Sparkline SQL extensions

Clear Druid Cache

clear druid cache

Druid Historical Server Segment assignments and other metadata about the Druid Segments and Servers are cached in the Driver(thriftserver or spark-shell). This metadata is exposed via Runtime View Any changes in the historical servers in the Druid Cluster are automatically picked up and these caches are rebuild. Use this command to explicitly clear the cached metadata; the next time a query is executed these caches will be rebuild.

Execute any Druid Query

Use this to execute arbitrary Druid Queries against a Druid DataSource. Note the json must include 'jsonClass' information for the json to be converted to a Druid Query Spec. The easiest way to get started is to cut-and-paste a query from the d$druidquery view

on druiddatasource orderLineItemPartSupplier_historical execute query {
    "jsonClass" : "GroupByQuerySpec",
    "queryType" : "groupBy",
    "dataSource" : "tpch",
    "dimensions" : [ {
      "jsonClass" : "ExtractionDimensionSpec",
      "type" : "extraction",
      "dimension" : "__time",
      "outputName" : "alias-9",
      "extractionFn" : {
        "jsonClass" : "TimeFormatExtractionFunctionSpec",
        "type" : "timeFormat",
        "format" : "YYYY-MM-dd 00:00:00",
        "timeZone" : "UTC",
        "locale" : "en_US"
      }
    } ],
    "granularity" : "all",
    "aggregations" : [ {
      "jsonClass" : "FunctionAggregationSpec",
      "type" : "doubleSum",
      "name" : "alias-10",
      "fieldName" : "l_extendedprice"
    } ],
    "intervals" : [ "1993-01-01T00:00:00.000Z/1997-12-31T00:00:01.000Z" ]
  }