Source code

001/**
002 * Licensed to the Apache Software Foundation (ASF) under one
003 * or more contributor license agreements.  See the NOTICE file
004 * distributed with this work for additional information
005 * regarding copyright ownership.  The ASF licenses this file
006 * to you under the Apache License, Version 2.0 (the
007 * "License"); you may not use this file except in compliance
008 * with the License.  You may obtain a copy of the License at
009 *
010 *      http://www.apache.org/licenses/LICENSE-2.0
011 *
012 * Unless required by applicable law or agreed to in writing, software
013 * distributed under the License is distributed on an "AS IS" BASIS,
014 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
015 * See the License for the specific language governing permissions and
016 * limitations under the License.
017 */
018
019package org.apache.oozie.action.hadoop;
020
021import java.io.BufferedReader;
022import java.io.IOException;
023import java.io.InputStream;
024import java.io.InputStreamReader;
025import java.io.OutputStream;
026import java.math.BigInteger;
027import java.security.MessageDigest;
028import java.security.NoSuchAlgorithmException;
029import java.security.PrivilegedExceptionAction;
030import java.util.ArrayList;
031import java.util.Collection;
032import java.util.HashMap;
033import java.util.List;
034import java.util.Map;
035import java.util.Properties;
036
037import org.apache.hadoop.conf.Configuration;
038import org.apache.hadoop.fs.FileSystem;
039import org.apache.hadoop.fs.Path;
040import org.apache.hadoop.io.SequenceFile;
041import org.apache.hadoop.io.Text;
042import org.apache.hadoop.mapred.JobConf;
043import org.apache.hadoop.mapred.RunningJob;
044import org.apache.hadoop.mapred.Counters;
045import org.apache.hadoop.security.UserGroupInformation;
046import org.apache.oozie.client.OozieClient;
047import org.apache.oozie.client.WorkflowAction;
048import org.apache.oozie.service.HadoopAccessorException;
049import org.apache.oozie.service.HadoopAccessorService;
050import org.apache.oozie.service.Services;
051import org.apache.oozie.service.URIHandlerService;
052import org.apache.oozie.service.UserGroupInformationService;
053import org.apache.oozie.util.IOUtils;
054import org.apache.oozie.util.PropertiesUtils;
055
056public class LauncherMapperHelper {
057
058    public static String getRecoveryId(Configuration launcherConf, Path actionDir, String recoveryId)
059            throws HadoopAccessorException, IOException {
060        String jobId = null;
061        Path recoveryFile = new Path(actionDir, recoveryId);
062        FileSystem fs = Services.get().get(HadoopAccessorService.class)
063                .createFileSystem(launcherConf.get("user.name"),recoveryFile.toUri(), launcherConf);
064
065        if (fs.exists(recoveryFile)) {
066            InputStream is = fs.open(recoveryFile);
067            BufferedReader reader = new BufferedReader(new InputStreamReader(is));
068            jobId = reader.readLine();
069            reader.close();
070        }
071        return jobId;
072
073    }
074
075    public static void setupMainClass(Configuration launcherConf, String javaMainClass) {
076        // Only set the javaMainClass if its not null or empty string, this way the user can override the action's main class via
077        // <configuration> property
078        if (javaMainClass != null && !javaMainClass.equals("")) {
079            launcherConf.set(LauncherMapper.CONF_OOZIE_ACTION_MAIN_CLASS, javaMainClass);
080        }
081    }
082
083    public static void setupLauncherURIHandlerConf(Configuration launcherConf) {
084        for(Map.Entry<String, String> entry : Services.get().get(URIHandlerService.class).getLauncherConfig()) {
085            launcherConf.set(entry.getKey(), entry.getValue());
086        }
087    }
088
089    public static void setupMainArguments(Configuration launcherConf, String[] args) {
090        launcherConf.setInt(LauncherMapper.CONF_OOZIE_ACTION_MAIN_ARG_COUNT, args.length);
091        for (int i = 0; i < args.length; i++) {
092            launcherConf.set(LauncherMapper.CONF_OOZIE_ACTION_MAIN_ARG_PREFIX + i, args[i]);
093        }
094    }
095
096    public static void setupMaxOutputData(Configuration launcherConf, int maxOutputData) {
097        launcherConf.setInt(LauncherMapper.CONF_OOZIE_ACTION_MAX_OUTPUT_DATA, maxOutputData);
098    }
099
100    /**
101     * Set the maximum value of stats data
102     *
103     * @param launcherConf the oozie launcher configuration
104     * @param maxStatsData the maximum allowed size of stats data
105     */
106    public static void setupMaxExternalStatsSize(Configuration launcherConf, int maxStatsData){
107        launcherConf.setInt(LauncherMapper.CONF_OOZIE_EXTERNAL_STATS_MAX_SIZE, maxStatsData);
108    }
109
110    /**
111     * Set the maximum number of globbed files/dirs
112     *
113     * @param launcherConf the oozie launcher configuration
114     * @param fsGlobMax the maximum number of files/dirs for FS operation
115     */
116    public static void setupMaxFSGlob(Configuration launcherConf, int fsGlobMax){
117        launcherConf.setInt(LauncherMapper.CONF_OOZIE_ACTION_FS_GLOB_MAX, fsGlobMax);
118    }
119
120    public static void setupLauncherInfo(JobConf launcherConf, String jobId, String actionId, Path actionDir,
121            String recoveryId, Configuration actionConf, String prepareXML) throws IOException, HadoopAccessorException {
122
123        launcherConf.setMapperClass(LauncherMapper.class);
124        launcherConf.setSpeculativeExecution(false);
125        launcherConf.setNumMapTasks(1);
126        launcherConf.setNumReduceTasks(0);
127
128        launcherConf.set(LauncherMapper.OOZIE_JOB_ID, jobId);
129        launcherConf.set(LauncherMapper.OOZIE_ACTION_ID, actionId);
130        launcherConf.set(LauncherMapper.OOZIE_ACTION_DIR_PATH, actionDir.toString());
131        launcherConf.set(LauncherMapper.OOZIE_ACTION_RECOVERY_ID, recoveryId);
132        launcherConf.set(LauncherMapper.ACTION_PREPARE_XML, prepareXML);
133
134        actionConf.set(LauncherMapper.OOZIE_JOB_ID, jobId);
135        actionConf.set(LauncherMapper.OOZIE_ACTION_ID, actionId);
136
137        if (Services.get().getConf().getBoolean("oozie.hadoop-2.0.2-alpha.workaround.for.distributed.cache", false)) {
138          List<String> purgedEntries = new ArrayList<String>();
139          Collection<String> entries = actionConf.getStringCollection("mapreduce.job.cache.files");
140          for (String entry : entries) {
141            if (entry.contains("#")) {
142              purgedEntries.add(entry);
143            }
144          }
145          actionConf.setStrings("mapreduce.job.cache.files", purgedEntries.toArray(new String[purgedEntries.size()]));
146          launcherConf.setBoolean("oozie.hadoop-2.0.2-alpha.workaround.for.distributed.cache", true);
147        }
148
149        FileSystem fs =
150          Services.get().get(HadoopAccessorService.class).createFileSystem(launcherConf.get("user.name"),
151                                                                           actionDir.toUri(), launcherConf);
152        fs.mkdirs(actionDir);
153
154        OutputStream os = fs.create(new Path(actionDir, LauncherMapper.ACTION_CONF_XML));
155        try {
156            actionConf.writeXml(os);
157        } finally {
158            IOUtils.closeSafely(os);
159        }
160
161        launcherConf.setInputFormat(OozieLauncherInputFormat.class);
162        launcherConf.set("mapred.output.dir", new Path(actionDir, "output").toString());
163    }
164
165    public static void setupYarnRestartHandling(JobConf launcherJobConf, Configuration actionConf, String launcherTag)
166            throws NoSuchAlgorithmException {
167        launcherJobConf.setLong(LauncherMainHadoopUtils.OOZIE_JOB_LAUNCH_TIME, System.currentTimeMillis());
168        // Tags are limited to 100 chars so we need to hash them to make sure (the actionId otherwise doesn't have a max length)
169        String tag = getTag(launcherTag);
170        // keeping the oozie.child.mapreduce.job.tags instead of mapreduce.job.tags to avoid killing launcher itself.
171        // mapreduce.job.tags should only go to child job launch by launcher.
172        actionConf.set(LauncherMainHadoopUtils.CHILD_MAPREDUCE_JOB_TAGS, tag);
173    }
174
175    private static String getTag(String launcherTag) throws NoSuchAlgorithmException {
176        MessageDigest digest = MessageDigest.getInstance("MD5");
177        digest.update(launcherTag.getBytes(), 0, launcherTag.length());
178        String md5 = "oozie-" + new BigInteger(1, digest.digest()).toString(16);
179        return md5;
180    }
181
182    public static boolean isMainDone(RunningJob runningJob) throws IOException {
183        return runningJob.isComplete();
184    }
185
186    public static boolean isMainSuccessful(RunningJob runningJob) throws IOException {
187        boolean succeeded = runningJob.isSuccessful();
188        if (succeeded) {
189            Counters counters = runningJob.getCounters();
190            if (counters != null) {
191                Counters.Group group = counters.getGroup(LauncherMapper.COUNTER_GROUP);
192                if (group != null) {
193                    succeeded = group.getCounter(LauncherMapper.COUNTER_LAUNCHER_ERROR) == 0;
194                }
195            }
196        }
197        return succeeded;
198    }
199
200    /**
201     * Determine whether action has external child jobs or not
202     * @param actionData
203     * @return true/false
204     * @throws IOException
205     */
206    public static boolean hasExternalChildJobs(Map<String, String> actionData) throws IOException {
207        return actionData.containsKey(LauncherMapper.ACTION_DATA_EXTERNAL_CHILD_IDS);
208    }
209
210    /**
211     * Determine whether action has output data or not
212     * @param actionData
213     * @return true/false
214     * @throws IOException
215     */
216    public static boolean hasOutputData(Map<String, String> actionData) throws IOException {
217        return actionData.containsKey(LauncherMapper.ACTION_DATA_OUTPUT_PROPS);
218    }
219
220    /**
221     * Determine whether action has external stats or not
222     * @param actionData
223     * @return true/false
224     * @throws IOException
225     */
226    public static boolean hasStatsData(Map<String, String> actionData) throws IOException{
227        return actionData.containsKey(LauncherMapper.ACTION_DATA_STATS);
228    }
229
230    /**
231     * Determine whether action has new id (id swap) or not
232     * @param actionData
233     * @return true/false
234     * @throws IOException
235     */
236    public static boolean hasIdSwap(Map<String, String> actionData) throws IOException {
237        return actionData.containsKey(LauncherMapper.ACTION_DATA_NEW_ID);
238    }
239
240    /**
241     * Get the sequence file path storing all action data
242     * @param actionDir
243     * @return
244     */
245    public static Path getActionDataSequenceFilePath(Path actionDir) {
246        return new Path(actionDir, LauncherMapper.ACTION_DATA_SEQUENCE_FILE);
247    }
248
249    /**
250     * Utility function to load the contents of action data sequence file into
251     * memory object
252     *
253     * @param fs Action Filesystem
254     * @param actionDir Path
255     * @param conf Configuration
256     * @return Map action data
257     * @throws IOException
258     * @throws InterruptedException
259     */
260    public static Map<String, String> getActionData(final FileSystem fs, final Path actionDir, final Configuration conf)
261            throws IOException, InterruptedException {
262        UserGroupInformationService ugiService = Services.get().get(UserGroupInformationService.class);
263        UserGroupInformation ugi = ugiService.getProxyUser(conf.get(OozieClient.USER_NAME));
264
265        return ugi.doAs(new PrivilegedExceptionAction<Map<String, String>>() {
266            @Override
267            public Map<String, String> run() throws IOException {
268                Map<String, String> ret = new HashMap<String, String>();
269                Path seqFilePath = getActionDataSequenceFilePath(actionDir);
270                if (fs.exists(seqFilePath)) {
271                    SequenceFile.Reader seqFile = new SequenceFile.Reader(fs, seqFilePath, conf);
272                    Text key = new Text(), value = new Text();
273                    while (seqFile.next(key, value)) {
274                        ret.put(key.toString(), value.toString());
275                    }
276                    seqFile.close();
277                }
278                else { // maintain backward-compatibility. to be deprecated
279                    org.apache.hadoop.fs.FileStatus[] files = fs.listStatus(actionDir);
280                    InputStream is;
281                    BufferedReader reader = null;
282                    Properties props;
283                    if (files != null && files.length > 0) {
284                        for (int x = 0; x < files.length; x++) {
285                            Path file = files[x].getPath();
286                            if (file.equals(new Path(actionDir, "externalChildIds.properties"))) {
287                                is = fs.open(file);
288                                reader = new BufferedReader(new InputStreamReader(is));
289                                ret.put(LauncherMapper.ACTION_DATA_EXTERNAL_CHILD_IDS,
290                                        IOUtils.getReaderAsString(reader, -1));
291                            }
292                            else if (file.equals(new Path(actionDir, "newId.properties"))) {
293                                is = fs.open(file);
294                                reader = new BufferedReader(new InputStreamReader(is));
295                                props = PropertiesUtils.readProperties(reader, -1);
296                                ret.put(LauncherMapper.ACTION_DATA_NEW_ID, props.getProperty("id"));
297                            }
298                            else if (file.equals(new Path(actionDir, LauncherMapper.ACTION_DATA_OUTPUT_PROPS))) {
299                                int maxOutputData = conf.getInt(LauncherMapper.CONF_OOZIE_ACTION_MAX_OUTPUT_DATA,
300                                        2 * 1024);
301                                is = fs.open(file);
302                                reader = new BufferedReader(new InputStreamReader(is));
303                                ret.put(LauncherMapper.ACTION_DATA_OUTPUT_PROPS, PropertiesUtils
304                                        .propertiesToString(PropertiesUtils.readProperties(reader, maxOutputData)));
305                            }
306                            else if (file.equals(new Path(actionDir, LauncherMapper.ACTION_DATA_STATS))) {
307                                int statsMaxOutputData = conf.getInt(LauncherMapper.CONF_OOZIE_EXTERNAL_STATS_MAX_SIZE,
308                                        Integer.MAX_VALUE);
309                                is = fs.open(file);
310                                reader = new BufferedReader(new InputStreamReader(is));
311                                ret.put(LauncherMapper.ACTION_DATA_STATS, PropertiesUtils
312                                        .propertiesToString(PropertiesUtils.readProperties(reader, statsMaxOutputData)));
313                            }
314                            else if (file.equals(new Path(actionDir, LauncherMapper.ACTION_DATA_ERROR_PROPS))) {
315                                is = fs.open(file);
316                                reader = new BufferedReader(new InputStreamReader(is));
317                                ret.put(LauncherMapper.ACTION_DATA_ERROR_PROPS, IOUtils.getReaderAsString(reader, -1));
318                            }
319                        }
320                    }
321                }
322                return ret;
323            }
324        });
325    }
326}