All Downloads are FREE. Search and download functionalities are using the official Maven repository.

ml.shifu.guagua.master.MasterContext Maven / Gradle / Ivy

There is a newer version: 0.7.10
Show newest version
/*
 * Copyright [2013-2014] eBay Software Foundation
 *  
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *    http://www.apache.org/licenses/LICENSE-2.0
 *  
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
package ml.shifu.guagua.master;

import java.util.Properties;

import ml.shifu.guagua.GuaguaConstants;
import ml.shifu.guagua.io.Bytable;

/**
 * {@link MasterContext} is a context object which contains all useful info used in master computation.
 * 
 * 

* The info includes: *

    *
  • Application ID: Job ID for Hadoop mapreduce Job, application ID for YARN application.
  • *
  • Container ID: Task index for Hadoop mapreduce task, Task index for YARN Container.
  • *
  • Total iteration number.
  • *
  • Current iteration number.
  • *
  • Worker result list for current iteration.
  • *
  • Master result for current iteration which is used to be sent to workers in next iteration.
  • *
* * @param * master result for computation in each iteration. * @param * worker result for computation in each iteration. */ public class MasterContext { /** * Worker results for current iteration, should be set in coordination preIteration function. Type of * {@link #workerResults} is set as {@link Iterable} to save memory. If type is set as list, all results of workers * should be set into memory while by using {@link Iterable}, only one worker result when iterating is set into * memory, this is optimization for memory consumption in master. * * @see AbstractMasterCoordinator#setWorkerResults */ private Iterable workerResults; /** * Master result for current iteration, should be sent in coordination postIteration function. */ private MASTER_RESULT masterResult; /** * Current iteration, start from 1. 0 is used for initialization. */ private int currentIteration; /** * Total iteration number, set by client. */ private final int totalIteration; /** * How many workers, set by client, for MapReduce, it is set by getSplits */ private final int workers; /** * This props is used to store any key-value pairs for configurations. It will be set from hadoop configuration. The * reason we don't want to use Configuration is that we don't want to depend on hadoop for guagua-core. */ private final Properties props; /** * App id for whole application. For example: Job id in MapReduce(hadoop 1.0) or application id in Yarn. */ private final String appId; /** * Class name for master result in each iteration. We must have this field because of reflection need it to create a * new instance. */ private final String masterResultClassName; /** * Class name for worker result in each iteration. We must have this field because of reflection need it to create a * new instance. */ private final String workerResultClassName; /** * Container id in yarn, task attempt id in mapreduce. */ private final String containerId; /** * The ratio of minimal workers which are done to determine done of that iteration. */ private final double minWorkersRatio; /** * After this time elapsed, we can use {@link #minWorkersRatio} to determine done of that iteration. */ private final long minWorkersTimeOut; /** * This attachment is for {@link MasterComputable} and {@link MasterInterceptor} to transfer object. It can be set * by user for running time usage. * * @since 0.4.1 */ private Object attachment; public MasterContext(int totalIteration, int workers, Properties props, String appId, String containerId, String masterResultClassName, String workerResultClassName, double minWorkersRatio, long minWorkersTimeOut) { this.totalIteration = totalIteration; this.workers = workers; this.props = props; this.appId = appId; this.containerId = containerId; this.masterResultClassName = masterResultClassName; this.workerResultClassName = workerResultClassName; this.minWorkersRatio = minWorkersRatio; this.minWorkersTimeOut = minWorkersTimeOut; } public String getContainerId() { return containerId; } public Iterable getWorkerResults() { return workerResults; } public void setWorkerResults(Iterable workerResults) { this.workerResults = workerResults; } public int getTotalIteration() { return totalIteration; } public int getCurrentIteration() { return currentIteration; } public void setCurrentIteration(int currentIteration) { this.currentIteration = currentIteration; } public int getWorkers() { return workers; } public Properties getProps() { return props; } public String getAppId() { return appId; } public MASTER_RESULT getMasterResult() { return masterResult; } public void setMasterResult(MASTER_RESULT masterResult) { this.masterResult = masterResult; } public String getMasterResultClassName() { return masterResultClassName; } public String getWorkerResultClassName() { return workerResultClassName; } public double getMinWorkersRatio() { return minWorkersRatio; } public long getMinWorkersTimeOut() { return minWorkersTimeOut; } public Object getAttachment() { return attachment; } public void setAttachment(Object attachment) { this.attachment = attachment; } public boolean isFirstIteration() { return getCurrentIteration() == GuaguaConstants.GUAGUA_FIRST_ITERATION; } @Override public String toString() { return String .format("MasterContext [workerResults=%s, masterResult=%s, currentIteration=%s, totalIteration=%s, workers=%s, appId=%s, masterResultClassName=%s, workerResultClassName=%s, containerId=%s, minWorkersRatio=%s, minWorkersTimeOut=%s]", workerResults, masterResult, currentIteration, totalIteration, workers, appId, masterResultClassName, workerResultClassName, containerId, minWorkersRatio, minWorkersTimeOut); } }




© 2015 - 2024 Weber Informatics LLC | Privacy Policy