EmrConfiguration - AWS Data Pipeline


The EmrConfiguration object is the configuration used for EMR clusters with releases 4.0.0 or greater. Configurations (as a list) is a parameter to the RunJobFlow API call. The configuration API for Amazon EMR takes a classification and properties. AWS Data Pipeline uses EmrConfiguration with corresponding Property objects to configure an EmrCluster application such as Hadoop, Hive, Spark, or Pig on EMR clusters launched in a pipeline execution. Because configuration can only be changed for new clusters, you cannot provide a EmrConfiguration object for existing resources. For more information, see http://docs.aws.amazon.com/ElasticMapReduce/latest/ReleaseGuide/.


The following configuration object sets the io.file.buffer.size and fs.s3.block.size properties in core-site.xml:

[ { "classification":"core-site", "properties": { "io.file.buffer.size": "4096", "fs.s3.block.size": "67108864" } } ]

The corresponding pipeline object definition uses a EmrConfiguration object and a list of Property objects in the property field:

{ "objects": [ { "name": "ReleaseLabelCluster", "releaseLabel": "emr-4.1.0", "applications": ["spark", "hive", "pig"], "id": "ResourceId_I1mCc", "type": "EmrCluster", "configuration": { "ref": "coresite" } }, { "name": "coresite", "id": "coresite", "type": "EmrConfiguration", "classification": "core-site", "property": [{ "ref": "io-file-buffer-size" }, { "ref": "fs-s3-block-size" } ] }, { "name": "io-file-buffer-size", "id": "io-file-buffer-size", "type": "Property", "key": "io.file.buffer.size", "value": "4096" }, { "name": "fs-s3-block-size", "id": "fs-s3-block-size", "type": "Property", "key": "fs.s3.block.size", "value": "67108864" } ] }

The following example is a nested configuration used to set the Hadoop environment with the hadoop-env classification:

[ { "classification": "hadoop-env", "properties": {}, "configurations": [ { "classification": "export", "properties": { "YARN_PROXYSERVER_HEAPSIZE": "2396" } } ] } ]

The corresponding pipeline definition object that uses this configuration is below:

{ "objects": [ { "name": "ReleaseLabelCluster", "releaseLabel": "emr-4.0.0", "applications": ["spark", "hive", "pig"], "id": "ResourceId_I1mCc", "type": "EmrCluster", "configuration": { "ref": "hadoop-env" } }, { "name": "hadoop-env", "id": "hadoop-env", "type": "EmrConfiguration", "classification": "hadoop-env", "configuration": { "ref": "export" } }, { "name": "export", "id": "export", "type": "EmrConfiguration", "classification": "export", "property": { "ref": "yarn-proxyserver-heapsize" } }, { "name": "yarn-proxyserver-heapsize", "id": "yarn-proxyserver-heapsize", "type": "Property", "key": "YARN_PROXYSERVER_HEAPSIZE", "value": "2396" }, ] }

The following example modifies a Hive-specific property for an EMR cluster:

{ "objects": [ { "name": "hivesite", "id": "hivesite", "type": "EmrConfiguration", "classification": "hive-site", "property": [ { "ref": "hive-client-timeout" } ] }, { "name": "hive-client-timeout", "id": "hive-client-timeout", "type": "Property", "key": "hive.metastore.client.socket.timeout", "value": "2400s" } ] }


This object includes the following fields.

Required Fields Description Slot Type
classification Classification for the configuration. String

Optional Fields Description Slot Type
configuration Sub-configuration for this configuration. Reference Object, e.g. "configuration":{"ref":"myEmrConfigurationId"}
parent Parent of the current object from which slots will be inherited. Reference Object, e.g. "parent":{"ref":"myBaseObjectId"}
property Configuration property. Reference Object, e.g. "property":{"ref":"myPropertyId"}

Runtime Fields Description Slot Type
@version Pipeline version the object was created with. String

System Fields Description Slot Type
@error Error describing the ill-formed object String
@pipelineId Id of the pipeline to which this object belongs to String
@sphere The sphere of an object denotes its place in the lifecycle: Component Objects give rise to Instance Objects which execute Attempt Objects String

See Also