Skip to content



WherobotsDB supports many parameters if used together with Apache Spark. To change their values,

  1. Set it through SparkConf:
val config = SedonaContext.builder().
      config("spark.kryo.registrator", "org.apache.sedona.core.serde.SedonaKryoRegistrator").
  1. Check your current WherobotsDB configuration:
val sedonaConf = new SedonaConf(sedona.conf)
  1. WherobotsDB parameters can be changed at runtime:

Tuning for Spatial Join

WherobotsDB features an advanced spatial join algorithm since v1.2.1, which does not require tuning to achieve good performance. Advanced spatial join would analyze both joined datasets and tune spatial join parameters automatically. The following parameters for tuning spatial join won't work when using advanced spatial join:

  • sedona.join.indexbuildside
  • sedona.join.spatitionside

The advanced spatial join algorithm is enabled by default, users can disable advanced spatial join by setting sedona.join.advanced to false and tune spatial join parameters manually.


  • sedona.join.advanced
    • Using advanced spatial join algorithm
    • Default: true
    • Possible values: true, false
    • Use spatial index (currently, only supports in SQL range join and SQL distance join), only valid when "sedona.join.advanced" is false
    • Default: true
    • Possible values: true, false
    • Spatial index type, only valid when "" is true and "sedona.join.advanced" is false
    • Default: rtree
    • Possible values: rtree, quadtree
  • sedona.join.autoBroadcastJoinThreshold
    • Configures the maximum size in bytes for a table that will be broadcast to all worker nodes when performing a join. By setting this value to -1 automatic broadcasting can be disabled.
    • Default: The default value is the same as spark.sql.autoBroadcastJoinThreshold
    • Possible values: any integer with a byte suffix i.e. 10MB or 512KB
  • sedona.join.gridtype
    • Spatial partitioning grid type for join query
    • Default: kdbtree
    • Possible values: quadtree, kdbtree
  • sedona.join.indexbuildside (Advanced users only!)
    • The side which Sedona builds spatial indices on, only valid when "sedona.join.advanced" is false
    • Default: left
    • Possible values: left, right
  • sedona.join.numpartition (Advanced users only!)
    • Number of partitions for both sides in a join query
    • Default: -1, in this case it will be automatically tuned according to the size of both datasets when using advanced spatial join algorithm; when not using advanced spatial join it means use the existing partitions of the dominant side.
    • Possible values: any integers
  • sedona.join.spatitionside (Advanced users only!)
    • The dominant side in spatial partitioning stage, only valid when "sedona.join.advanced" is false
    • Default: left
    • Possible values: left, right
  • sedona.join.optimizationmode (Advanced users only!)
    • When should Sedona optimize spatial join SQL queries
    • Default: nonequi
    • Possible values:
      • all: Always optimize spatial join queries, even for equi-joins.
      • none: Disable optimization for spatial joins.
      • nonequi: Optimize spatial join queries that are not equi-joins.

Last update: May 20, 2024 02:13:05