GIS Cloud Suite

Open Distributed Analyst Service

Please refer to the introduction of Computing Resources( GIS Cloud Suite > Computing Resources) to add the computing resource before opening Distributed Analyst Service, and follow the steps below to open Distributed Analyst Service:

1. Clicks Service Management > Distributed Analyst Service on the left navigation bar;

2. Clicks on Enable;

   

3. Select a computing resource;

   

   • Publish analysis results as a service: The results will be published to a service automatically if enable the button.

   If you use built-in Spark, you can set the following parameters in Advanced setting:

   • spark.executor.memory：The memory size of each executor porcess.
   • spark.exector.cores：The number of CPU cores used in each executor process.
   • spark.cores.max：The maximum number of CPU cores used in Distributed Analyst Service.
   • spark.driver.maxResultSize：The maximum serialized results for all partitions of the spark operation(e,g. collect), 0 means unlimited.
   • spark.driver.memory：The memory size of Distributed Analyst Service.
   • spark.driver.memoryOverhead：The off-heap memory of Distributed Analyst Service.
   • spark.memory.fraction：The rate of executor memory and total memory.The range is between 0 - 1.
   • spark.memory.offHeap.enabled：Allows the Spark to use off-heap memory.
   • spark.memory.offHeap.size：The absolute amount of off-heap memory, it is necessary to open the park.driver.offHeap.enabled first.
   • spark.network.timeout：The timeout of network interactions.

4. Clicks on OK, the prompt would appeard on the upper right side;

   

5. After the Status change to “1/1”, the Distributed Analyst Service is ready to use;

6. Clicks on the link under the Address to enter the Distributed Analyst Service interface.

   Notes:

   More details of how to use Distributed Analyst Service please refer to SuperMap iServer documentation(Using iServer > Using distributed analysis service).