Mailing-List: contact issues-help@flink.incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@flink.incubator.apache.org
From: fhueske <git@git.apache.org>
To: issues@flink.incubator.apache.org
Reply-To: issues@flink.incubator.apache.org
References: <git-pr-37-incubator-flink@git.apache.org>
In-Reply-To: <git-pr-37-incubator-flink@git.apache.org>
Subject: [GitHub] incubator-flink pull request: hadoopcompatibility:
 Implementations...
Content-Type: text/plain
Message-Id: <20140729091101.4531E9B9C99@tyr.zones.apache.org>
Date: Tue, 29 Jul 2014 09:11:00 +0000 (UTC)

Github user fhueske commented on a diff in the pull request:

    https://github.com/apache/incubator-flink/pull/37#discussion_r15512807
  
    --- Diff: flink-addons/flink-hadoop-compatibility/src/main/java/org/apache/flink/hadoopcompatibility/mapred/FlinkHadoopJobClient.java ---
    @@ -0,0 +1,427 @@
    +/**
    + * Licensed to the Apache Software Foundation (ASF) under one
    + * or more contributor license agreements.  See the NOTICE file
    + * distributed with this work for additional information
    + * regarding copyright ownership.  The ASF licenses this file
    + * to you 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 org.apache.flink.hadoopcompatibility.mapred;
    +
    +import org.apache.commons.logging.Log;
    +import org.apache.commons.logging.LogFactory;
    +import org.apache.flink.api.common.io.OutputFormat;
    +import org.apache.flink.api.java.ExecutionEnvironment;
    +import org.apache.flink.api.java.operators.DataSource;
    +import org.apache.flink.api.java.operators.FlatMapOperator;
    +import org.apache.flink.api.java.operators.ReduceGroupOperator;
    +import org.apache.flink.api.java.operators.UnsortedGrouping;
    +import org.apache.flink.api.java.typeutils.TypeExtractor;
    +import org.apache.flink.configuration.ConfigConstants;
    +import org.apache.flink.configuration.GlobalConfiguration;
    +import org.apache.flink.types.TypeInformation;
    +import org.apache.hadoop.conf.Configuration;
    +import org.apache.hadoop.fs.FileSystem;
    +import org.apache.hadoop.fs.Path;
    +import org.apache.hadoop.io.Text;
    +import org.apache.hadoop.mapred.ClusterStatus;
    +import org.apache.hadoop.mapred.Counters;
    +import org.apache.hadoop.mapred.InputFormat;
    +import org.apache.hadoop.mapred.JobClient;
    +import org.apache.hadoop.mapred.JobConf;
    +import org.apache.hadoop.mapred.JobID;
    +import org.apache.hadoop.mapred.JobQueueInfo;
    +import org.apache.hadoop.mapred.JobStatus;
    +import org.apache.hadoop.mapred.Reducer;
    +import org.apache.hadoop.mapred.RunningJob;
    +import org.apache.hadoop.mapred.TaskAttemptID;
    +import org.apache.hadoop.mapred.TaskCompletionEvent;
    +import org.apache.hadoop.mapred.TaskReport;
    +import org.apache.hadoop.mapred.lib.HashPartitioner;
    +import org.apache.hadoop.mapreduce.security.token.delegation.DelegationTokenIdentifier;
    +import org.apache.hadoop.security.token.Token;
    +
    +import java.io.IOException;
    +
    +/**
    + * The user's view of a Hadoop Job executed on a Flink cluster.
    + */
    +public class FlinkHadoopJobClient extends JobClient {
    +
    +	private final static int TASK_SLOTS = GlobalConfiguration.getInteger(ConfigConstants.TASK_MANAGER_NUM_TASK_SLOTS, -1);
    +	private static final Log LOG = LogFactory.getLog(FlinkHadoopJobClient.class);
    +
    +
    +	private ExecutionEnvironment environment;
    +	private Configuration hadoopConf;
    +
    +	public FlinkHadoopJobClient() throws IOException {
    +		this(new Configuration());
    +	}
    +
    +	public FlinkHadoopJobClient(JobConf jobConf) throws IOException {
    +		this(new Configuration(jobConf));
    +	}
    +
    +	public FlinkHadoopJobClient(Configuration hadoopConf) throws IOException{
    +		this(hadoopConf, (ExecutionEnvironment.getExecutionEnvironment()));
    +	}
    +
    +	public FlinkHadoopJobClient(Configuration hadoopConf, ExecutionEnvironment environment) throws IOException {
    +		this.hadoopConf = hadoopConf;
    +		this.environment = environment;
    +	}
    +
    +	public void init(JobConf conf) throws IOException {
    +		this.hadoopConf = conf;
    +	}
    +
    +	/**
    +	 * Submits a Hadoop job to Flink (as described by the JobConf) and returns after the job has been completed.
    +	 */
    +	public static RunningJob runJob(JobConf hadoopJobConf) throws IOException{
    +		final FlinkHadoopJobClient jobClient = new FlinkHadoopJobClient(hadoopJobConf);
    +		final RunningJob job = jobClient.submitJob(hadoopJobConf);
    +		job.waitForCompletion();
    +		return job;
    +	}
    +
    +	/**
    +	 * Submits a job to Flink and returns a RunningJob instance which can be scheduled and monitored
    +	 * without blocking by default. Use waitForCompletion() to block until the job is finished.
    +	 */
    +	@Override
    +	@SuppressWarnings("unchecked")
    +	public RunningJob submitJob(JobConf hadoopJobConf) throws IOException{
    +
    +		final int mapParallelism = getMapParallelism(hadoopJobConf);
    +		final int reduceParallelism = getReduceParallelism(hadoopJobConf);
    +
    +		//setting up the inputFormat for the job
    +		final DataSource<?> input = environment.createInput(getFlinkInputFormat(hadoopJobConf));
    +		input.setParallelism(mapParallelism);
    +
    +		final FlatMapOperator mapped = input.flatMap(new HadoopMapFunction(hadoopJobConf));
    +		mapped.setParallelism(getMapParallelism(hadoopJobConf));
    +
    +		//Partitioning
    +		final Class partitionerClass = hadoopJobConf.getPartitionerClass();
    +		if (! partitionerClass.equals(HashPartitioner.class)) {
    +			throw new UnsupportedOperationException("Custom partitioners are not supported yet.");
    +		}
    +		final UnsortedGrouping<?> grouping = mapped.groupBy(0);
    +
    +		final ReduceGroupOperator reduceOp = grouping.reduceGroup(new HadoopReduceFunction(hadoopJobConf));
    +		final Class<? extends Reducer> combinerClass = hadoopJobConf.getCombinerClass();
    +		if (combinerClass != null) {
    +			reduceOp.setCombinable(true);
    +		}
    +		reduceOp.setParallelism(reduceParallelism);
    +
    +		//Wrapping the output format.
    +		final OutputFormat<?> outputFormat = new HadoopOutputFormat(hadoopJobConf.getOutputFormat() ,hadoopJobConf);
    +		reduceOp.output(outputFormat).setParallelism(reduceParallelism);
    +
    +		return new DummyFlinkRunningJob(hadoopJobConf.getJobName());
    +	}
    +
    +	@SuppressWarnings("unchecked")
    +	private HadoopInputFormat<?,?> getFlinkInputFormat(JobConf jobConf) throws IOException{
    +		final InputFormat inputFormat = jobConf.getInputFormat();
    +		final Class inputFormatClass = inputFormat.getClass();
    +
    +		final TypeInformation keyTypeInfo = TypeExtractor.createTypeInfo(InputFormat.class, inputFormatClass,
    +				0, null, null);
    +		final Class keyClass = keyTypeInfo.getTypeClass();
    +
    +		final TypeInformation valueTypeInfo = TypeExtractor.createTypeInfo(InputFormat.class, inputFormatClass,
    +				1, null, null);
    +		final Class valueClass = valueTypeInfo.getTypeClass();
    +
    +		return new HadoopInputFormat(inputFormat, keyClass, valueClass, jobConf);
    +	}
    +
    +	/**
    +	 * The number of map tasks that can be run in parallel is the number of inputSplits.
    +	 * The usage of JobConf.setNumMapTasks() is deprecated.
    +	 * The upper bound for the number of parallel map tasks is the number of task slots.
    +	 */
    +	private int getMapParallelism(JobConf conf) throws IOException{
    --- End diff --
    
    the number of map tasks should be the minimum of
    - user specified value `conf.getNumMapTasks()`
    - the number of splits `conf.getInputFormat().getSplits(conf, 0).length`
    - the number of slots `TASK_SLOTS`


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastructure@apache.org or file a JIRA ticket
with INFRA.
---