Package org.apache.hadoop.hive.ql.stats

Examples of org.apache.hadoop.hive.ql.stats.StatsPublisher

76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
  public void testStatsPublisherOneStat() throws Throwable {
    try {
      System.out.println("StatsPublisher - one stat published per key - aggregating matching key");

      // instantiate stats publisher
      StatsPublisher statsPublisher = Utilities.getStatsPublisher((JobConf) conf);
      assertNotNull(statsPublisher);
      assertTrue(statsPublisher.init(conf));
      assertTrue(statsPublisher.connect(conf));

      // instantiate stats aggregator
      StatsAggregator statsAggregator = factory.getStatsAggregator();
      assertNotNull(statsAggregator);
      assertTrue(statsAggregator.connect(conf, null));

      // publish stats
      fillStatMap("200", "1000");
      assertTrue(statsPublisher.publishStat("file_00000", stats));
      fillStatMap("400", "3000");
      assertTrue(statsPublisher.publishStat("file_00001", stats));


      // aggregate existing stats
      String rows0 = statsAggregator.aggregateStats("file_00000", StatsSetupConst.ROW_COUNT);
      assertEquals("200", rows0);
      String usize0 = statsAggregator.aggregateStats("file_00000",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("1000", usize0);

      String rows1 = statsAggregator.aggregateStats("file_00001", StatsSetupConst.ROW_COUNT);
      assertEquals("400", rows1);
      String usize1 = statsAggregator.aggregateStats("file_00001",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("3000", usize1);

      // close connections
      assertTrue(statsPublisher.closeConnection());
      assertTrue(statsAggregator.closeConnection());

      System.out
          .println("StatsPublisher - one stat published per key - aggregating matching key - OK");
    } catch (Throwable e) {
View Full Code Here
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
  public void testStatsPublisher() throws Throwable {
    try {
      System.out.println("StatsPublisher - basic functionality");

      // instantiate stats publisher
      StatsPublisher statsPublisher = Utilities.getStatsPublisher(
          (JobConf) conf);
      assertNotNull(statsPublisher);
      assertTrue(statsPublisher.init(conf));
      assertTrue(statsPublisher.connect(conf));

      // instantiate stats aggregator
      StatsAggregator statsAggregator = factory.getStatsAggregator();
      assertNotNull(statsAggregator);
      assertTrue(statsAggregator.connect(conf, null));
      // statsAggregator.cleanUp("file_0000");
      // assertTrue(statsAggregator.connect(conf));

      // publish stats
      fillStatMap("200", "1000");
      assertTrue(statsPublisher.publishStat("file_00000_a", stats));
      fillStatMap("300", "2000");
      assertTrue(statsPublisher.publishStat("file_00000_b", stats));

      fillStatMap("400", "3000");
      assertTrue(statsPublisher.publishStat("file_00001_a", stats));
      fillStatMap("500", "4000");
      assertTrue(statsPublisher.publishStat("file_00001_b", stats));

      // aggregate existing stats
      String rows0 = statsAggregator.aggregateStats("file_00000", StatsSetupConst.ROW_COUNT);
      assertEquals("500", rows0);
      String usize0 = statsAggregator.aggregateStats("file_00000",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("3000", usize0);

      String rows1 = statsAggregator.aggregateStats("file_00001", StatsSetupConst.ROW_COUNT);
      assertEquals("900", rows1);
      String usize1 = statsAggregator.aggregateStats("file_00001",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("7000", usize1);

      // aggregate non-existent stats
      String rowsX = statsAggregator.aggregateStats("file_00002", StatsSetupConst.ROW_COUNT);
      assertEquals("0", rowsX);
      String usizeX = statsAggregator.aggregateStats("file_00002",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("0", usizeX);

      assertTrue(statsAggregator.cleanUp("file_0000"));

      // close connections
      assertTrue(statsPublisher.closeConnection());
      assertTrue(statsAggregator.closeConnection());

      System.out.println("StatsPublisher - basic functionality - OK");
    } catch (Throwable e) {
      e.printStackTrace();
View Full Code Here
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
  public void testStatsPublisherMultipleUpdates() throws Throwable {
    try {
      System.out.println("StatsPublisher - multiple updates");

      // instantiate stats publisher
      StatsPublisher statsPublisher = Utilities.getStatsPublisher((JobConf) conf);
      assertNotNull(statsPublisher);
      assertTrue(statsPublisher.init(conf));
      assertTrue(statsPublisher.connect(conf));

      // instantiate stats aggregator
      StatsAggregator statsAggregator = factory.getStatsAggregator();
      assertNotNull(statsAggregator);
      assertTrue(statsAggregator.connect(conf, null));

      // publish stats
      fillStatMap("200", "1000");
      assertTrue(statsPublisher.publishStat("file_00000_a", stats));
      fillStatMap("300", "2000");
      assertTrue(statsPublisher.publishStat("file_00000_b", stats));

      fillStatMap("400", "3000");
      assertTrue(statsPublisher.publishStat("file_00001_a", stats));
      fillStatMap("500", "4000");
      assertTrue(statsPublisher.publishStat("file_00001_b", stats));

      // update which should not take any effect
      fillStatMap("190", "1000");
      assertTrue(statsPublisher.publishStat("file_00000_a", stats));
      fillStatMap("290", "2000");
      assertTrue(statsPublisher.publishStat("file_00000_b", stats));

      // update that should take effect
      fillStatMap("500", "5000");
      assertTrue(statsPublisher.publishStat("file_00001_a", stats));
      fillStatMap("600", "6000");
      assertTrue(statsPublisher.publishStat("file_00001_b", stats));

      // aggregate existing stats
      String rows0 = statsAggregator.aggregateStats("file_00000", StatsSetupConst.ROW_COUNT);
      assertEquals("500", rows0);
      String usize0 = statsAggregator.aggregateStats("file_00000",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("3000", usize0);

      String rows1 = statsAggregator.aggregateStats("file_00001", StatsSetupConst.ROW_COUNT);
      assertEquals("1100", rows1);
      String usize1 = statsAggregator.aggregateStats("file_00001",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("11000", usize1);

      assertTrue(statsAggregator.cleanUp("file_0000"));

      // close connections
      assertTrue(statsPublisher.closeConnection());
      assertTrue(statsAggregator.closeConnection());

      System.out.println("StatsPublisher - multiple updates - OK");
    } catch (Throwable e) {
      e.printStackTrace();
View Full Code Here
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
    try {
      System.out
          .println("StatsPublisher - (multiple updates + publishing subset of supported statistics)");

      // instantiate stats publisher
      StatsPublisher statsPublisher = Utilities.getStatsPublisher((JobConf) conf);
      assertNotNull(statsPublisher);
      assertTrue(statsPublisher.init(conf));
      assertTrue(statsPublisher.connect(conf));

      // instantiate stats aggregator
      StatsAggregator statsAggregator = factory.getStatsAggregator();
      assertNotNull(statsAggregator);
      assertTrue(statsAggregator.connect(conf, null));

      // publish stats
      fillStatMap("200", "");
      assertTrue(statsPublisher.publishStat("file_00000_a", stats));
      fillStatMap("300", "2000");
      assertTrue(statsPublisher.publishStat("file_00000_b", stats));


      // aggregate existing stats
      String rows0 = statsAggregator.aggregateStats("file_00000", StatsSetupConst.ROW_COUNT);
      assertEquals("500", rows0);
      String usize0 = statsAggregator.aggregateStats("file_00000",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("2000", usize0);

      // update which should not take any effect - plus the map published is a supset of supported
      // stats
      fillStatMap("190", "");
      assertTrue(statsPublisher.publishStat("file_00000_a", stats));
      fillStatMap("290", "");
      assertTrue(statsPublisher.publishStat("file_00000_b", stats));

      // nothing changed
      rows0 = statsAggregator.aggregateStats("file_00000", StatsSetupConst.ROW_COUNT);
      assertEquals("500", rows0);
      usize0 = statsAggregator.aggregateStats("file_00000",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("2000", usize0);

      fillStatMap("500", "");
      assertTrue(statsPublisher.publishStat("file_00000_a", stats));
      fillStatMap("500", "");
      assertTrue(statsPublisher.publishStat("file_00000_b", stats));

      // changed + the rawDataSize size was overwriten !!!
      rows0 = statsAggregator.aggregateStats("file_00000", StatsSetupConst.ROW_COUNT);
      assertEquals("1000", rows0);
      usize0 = statsAggregator.aggregateStats("file_00000",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("0", usize0);

      assertTrue(statsAggregator.cleanUp("file_0000"));

      // close connections
      assertTrue(statsPublisher.closeConnection());
      assertTrue(statsAggregator.closeConnection());

      System.out
          .println("StatsPublisher - (multiple updates + publishing subset of supported statistics) - OK");
    } catch (Throwable e) {
View Full Code Here
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
  public void testStatsAggregatorCleanUp() throws Throwable {
    try {
      System.out.println("StatsAggregator - clean-up");

      // instantiate stats publisher
      StatsPublisher statsPublisher = Utilities.getStatsPublisher((JobConf) conf);
      assertNotNull(statsPublisher);
      assertTrue(statsPublisher.init(conf));
      assertTrue(statsPublisher.connect(conf));

      // instantiate stats aggregator
      StatsAggregator statsAggregator = factory.getStatsAggregator();
      assertNotNull(statsAggregator);
      assertTrue(statsAggregator.connect(conf, null));

      // publish stats
      fillStatMap("200", "1000");
      assertTrue(statsPublisher.publishStat("file_00000_a", stats));
      fillStatMap("300", "2000");
      assertTrue(statsPublisher.publishStat("file_00000_b", stats));

      fillStatMap("400", "3000");
      assertTrue(statsPublisher.publishStat("file_00001_a", stats));
      fillStatMap("500", "4000");
      assertTrue(statsPublisher.publishStat("file_00001_b", stats));

      // cleanUp
      assertTrue(statsAggregator.cleanUp("file_00000"));

      // now clean-up just for one key
      String rows0 = statsAggregator.aggregateStats("file_00000", StatsSetupConst.ROW_COUNT);
      assertEquals("0", rows0);
      String usize0 = statsAggregator.aggregateStats("file_00000",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("0", usize0);

      // this should still be in the table
      String rows1 = statsAggregator.aggregateStats("file_00001", StatsSetupConst.ROW_COUNT);
      assertEquals("900", rows1);
      String usize1 = statsAggregator.aggregateStats("file_00001",
          StatsSetupConst.RAW_DATA_SIZE);
      assertEquals("7000", usize1);

      assertTrue(statsAggregator.cleanUp("file_0000"));

      // close connections
      assertTrue(statsPublisher.closeConnection());
      assertTrue(statsAggregator.closeConnection());

      System.out.println("StatsAggregator - clean-up - OK");
    } catch (Throwable e) {
      e.printStackTrace();
View Full Code Here
266
267
268
269
270
271
272
273
274
275
276
277
    if (factory == null) {
      throw new HiveException(ErrorMsg.STATSPUBLISHER_NOT_OBTAINED.getErrorCodedMsg());
    }
    // initialize stats publishing table for noscan which has only stats task
    // the rest of MR task following stats task initializes it in ExecDriver.java
    StatsPublisher statsPublisher = factory.getStatsPublisher();
    if (!statsPublisher.init(conf)) { // creating stats table if not exists
      throw new HiveException(ErrorMsg.STATSPUBLISHER_INITIALIZATION_ERROR.getErrorCodedMsg());
    }
    Task sourceTask = getWork().getSourceTask();
    if (sourceTask == null) {
      throw new HiveException(ErrorMsg.STATSAGGREGATOR_SOURCETASK_NULL.getErrorCodedMsg());
View Full Code Here
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
  private void publishStats() throws HiveException {
    boolean isStatsReliable = conf.isStatsReliable();

    // Initializing a stats publisher
    StatsPublisher statsPublisher = Utilities.getStatsPublisher(jc);
    if (!statsPublisher.connect(jc)) {
      // just return, stats gathering should not block the main query.
      LOG.info("StatsPublishing error: cannot connect to database.");
      if (isStatsReliable) {
        throw new HiveException(ErrorMsg.STATSPUBLISHER_CONNECTION_ERROR.getErrorCodedMsg());
      }
      return;
    }

    String taskID = Utilities.getTaskIdFromFilename(Utilities.getTaskId(hconf));
    Map<String, String> statsToPublish = new HashMap<String, String>();

    for (String pspecs : stats.keySet()) {
      statsToPublish.clear();
      String prefix = Utilities.join(conf.getStatsAggPrefix(), pspecs);

      int maxKeyLength = conf.getMaxStatsKeyPrefixLength();
      String key = Utilities.getHashedStatsPrefix(prefix, maxKeyLength);
      if (!(statsPublisher instanceof StatsCollectionTaskIndependent)) {
        // stats publisher except counter or fs type needs postfix 'taskID'
        key = Utilities.join(prefix, taskID);
      }
      for(String statType : stats.get(pspecs).getStoredStats()) {
        statsToPublish.put(statType, Long.toString(stats.get(pspecs).getStat(statType)));
      }
      if (!statsPublisher.publishStat(key, statsToPublish)) {
        if (isStatsReliable) {
          throw new HiveException(ErrorMsg.STATSPUBLISHER_PUBLISHING_ERROR.getErrorCodedMsg());
        }
      }
      LOG.info("publishing : " + key + " : " + statsToPublish.toString());
    }
    if (!statsPublisher.closeConnection()) {
      if (isStatsReliable) {
        throw new HiveException(ErrorMsg.STATSPUBLISHER_CLOSING_ERROR.getErrorCodedMsg());
      }
    }
  }
View Full Code Here
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
      throw new HiveException(ErrorMsg.GENERIC_ERROR.getErrorCodedMsg());
    }

    // initialize stats publisher if necessary
    if (work.isGatheringStats()) {
      StatsPublisher statsPublisher;
      StatsFactory factory = StatsFactory.newFactory(conf);
      if (factory != null) {
        statsPublisher = factory.getStatsPublisher();
        if (!statsPublisher.init(conf)) { // creating stats table if not exists
          if (HiveConf.getBoolVar(conf, HiveConf.ConfVars.HIVE_STATS_RELIABLE)) {
            throw
              new HiveException(ErrorMsg.STATSPUBLISHER_INITIALIZATION_ERROR.getErrorCodedMsg());
          }
        }
View Full Code Here
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
      // make this client wait if job trcker is not behaving well.
      Throttle.checkJobTracker(job, LOG);

      if (work.isGatheringStats()) {
        // initialize stats publishing table
        StatsPublisher statsPublisher;
        StatsFactory factory = StatsFactory.newFactory(job);
        if (factory != null) {
          statsPublisher = factory.getStatsPublisher();
          if (!statsPublisher.init(job)) { // creating stats table if not exists
            if (HiveConf.getBoolVar(job, HiveConf.ConfVars.HIVE_STATS_RELIABLE)) {
              throw
                new HiveException(ErrorMsg.STATSPUBLISHER_INITIALIZATION_ERROR.getErrorCodedMsg());
            }
          }
View Full Code Here
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
   *
   * @throws HiveException
   */
  private void publishStats() throws HiveException {
    // Initializing a stats publisher
    StatsPublisher statsPublisher = Utilities.getStatsPublisher(jc);

    if (statsPublisher == null) {
      // just return, stats gathering should not block the main query
      LOG.error("StatsPublishing error: StatsPublisher is not initialized.");
      throw new HiveException(ErrorMsg.STATSPUBLISHER_NOT_OBTAINED.getErrorCodedMsg());
    }

    if (!statsPublisher.connect(jc)) {
      // should fail since stats gathering is main purpose of the job
      LOG.error("StatsPublishing error: cannot connect to database");
      throw new HiveException(ErrorMsg.STATSPUBLISHER_CONNECTION_ERROR.getErrorCodedMsg());
    }

    int maxPrefixLength = StatsFactory.getMaxPrefixLength(jc);
    // construct key used to store stats in intermediate db
    String key = Utilities.getHashedStatsPrefix(statsAggKeyPrefix, maxPrefixLength);
    if (!(statsPublisher instanceof CounterStatsPublisher)) {
      String taskID = Utilities.getTaskIdFromFilename(Utilities.getTaskId(jc));
      key = Utilities.join(key, taskID);
    }

    // construct statistics to be stored
    Map<String, String> statsToPublish = new HashMap<String, String>();
    statsToPublish.put(StatsSetupConst.RAW_DATA_SIZE, Long.toString(uncompressedFileSize));
    statsToPublish.put(StatsSetupConst.ROW_COUNT, Long.toString(rowNo));

    if (!statsPublisher.publishStat(key, statsToPublish)) {
      // The original exception is lost.
      // Not changing the interface to maintain backward compatibility
      throw new HiveException(ErrorMsg.STATSPUBLISHER_PUBLISHING_ERROR.getErrorCodedMsg());
    }

    if (!statsPublisher.closeConnection()) {
      // The original exception is lost.
      // Not changing the interface to maintain backward compatibility
      throw new HiveException(ErrorMsg.STATSPUBLISHER_CLOSING_ERROR.getErrorCodedMsg());
    }
  }
View Full Code Here
TOP

Related Classes of org.apache.hadoop.hive.ql.stats.StatsPublisher

  • org.apache.hadoop.hive.ql.exec.ExecDriver
  • org.apache.hadoop.hive.ql.exec.FileSinkOperator
  • org.apache.hadoop.hive.ql.exec.mr.ExecDriver
  • org.apache.hadoop.hive.ql.exec.StatsTask
  • org.apache.hadoop.hive.ql.exec.TableScanOperator
  • org.apache.hadoop.hive.ql.exec.TestStatsPublisherEnhanced
  • org.apache.hadoop.hive.ql.exec.tez.DagUtils
  • org.apache.hadoop.hive.ql.io.rcfile.stats.PartialScanMapper
  • org.apache.hadoop.hive.ql.io.rcfile.stats.PartialScanTask
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.