org.khaleesi.carfield.tools.sparkjobserver.api.ISparkJobServerClientConstants Maven / Gradle / Ivy
Show all versions of spark-job-server-client Show documentation
package org.khaleesi.carfield.tools.sparkjobserver.api;
/**
* Defines the constants used in public for
* the spark job server client.
*
* @author bluebreezecf
* @since 2014-09-11
*
*/
public interface ISparkJobServerClientConstants {
/**
* Parameter used to create job context, indicates the number of cpu cores
* for current spark job context.
*
* It's value is integer.
*
* It is equivalent to PARAM_SPARK_CORES_MAX.
*/
String PARAM_NUM_CPU_CORES = "num-cpu-cores";
/**
* Parameter used to create job context, indicates the number of cpu cores
* for current spark job context.
*
* It's value is integer.
*
* It is equivalent to PARAM_NUM_CPU_CORES.
*/
String PARAM_SPARK_CORES_MAX = "spark.cores.max";
/**
* Parameter used to create job context, indicates the number of memeory
* for each spark job executor.
*
* It's value is 'xxxm' such as 512m, where xxx is a integer, and m means MB.
*
* It is equivalent to PARAM_SPARK_EXECUTOR_MEMORY.
*/
String PARAM_MEM_PER_NODE = "mem-per-node";
/**
* Parameter used to create job context, indicates the number of memeory
* for each spark job executor.
*
* It's value is 'xxxm' such as 512m, where xxx is a integer, and m means MB.
*
* It is equivalent to PARAM_MEM_PER_NODE.
*/
String PARAM_SPARK_EXECUTOR_MEMORY = "spark.executor.memory";
/**
* It's a necessary parameter used to assign a application name when
* trying to start a new job.
*/
String PARAM_APP_NAME = "appName";
/**
* It's a necessary parameter which is a full qualified class name such as
* 'spark.jobserver.WorkCountExample', used to assign a spark
* job main class when trying to start a new job.
*/
String PARAM_CLASS_PATH = "classPath";
/**
* It's an optional parameter which is an existing context in the calling of
* GET /contexts, used to assign a context in which the new
* created job will run.
*
*
* When it wants the new started job to run synchronize and gets the result,
* It becomes a necessary one.
*/
String PARAM_CONTEXT = "context";
/**
* It's an optional parameter which could be "true" or "false",
* which is used to identify the client will wait for created
* job 's finish and get the results.
*
*
* When it wants the new started job to run synchronize and gets the result,
* It becomes a necessary one.
*/
String PARAM_SYNC = "sync";
}