def test_allow_string_cluster(self):
spot = "2"
withSpotPrice = "WithSpotPrice"
cluster = emr.Cluster(
'Cluster',
# AdditionalInfo="Additional Info",
Applications=[
emr.Application(Name="Hadoop"),
emr.Application(Name="Hive"),
emr.Application(Name="Mahout"),
emr.Application(Name="Pig"),
emr.Application(Name="Spark")
],
BootstrapActions=[
emr.BootstrapActionConfig(
Name='Dummy bootstrap action',
ScriptBootstrapAction=emr.ScriptBootstrapActionConfig(
Path='file:/usr/share/aws/emr/scripts/install-hue',
Args=["dummy", "parameter"]))
],
Configurations=[
emr.Configuration(Classification="core-site",
ConfigurationProperties={
'hadoop.security.groups.cache.secs':
'250'
})
],
Instances=emr.JobFlowInstancesConfig(
Ec2KeyName="KeyName",
Ec2SubnetId="SubnetId",
MasterInstanceGroup=emr.InstanceGroupConfigProperty(
InstanceCount="1",
InstanceType=M4_LARGE,
AutoScalingPolicy=emr.AutoScalingPolicy(
Constraints=emr.ScalingConstraints(MinCapacity="1",
MaxCapacity="3"),
Rules=self.generate_rules("MasterAutoScalingPolicy")),
),
CoreInstanceGroup=emr.InstanceGroupConfigProperty(
Name="Core Instance",
BidPrice=If(withSpotPrice, Ref(spot), Ref("AWS::NoValue")),
Market=If(withSpotPrice, "SPOT", "ON_DEMAND"),
InstanceCount="1",
InstanceType=M4_LARGE,
AutoScalingPolicy=emr.AutoScalingPolicy(
Constraints=emr.ScalingConstraints(MinCapacity="1",
MaxCapacity="3"),
Rules=self.generate_rules("CoreAutoScalingPolicy"),
)),
),
JobFlowRole="EMRJobFlowRole",
LogUri="s3://cluster-logs",
Name="EMR Cluster",
ReleaseLabel="emr-5.5.0",
ServiceRole="EMRServiceRole",
AutoScalingRole="EMR_AutoScaling_DefaultRole",
VisibleToAllUsers="true",
Tags=Tags(Name="EMR Sample Cluster"))
cluster.to_dict()
JobFlowRole=Ref(emr_instance_profile),
ServiceRole=Ref(emr_service_role),
Instances=emr.JobFlowInstancesConfig(
MasterInstanceGroup=emr.InstanceGroupConfigProperty(
Name="Master Instance",
InstanceCount="1",
InstanceType="m3.xlarge",
Market="ON_DEMAND"),
CoreInstanceGroup=emr.InstanceGroupConfigProperty(
Name="Core Instance",
BidPrice="20",
InstanceCount="1",
InstanceType="m3.xlarge",
Market="SPOT")),
Applications=[
emr.Application(Name="Hadoop"),
emr.Application(Name="Hive"),
emr.Application(Name="Mahout"),
emr.Application(Name="Pig"),
emr.Application(Name="Spark")
],
VisibleToAllUsers="true",
Tags=Tags(Name="EMR Sample Cluster")))
step = template.add_resource(
emr.Step('TestStep',
Name="TestStep",
ActionOnFailure='CONTINUE',
HadoopJarStep=emr.HadoopJarStepConfig(
Args=["5", "10"],
Jar="s3://emr-cfn-test/hadoop-mapreduce-examples-2.6.0.jar",
def test_allow_string_cluster(self):
cluster_security_configuration = emr.SecurityConfiguration(
'emrsecurityconfiguration',
Name="EMRSecurityConfiguration",
SecurityConfiguration=security_configuration)
spot = "2"
withSpotPrice = "WithSpotPrice"
cluster = emr.Cluster(
'Cluster',
# AdditionalInfo="Additional Info",
Applications=[
emr.Application(Name="Hadoop"),
emr.Application(Name="Hive"),
emr.Application(Name="Mahout"),
emr.Application(Name="Pig"),
emr.Application(Name="Spark")
],
BootstrapActions=[
emr.BootstrapActionConfig(
Name='Dummy bootstrap action',
ScriptBootstrapAction=emr.ScriptBootstrapActionConfig(
Path='file:/usr/share/aws/emr/scripts/install-hue',
Args=["dummy", "parameter"]))
],
Configurations=[
emr.Configuration(Classification="core-site",
ConfigurationProperties={
'hadoop.security.groups.cache.secs':
'250'
})
],
Instances=emr.JobFlowInstancesConfig(
Ec2KeyName="KeyName",
Ec2SubnetId="SubnetId",
MasterInstanceGroup=emr.InstanceGroupConfigProperty(
InstanceCount="1",
InstanceType=M4_LARGE,
AutoScalingPolicy=emr.AutoScalingPolicy(
Constraints=emr.ScalingConstraints(MinCapacity="1",
MaxCapacity="3"),
Rules=self.generate_rules("MasterAutoScalingPolicy")),
),
CoreInstanceGroup=emr.InstanceGroupConfigProperty(
Name="Core Instance",
BidPrice=If(withSpotPrice, Ref(spot), Ref("AWS::NoValue")),
Market=If(withSpotPrice, "SPOT", "ON_DEMAND"),
InstanceCount="1",
InstanceType=M4_LARGE,
AutoScalingPolicy=emr.AutoScalingPolicy(
Constraints=emr.ScalingConstraints(MinCapacity="1",
MaxCapacity="3"),
Rules=self.generate_rules("CoreAutoScalingPolicy"),
)),
),
JobFlowRole="EMRJobFlowRole",
LogUri="s3://cluster-logs",
Name="EMR Cluster",
ReleaseLabel="emr-5.5.0",
SecurityConfiguration=Ref(cluster_security_configuration),
ServiceRole="EMRServiceRole",
AutoScalingRole="EMR_AutoScaling_DefaultRole",
VisibleToAllUsers="true",
Tags=Tags(Name="EMR Sample Cluster"))
cluster.to_dict()
autoscale_policy = emr.AutoScalingPolicy(
Constraints=emr.ScalingConstraints(MinCapacity=0, MaxCapacity=5),
Rules=[
emr.ScalingRule(
Name='ScaleUpContainerPending',
Description='Scale up on over-provisioned '
'containers',
Action=emr.ScalingAction(
SimpleScalingPolicyConfiguration=emr.
SimpleScalingPolicyConfiguration(
AdjustmentType=emr.CHANGE_IN_CAPACITY,
CoolDown=300,
ScalingAdjustment=1)),
Trigger=emr.ScalingTrigger(
CloudWatchAlarmDefinition=emr.
CloudWatchAlarmDefinition(
ComparisonOperator='GREATER_THAN',
MetricName='ContainerPendingRatio',
Period=300,
Threshold=0.75,
Dimensions=[
emr.MetricDimension(Key='JobFlowId',
Value='${emr.clusterId}')
]))),
emr.ScalingRule(
Name='ScaleUpMemory',
Description='Scale up on low memory',
Action=emr.ScalingAction(
SimpleScalingPolicyConfiguration=emr.
SimpleScalingPolicyConfiguration(
AdjustmentType='CHANGE_IN_CAPACITY',
CoolDown=300,
ScalingAdjustment=1)),
Trigger=emr.ScalingTrigger(
CloudWatchAlarmDefinition=emr.
CloudWatchAlarmDefinition(
ComparisonOperator='LESS_THAN',
MetricName='YARNMemoryAvailablePercentage',
Period=300,
Threshold=15,
Dimensions=[
emr.MetricDimension(Key='JobFlowId',
Value='${emr.clusterId}')
]))),
emr.ScalingRule(
Name='ScaleDownMemory',
Description='Scale down on high memory',
Action=emr.ScalingAction(
SimpleScalingPolicyConfiguration=emr.
SimpleScalingPolicyConfiguration(
AdjustmentType=emr.CHANGE_IN_CAPACITY,
CoolDown=300,
ScalingAdjustment=-1)),
Trigger=emr.ScalingTrigger(
CloudWatchAlarmDefinition=emr.
CloudWatchAlarmDefinition(
ComparisonOperator='GREATER_THAN',
MetricName='YARNMemoryAvailablePercentage',
Period=300,
Threshold=75,
Dimensions=[
emr.MetricDimension(Key='JobFlowId',
Value='${emr.clusterId}')
])))
])
emr.InstanceGroupConfig('TaskInstanceGroup',
AutoScalingPolicy=autoscale_policy,
InstanceCount=0,
InstanceType=M4_LARGE,
InstanceRole='TASK',
Market='ON_DEMAND',
Name='Task Instance',
JobFlowId=Ref(cluster))
"PYTHONPATH":
os.path.join('/home/hadoop/miniconda',
'bin/python') +
":/usr/lib/spark/python/:$PYTHONPATH",
"PYSPARK_DRIVER_PYTHON":
os.path.join('/home/hadoop/miniconda',
'bin/python'),
"SPARK_HOME":
"/usr/lib/spark",
"PYTHONHASHSEED":
"123"
})
]),
],
Applications=[
emr.Application(Name=app) for app in cfg['applications']
],
VisibleToAllUsers='true',
Tags=Tags(Name='jupyter-cluster')))
# Outputs
template.add_output([
Output("EMRCluster", Description="EMRCluster", Value=Ref(cluster)),
Output("EMRClusterMasterDNS",
Description="EMRCluster",
Value=GetAtt(cluster, 'MasterPublicDNS')),
])
template_json = template.to_json(indent=4)
print(template_json)
def build(ssh_keypair_name):
template = Template()
template.set_version("2010-09-09")
keyname_param = template.add_parameter(
Parameter(
"KeyName",
ConstraintDescription=
"must be the name of an existing EC2 KeyPair.",
Description=
"Name of an existing EC2 KeyPair to enable SSH access to \
the instance",
Type="AWS::EC2::KeyPair::KeyName",
Default=ssh_keypair_name,
))
sshlocation_param = template.add_parameter(
Parameter(
"SSHLocation",
Description=
" The IP address range that can be used to SSH to the EC2 \
instances",
Type="String",
MinLength="9",
MaxLength="18",
Default="0.0.0.0/0",
AllowedPattern=
r"(\d{1,3})\.(\d{1,3})\.(\d{1,3})\.(\d{1,3})/(\d{1,2})",
ConstraintDescription=(
"must be a valid IP CIDR range of the form x.x.x.x/x."),
))
vpc = template.add_resource(VPC("VPC", CidrBlock="10.0.0.0/16"))
subnet = template.add_resource(
Subnet(
"Subnet",
CidrBlock="10.0.0.0/24",
VpcId=Ref(vpc),
))
internet_gateway = template.add_resource(
InternetGateway("InternetGateway"))
attach_gateway = template.add_resource(
VPCGatewayAttachment("AttachGateway",
VpcId=Ref(vpc),
InternetGatewayId=Ref(internet_gateway)))
route_table = template.add_resource(
RouteTable("RouteTable", VpcId=Ref(vpc)))
template.add_resource(
Route(
"Route",
DependsOn=attach_gateway,
GatewayId=Ref(internet_gateway),
DestinationCidrBlock="0.0.0.0/0",
RouteTableId=Ref(route_table),
))
template.add_resource(
SubnetRouteTableAssociation(
"SubnetRouteTableAssociation",
SubnetId=Ref(subnet),
RouteTableId=Ref(route_table),
))
network_acl = template.add_resource(
NetworkAcl("NetworkAcl", VpcId=Ref(vpc)))
template.add_resource(
NetworkAclEntry(
"InboundSSHNetworkAclEntry",
NetworkAclId=Ref(network_acl),
RuleNumber="101",
Protocol="6",
PortRange=PortRange(To="22", From="22"),
Egress="false",
RuleAction="allow",
CidrBlock="0.0.0.0/0",
))
template.add_resource(
NetworkAclEntry(
"InboundResponsePortsNetworkAclEntry",
NetworkAclId=Ref(network_acl),
RuleNumber="102",
Protocol="6",
PortRange=PortRange(To="65535", From="1024"),
Egress="false",
RuleAction="allow",
CidrBlock="0.0.0.0/0",
))
template.add_resource(
NetworkAclEntry(
"OutBoundResponsePortsNetworkAclEntry",
NetworkAclId=Ref(network_acl),
RuleNumber="103",
Protocol="6",
PortRange=PortRange(To="65535", From="1024"),
Egress="true",
RuleAction="allow",
CidrBlock="0.0.0.0/0",
))
template.add_resource(
NetworkAclEntry(
"OutBoundHTTPPortsNetworkAclEntry",
NetworkAclId=Ref(network_acl),
RuleNumber="104",
Protocol="6",
PortRange=PortRange(To="80", From="80"),
Egress="true",
RuleAction="allow",
CidrBlock="0.0.0.0/0",
))
template.add_resource(
NetworkAclEntry(
"OutBoundHTTPSPortsNetworkAclEntry",
NetworkAclId=Ref(network_acl),
RuleNumber="105",
Protocol="6",
PortRange=PortRange(To="443", From="443"),
Egress="true",
RuleAction="allow",
CidrBlock="0.0.0.0/0",
))
template.add_resource(
NetworkAclEntry(
"OutBoundSSHPortsNetworkAclEntry",
NetworkAclId=Ref(network_acl),
RuleNumber="106",
Protocol="6",
PortRange=PortRange(To="22", From="22"),
Egress="true",
RuleAction="allow",
CidrBlock="0.0.0.0/0",
))
template.add_resource(
SubnetNetworkAclAssociation("SubnetNetworkAclAssociation",
SubnetId=Ref(subnet),
NetworkAclId=Ref(network_acl)))
emr_security_group = template.add_resource(
SecurityGroup(
"EMRSecurityGroup",
GroupDescription="Enable SSH access via port 22",
SecurityGroupIngress=[
SecurityGroupRule(IpProtocol="tcp",
FromPort="22",
ToPort="22",
CidrIp=Ref(sshlocation_param)),
],
VpcId=Ref(vpc),
))
emr_service_role = template.add_resource(
iam.Role(
"EMRServiceRole",
AssumeRolePolicyDocument={
"Statement": [{
"Effect": "Allow",
"Principal": {
"Service": ["elasticmapreduce.amazonaws.com"]
},
"Action": ["sts:AssumeRole"],
}]
},
ManagedPolicyArns=[
"arn:aws:iam::aws:policy/service-role/AmazonElasticMapReduceRole"
],
))
emr_job_flow_role = template.add_resource(
iam.Role(
"EMRJobFlowRole",
AssumeRolePolicyDocument={
"Statement": [{
"Effect": "Allow",
"Principal": {
"Service": ["ec2.amazonaws.com"]
},
"Action": ["sts:AssumeRole"]
}]
},
ManagedPolicyArns=[
"arn:aws:iam::aws:policy/service-role/AmazonElasticMapReduceforEC2Role"
],
))
emr_instance_profile = template.add_resource(
iam.InstanceProfile("EMRInstanceProfile",
Roles=[Ref(emr_job_flow_role)]))
cluster = template.add_resource(
emr.Cluster(
"EMRCluster",
ReleaseLabel="emr-6.0.0",
JobFlowRole=Ref(emr_instance_profile),
ServiceRole=Ref(emr_service_role),
Instances=emr.JobFlowInstancesConfig(
Ec2KeyName=Ref(keyname_param),
Ec2SubnetId=Ref(subnet),
EmrManagedMasterSecurityGroup=Ref(emr_security_group),
EmrManagedSlaveSecurityGroup=Ref(emr_security_group),
MasterInstanceGroup=emr.InstanceGroupConfigProperty(
Name="Master Instance",
InstanceCount="1",
InstanceType=M4_LARGE,
Market="ON_DEMAND",
),
CoreInstanceGroup=emr.InstanceGroupConfigProperty(
Name="Core Instance",
EbsConfiguration=emr.EbsConfiguration(
EbsBlockDeviceConfigs=[
emr.EbsBlockDeviceConfigs(
VolumeSpecification=emr.VolumeSpecification(
SizeInGB="10", VolumeType="gp2"),
VolumesPerInstance="1",
)
],
EbsOptimized="true",
),
InstanceCount="1",
InstanceType=M4_LARGE,
),
),
Applications=[emr.Application(Name="Spark")],
VisibleToAllUsers="true",
))
template.add_output(
[Output("MasterPublicDNS", Value=GetAtt(cluster, "MasterPublicDNS"))])
return template
