indexingTask.json

{
  "type" : "index_hadoop",
  "spec" : {
    "dataSchema" : {
      "dataSource" : "daily-data-source",
      "parser" : {
        "type" : "hadoopyString",
        "parseSpec" : {
          "format" : "json",
          "timestampSpec" : {
            "column" : "timestamp"
          },
          "dimensionsSpec" : {
            "dimensions" : [ ],
            "dimensionExclusions" : [ "timestamp", "d_x-hw-redirect", "d_ampce", "d_amprand" ],
            "spatialDimensions" : [ ]
          }
        }
      },
      "metricsSpec" : [ {
        "type" : "count",
        "name" : "count"
      }, {
        "type" : "hyperUnique",
        "name" : "gwid",
        "fieldName" : "gwid"
      } ],
      "granularitySpec" : {
        "type" : "uniform",
        "segmentGranularity" : "DAY",
        "queryGranularity" : "DAY",
        "intervals" : [ "2016-08-03T00:00:00.000Z/2016-08-04T00:00:00.000Z" ]
      }
    },
    "ioConfig" : {
      "type" : "hadoop",
      "inputSpec" : {
        "type" : "dataSource",
        "ingestionSpec" : {
          "dataSource" : "hourly-data-source",
          "intervals" : [ "2016-08-03T00:00:00.000Z/2016-08-04T00:00:00.000Z" ]
        }
      }
    },
    "tuningConfig" : {
      "type" : "hadoop",
      "combineText" : true,
      "useCombiner" : true,
      "jobProperties" : { },
      "numBackgroundPersistThreads" : 1
    }
  }
}