
ec2_key_name: # 'key-pair-name'

# Availability Zone (AZ) to launch instances in. An AZ in the US-East region is
# recommended since the Common Crawl corpus is stored there. Otherwise inter
# region data transfer charges will apply.
placement: 'us-east-1a'

# The AMI version to use when launching instances.
emr_ami_version: 'latest'

# Default instance profile
job_flow_role: 'EMR_EC2_DefaultRole'

# Default service role
service_role: 'EMR_DefaultRole'

# Subnet ID. Required for new Amazon accounts launching more powerful instance types.
ec2_subnet_id: 'subnet-name'

Version data entries

1 entries across 1 versions & 1 rubygems

Version	Path
elasticrawl-1.1.8	templates/cluster.yml