KYLIN-2242 write multiple files in FactDistinctColumnsReducer with MultipleOutputs
Project: http://git-wip-us.apache.org/repos/asf/kylin/repo Commit: http://git-wip-us.apache.org/repos/asf/kylin/commit/f6dda4fe Tree: http://git-wip-us.apache.org/repos/asf/kylin/tree/f6dda4fe Diff: http://git-wip-us.apache.org/repos/asf/kylin/diff/f6dda4fe Branch: refs/heads/KYLIN-2361 Commit: f6dda4fecd5a882a17379987445a5aa0a347ff63 Parents: 4c33e38 Author: kangkaisen <kangkai...@live.com> Authored: Sat Dec 17 14:12:48 2016 +0800 Committer: shaofengshi <shaofeng...@apache.org> Committed: Mon Jan 23 16:23:56 2017 +0800 ---------------------------------------------------------------------- .../apache/kylin/common/util/HadoopUtil.java | 16 ++ .../kylin/engine/mr/JobBuilderSupport.java | 2 +- .../kylin/engine/mr/common/BatchConstants.java | 9 +- .../engine/mr/steps/CreateDictionaryJob.java | 43 ++-- .../engine/mr/steps/FactDistinctColumnsJob.java | 32 ++- .../mr/steps/FactDistinctColumnsReducer.java | 240 +++++++------------ .../engine/mr/steps/SaveStatisticsStep.java | 10 +- .../mr/steps/UpdateCubeInfoAfterBuildStep.java | 10 +- 8 files changed, 175 insertions(+), 187 deletions(-) ---------------------------------------------------------------------- http://git-wip-us.apache.org/repos/asf/kylin/blob/f6dda4fe/core-common/src/main/java/org/apache/kylin/common/util/HadoopUtil.java ---------------------------------------------------------------------- diff --git a/core-common/src/main/java/org/apache/kylin/common/util/HadoopUtil.java b/core-common/src/main/java/org/apache/kylin/common/util/HadoopUtil.java index bdc4c3e..b9ffe38 100644 --- a/core-common/src/main/java/org/apache/kylin/common/util/HadoopUtil.java +++ b/core-common/src/main/java/org/apache/kylin/common/util/HadoopUtil.java @@ -26,8 +26,10 @@ import java.net.URISyntaxException; import org.apache.commons.lang.StringUtils; import org.apache.hadoop.conf.Configuration; +import org.apache.hadoop.fs.FileStatus; import org.apache.hadoop.fs.FileSystem; import org.apache.hadoop.fs.Path; +import org.apache.hadoop.fs.PathFilter; import org.apache.hadoop.io.Writable; import org.apache.kylin.common.KylinConfig; import org.slf4j.Logger; @@ -140,4 +142,18 @@ public class HadoopUtil { } } + public static Path getFilterOnlyPath(FileSystem fs, Path baseDir, final String filter) throws IOException { + FileStatus[] fileStatus = fs.listStatus(baseDir, new PathFilter() { + @Override + public boolean accept(Path path) { + return path.getName().startsWith(filter); + } + }); + + if (fileStatus.length == 1) { + return fileStatus[0].getPath(); + } else { + return null; + } + } } http://git-wip-us.apache.org/repos/asf/kylin/blob/f6dda4fe/engine-mr/src/main/java/org/apache/kylin/engine/mr/JobBuilderSupport.java ---------------------------------------------------------------------- diff --git a/engine-mr/src/main/java/org/apache/kylin/engine/mr/JobBuilderSupport.java b/engine-mr/src/main/java/org/apache/kylin/engine/mr/JobBuilderSupport.java index 696b22a..c34a904 100644 --- a/engine-mr/src/main/java/org/apache/kylin/engine/mr/JobBuilderSupport.java +++ b/engine-mr/src/main/java/org/apache/kylin/engine/mr/JobBuilderSupport.java @@ -171,7 +171,7 @@ public class JobBuilderSupport { } public String getStatisticsPath(String jobId) { - return getRealizationRootPath(jobId) + "/statistics"; + return getRealizationRootPath(jobId) + "/fact_distinct_columns/" + BatchConstants.CFG_OUTPUT_STATISTICS; } // ============================================================================ http://git-wip-us.apache.org/repos/asf/kylin/blob/f6dda4fe/engine-mr/src/main/java/org/apache/kylin/engine/mr/common/BatchConstants.java ---------------------------------------------------------------------- diff --git a/engine-mr/src/main/java/org/apache/kylin/engine/mr/common/BatchConstants.java b/engine-mr/src/main/java/org/apache/kylin/engine/mr/common/BatchConstants.java index 0281539..602b4bb 100644 --- a/engine-mr/src/main/java/org/apache/kylin/engine/mr/common/BatchConstants.java +++ b/engine-mr/src/main/java/org/apache/kylin/engine/mr/common/BatchConstants.java @@ -53,9 +53,16 @@ public interface BatchConstants { String CFG_STATISTICS_ENABLED = "statistics.enabled"; String CFG_STATISTICS_OUTPUT = "statistics.ouput";//spell error, for compatibility issue better not change it String CFG_STATISTICS_SAMPLING_PERCENT = "statistics.sampling.percent"; - String CFG_STATISTICS_CUBE_ESTIMATION_FILENAME = "cube_statistics.txt"; String CFG_STATISTICS_CUBOID_ESTIMATION_FILENAME = "cuboid_statistics.seq"; + String CFG_MAPRED_OUTPUT_COMPRESS = "mapred.output.compress"; + + String CFG_OUTPUT_COLUMN = "column"; + String CFG_OUTPUT_DICT = "dict"; + String CFG_OUTPUT_STATISTICS = "statistics"; + String CFG_OUTPUT_PARTITION = "partition"; + + /** * command line ARGuments */ http://git-wip-us.apache.org/repos/asf/kylin/blob/f6dda4fe/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/CreateDictionaryJob.java ---------------------------------------------------------------------- diff --git a/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/CreateDictionaryJob.java b/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/CreateDictionaryJob.java index 95d8cb1..e5d053b 100644 --- a/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/CreateDictionaryJob.java +++ b/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/CreateDictionaryJob.java @@ -18,15 +18,20 @@ package org.apache.kylin.engine.mr.steps; +import java.io.DataInputStream; import java.io.IOException; +import java.nio.ByteBuffer; import org.apache.commons.cli.Options; -import org.apache.commons.io.IOUtils; -import org.apache.hadoop.fs.FSDataInputStream; import org.apache.hadoop.fs.FileSystem; import org.apache.hadoop.fs.Path; +import org.apache.hadoop.io.BytesWritable; +import org.apache.hadoop.io.NullWritable; +import org.apache.hadoop.io.SequenceFile; import org.apache.hadoop.util.ToolRunner; import org.apache.kylin.common.KylinConfig; +import org.apache.kylin.common.util.ByteArray; +import org.apache.kylin.common.util.ByteBufferBackedInputStream; import org.apache.kylin.common.util.ClassUtil; import org.apache.kylin.common.util.Dictionary; import org.apache.kylin.common.util.HadoopUtil; @@ -63,21 +68,27 @@ public class CreateDictionaryJob extends AbstractHadoopJob { @Override public Dictionary<String> getDictionary(TblColRef col) throws IOException { - Path dictFile = new Path(factColumnsInputPath, col.getIdentity() + FactDistinctColumnsReducer.DICT_FILE_POSTFIX); - FileSystem fs = HadoopUtil.getWorkingFileSystem(); - if (fs.exists(dictFile) == false) + Path colDir = new Path(factColumnsInputPath, col.getName()); + FileSystem fs = HadoopUtil.getFileSystem(colDir.toString()); + + Path dictFile = HadoopUtil.getFilterOnlyPath(fs, colDir, col.getName() + FactDistinctColumnsReducer.DICT_FILE_POSTFIX); + if (dictFile == null) { return null; - - FSDataInputStream is = null; - try { - is = fs.open(dictFile); - String dictClassName = is.readUTF(); - Dictionary<String> dict = (Dictionary<String>) ClassUtil.newInstance(dictClassName); - dict.readFields(is); - logger.info("DictionaryProvider read dict from file: " + dictFile); - return dict; - } finally { - IOUtils.closeQuietly(is); + } + + try (SequenceFile.Reader reader = new SequenceFile.Reader(HadoopUtil.getCurrentConfiguration(), SequenceFile.Reader.file(dictFile))) { + NullWritable key = NullWritable.get(); + BytesWritable value = new BytesWritable(); + reader.next(key, value); + + ByteBuffer buffer = new ByteArray(value.getBytes()).asBuffer(); + try (DataInputStream is = new DataInputStream(new ByteBufferBackedInputStream(buffer))) { + String dictClassName = is.readUTF(); + Dictionary<String> dict = (Dictionary<String>) ClassUtil.newInstance(dictClassName); + dict.readFields(is); + logger.info("DictionaryProvider read dict from file: " + dictFile); + return dict; + } } } }); http://git-wip-us.apache.org/repos/asf/kylin/blob/f6dda4fe/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/FactDistinctColumnsJob.java ---------------------------------------------------------------------- diff --git a/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/FactDistinctColumnsJob.java b/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/FactDistinctColumnsJob.java index ce01eb6..aded600 100644 --- a/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/FactDistinctColumnsJob.java +++ b/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/FactDistinctColumnsJob.java @@ -23,11 +23,16 @@ import java.util.List; import org.apache.commons.cli.Options; import org.apache.hadoop.fs.Path; +import org.apache.hadoop.io.BytesWritable; +import org.apache.hadoop.io.LongWritable; import org.apache.hadoop.io.NullWritable; import org.apache.hadoop.io.Text; import org.apache.hadoop.mapreduce.Job; import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat; +import org.apache.hadoop.mapreduce.lib.output.LazyOutputFormat; +import org.apache.hadoop.mapreduce.lib.output.MultipleOutputs; import org.apache.hadoop.mapreduce.lib.output.SequenceFileOutputFormat; +import org.apache.hadoop.mapreduce.lib.output.TextOutputFormat; import org.apache.hadoop.util.ToolRunner; import org.apache.kylin.common.KylinConfig; import org.apache.kylin.cube.CubeInstance; @@ -82,8 +87,8 @@ public class FactDistinctColumnsJob extends AbstractHadoopJob { int uhcReducerCount = cube.getConfig().getUHCReducerCount(); int[] uhcIndex = cubeMgr.getUHCIndex(cube.getDescriptor()); - for(int index : uhcIndex) { - if(index == 1) { + for (int index : uhcIndex) { + if (index == 1) { reducerCount += uhcReducerCount - 1; } } @@ -92,7 +97,6 @@ public class FactDistinctColumnsJob extends AbstractHadoopJob { throw new IllegalArgumentException("The max reducer number for FactDistinctColumnsJob is 255, but now it is " + reducerCount + ", decrease 'kylin.engine.mr.uhc-reducer-count'"); } - job.getConfiguration().set(BatchConstants.CFG_CUBE_NAME, cubeName); job.getConfiguration().set(BatchConstants.CFG_CUBE_SEGMENT_ID, segmentID); job.getConfiguration().set(BatchConstants.CFG_STATISTICS_ENABLED, statistics_enabled); @@ -117,6 +121,12 @@ public class FactDistinctColumnsJob extends AbstractHadoopJob { attachCubeMetadata(cube, job.getConfiguration()); + /** + * don't compress the reducer output so that {@link CreateDictionaryJob} and {@link UpdateCubeInfoAfterBuildStep} + * could read the reducer file directly + */ + job.getConfiguration().set(BatchConstants.CFG_MAPRED_OUTPUT_COMPRESS, "false"); + return waitForCompletion(job); } finally { @@ -138,18 +148,22 @@ public class FactDistinctColumnsJob extends AbstractHadoopJob { private void setupReducer(Path output, int numberOfReducers) throws IOException { job.setReducerClass(FactDistinctColumnsReducer.class); - job.setOutputFormatClass(SequenceFileOutputFormat.class); - job.setOutputKeyClass(NullWritable.class); - job.setOutputValueClass(Text.class); job.setPartitionerClass(FactDistinctColumnPartitioner.class); job.setNumReduceTasks(numberOfReducers); - // important, reducer writes HDFS directly at the moment - job.setReduceSpeculativeExecution(false); - + //make each reducer output to respective dir + MultipleOutputs.addNamedOutput(job, BatchConstants.CFG_OUTPUT_COLUMN, SequenceFileOutputFormat.class, NullWritable.class, Text.class); + MultipleOutputs.addNamedOutput(job, BatchConstants.CFG_OUTPUT_DICT, SequenceFileOutputFormat.class, NullWritable.class, BytesWritable.class); + MultipleOutputs.addNamedOutput(job, BatchConstants.CFG_OUTPUT_STATISTICS, SequenceFileOutputFormat.class, LongWritable.class, BytesWritable.class); + MultipleOutputs.addNamedOutput(job, BatchConstants.CFG_OUTPUT_PARTITION, TextOutputFormat.class, NullWritable.class, LongWritable.class); + + FileOutputFormat.setOutputPath(job, output); job.getConfiguration().set(BatchConstants.CFG_OUTPUT_PATH, output.toString()); + //prevent to create zero-sized default output + LazyOutputFormat.setOutputFormatClass(job, SequenceFileOutputFormat.class); + deletePath(job.getConfiguration(), output); } http://git-wip-us.apache.org/repos/asf/kylin/blob/f6dda4fe/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/FactDistinctColumnsReducer.java ---------------------------------------------------------------------- diff --git a/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/FactDistinctColumnsReducer.java b/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/FactDistinctColumnsReducer.java index 711d991..5d2fb72 100644 --- a/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/FactDistinctColumnsReducer.java +++ b/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/FactDistinctColumnsReducer.java @@ -18,27 +18,25 @@ package org.apache.kylin.engine.mr.steps; +import java.io.DataOutputStream; import java.io.IOException; import java.nio.ByteBuffer; -import java.util.Collection; import java.util.Collections; import java.util.HashMap; import java.util.List; import java.util.Map; -import org.apache.commons.io.IOUtils; +import org.apache.commons.io.output.ByteArrayOutputStream; import org.apache.hadoop.conf.Configuration; -import org.apache.hadoop.fs.FSDataOutputStream; -import org.apache.hadoop.fs.FileSystem; -import org.apache.hadoop.fs.Path; +import org.apache.hadoop.io.BytesWritable; +import org.apache.hadoop.io.LongWritable; import org.apache.hadoop.io.NullWritable; import org.apache.hadoop.io.Text; +import org.apache.hadoop.mapreduce.lib.output.MultipleOutputs; import org.apache.kylin.common.KylinConfig; -import org.apache.kylin.common.util.ByteArray; import org.apache.kylin.common.util.Bytes; import org.apache.kylin.common.util.DateFormat; import org.apache.kylin.common.util.Dictionary; -import org.apache.kylin.common.util.HadoopUtil; import org.apache.kylin.cube.CubeInstance; import org.apache.kylin.cube.CubeManager; import org.apache.kylin.cube.model.CubeDesc; @@ -47,7 +45,7 @@ import org.apache.kylin.dict.IDictionaryBuilder; import org.apache.kylin.engine.mr.KylinReducer; import org.apache.kylin.engine.mr.common.AbstractHadoopJob; import org.apache.kylin.engine.mr.common.BatchConstants; -import org.apache.kylin.engine.mr.common.CubeStatsWriter; +import org.apache.kylin.measure.BufferedMeasureCodec; import org.apache.kylin.measure.hllc.HLLCounter; import org.apache.kylin.metadata.model.TblColRef; import org.slf4j.Logger; @@ -63,14 +61,12 @@ public class FactDistinctColumnsReducer extends KylinReducer<SelfDefineSortableK private static final Logger logger = LoggerFactory.getLogger(FactDistinctColumnsReducer.class); private List<TblColRef> columnList; - private String statisticsOutput = null; private List<Long> baseCuboidRowCountInMappers; protected Map<Long, HLLCounter> cuboidHLLMap = null; protected long baseCuboidId; protected CubeDesc cubeDesc; private long totalRowsBeforeMerge = 0; private int samplingPercentage; - private List<ByteArray> colValues; private TblColRef col = null; private boolean isStatistics = false; private KylinConfig cubeConfig; @@ -88,10 +84,14 @@ public class FactDistinctColumnsReducer extends KylinReducer<SelfDefineSortableK public static final String DICT_FILE_POSTFIX = ".rldict"; public static final String PARTITION_COL_INFO_FILE_POSTFIX = ".pci"; + private MultipleOutputs mos; + @Override protected void setup(Context context) throws IOException { super.bindCurrentConfiguration(context.getConfiguration()); Configuration conf = context.getConfiguration(); + mos = new MultipleOutputs(context); + KylinConfig config = AbstractHadoopJob.loadKylinPropsAndMetadata(); String cubeName = conf.get(BatchConstants.CFG_CUBE_NAME); CubeInstance cube = CubeManager.getInstance(config).getCube(cubeName); @@ -109,26 +109,20 @@ public class FactDistinctColumnsReducer extends KylinReducer<SelfDefineSortableK if (collectStatistics && (taskId == numberOfTasks - 1)) { // hll isStatistics = true; - statisticsOutput = conf.get(BatchConstants.CFG_STATISTICS_OUTPUT); baseCuboidRowCountInMappers = Lists.newArrayList(); cuboidHLLMap = Maps.newHashMap(); samplingPercentage = Integer.parseInt(context.getConfiguration().get(BatchConstants.CFG_STATISTICS_SAMPLING_PERCENT)); logger.info("Reducer " + taskId + " handling stats"); } else if (collectStatistics && (taskId == numberOfTasks - 2)) { // partition col - isStatistics = false; isPartitionCol = true; col = cubeDesc.getModel().getPartitionDesc().getPartitionDateColumnRef(); if (col == null) { logger.info("Do not have partition col. This reducer will keep empty"); } - colValues = Lists.newLinkedList(); - logger.info("Reducer " + taskId + " handling partition column " + col); } else { // normal col - isStatistics = false; col = columnList.get(reducerIdToColumnIndex.get(taskId)); - colValues = Lists.newLinkedList(); // local build dict isReducerLocalBuildDict = config.isReducerLocalBuildDict(); @@ -194,15 +188,13 @@ public class FactDistinctColumnsReducer extends KylinReducer<SelfDefineSortableK logAFewRows(value); builder.addValue(value); } else { - colValues.add(new ByteArray(Bytes.copy(key.getBytes(), 1, key.getLength() - 1))); - if (colValues.size() == 1000000) { //spill every 1 million - logger.info("spill values to disk..."); - outputDistinctValues(col, colValues, context); - colValues.clear(); - } + byte[] keyBytes = Bytes.copy(key.getBytes(), 1, key.getLength() - 1); + // output written to baseDir/colName/-r-00000 (etc) + String fileName = col.getName() + "/"; + mos.write(BatchConstants.CFG_OUTPUT_COLUMN, NullWritable.get(), new Text(keyBytes), fileName); } } - + rowCount++; } @@ -212,162 +204,104 @@ public class FactDistinctColumnsReducer extends KylinReducer<SelfDefineSortableK } } - private void outputDistinctValues(TblColRef col, Collection<ByteArray> values, Context context) throws IOException { - final Configuration conf = context.getConfiguration(); - final FileSystem fs = HadoopUtil.getWorkingFileSystem(conf); - final String outputPath = conf.get(BatchConstants.CFG_OUTPUT_PATH); - final Path colDir = new Path(outputPath, col.getIdentity()); - final String fileName = col.getIdentity() + "-" + taskId % uhcReducerCount; - final Path outputFile = new Path(colDir, fileName); - - FSDataOutputStream out = null; - try { - if (!fs.exists(colDir)) { - fs.mkdirs(colDir); - } - - if (fs.exists(outputFile)) { - out = fs.append(outputFile); - logger.info("append file " + outputFile); - } else { - out = fs.create(outputFile); - logger.info("create file " + outputFile); - } - - for (ByteArray value : values) { - out.write(value.array(), value.offset(), value.length()); - out.write('\n'); - } - } finally { - IOUtils.closeQuietly(out); - } - } - - private void outputDict(TblColRef col, Dictionary<String> dict, Context context) throws IOException { - final String fileName = col.getIdentity() + DICT_FILE_POSTFIX; - FSDataOutputStream out = getOutputStream(context, fileName); - try { - String dictClassName = dict.getClass().getName(); - out.writeUTF(dictClassName); - dict.write(out); - logger.info("reducer id is:+" + taskId + " colName:" + col.getName() + " writing dict at file : " + fileName + " dict class:" + dictClassName); - } finally { - IOUtils.closeQuietly(out); - } - } - - private void outputPartitionInfo(Context context) throws IOException { - final String fileName = col.getIdentity() + PARTITION_COL_INFO_FILE_POSTFIX; - FSDataOutputStream out = getOutputStream(context, fileName); - try { - out.writeLong(timeMinValue); - out.writeLong(timeMaxValue); - logger.info("write partition info for col : " + col.getName() + " minValue:" + timeMinValue + " maxValue:" + timeMaxValue); - } finally { - IOUtils.closeQuietly(out); - } - } - - private FSDataOutputStream getOutputStream(Context context, String outputFileName) throws IOException { - final Configuration conf = context.getConfiguration(); - final FileSystem fs = HadoopUtil.getWorkingFileSystem(conf); - final Path outputPath = new Path(conf.get(BatchConstants.CFG_OUTPUT_PATH)); - final Path outputFile = new Path(outputPath, outputFileName); - if (!fs.exists(outputPath)) { - fs.mkdirs(outputPath); - } - FSDataOutputStream out = fs.create(outputFile); - return out; - } - @Override protected void doCleanup(Context context) throws IOException, InterruptedException { if (isStatistics) { - // output the hll info - long grandTotal = 0; - for (HLLCounter hll : cuboidHLLMap.values()) { - grandTotal += hll.getCountEstimate(); - } - double mapperOverlapRatio = grandTotal == 0 ? 0 : (double) totalRowsBeforeMerge / grandTotal; - - int mapperNumber = baseCuboidRowCountInMappers.size(); + //output the hll info; + List<Long> allCuboids = Lists.newArrayList(); + allCuboids.addAll(cuboidHLLMap.keySet()); + Collections.sort(allCuboids); - writeMapperAndCuboidStatistics(context); // for human check - CubeStatsWriter.writeCuboidStatistics(context.getConfiguration(), new Path(statisticsOutput), // - cuboidHLLMap, samplingPercentage, mapperNumber, mapperOverlapRatio); + logMapperAndCuboidStatistics(allCuboids); // for human check + outputStatistics(allCuboids); } else if (isPartitionCol) { // partition col - if (col != null) { - outputPartitionInfo(context); - } + outputPartitionInfo(); } else { // normal col if (isReducerLocalBuildDict) { Dictionary<String> dict = builder.build(); - outputDict(col, dict, context); - } else { - if (colValues.size() > 0) { - outputDistinctValues(col, colValues, context); - colValues.clear(); - } + outputDict(col, dict); } } + + mos.close(); } - private void writeMapperAndCuboidStatistics(Context context) throws IOException { - Configuration conf = context.getConfiguration(); - FileSystem fs = HadoopUtil.getWorkingFileSystem(conf); - Path path = new Path(statisticsOutput, BatchConstants.CFG_STATISTICS_CUBE_ESTIMATION_FILENAME); - FSDataOutputStream out = fs.create(path); + private void outputPartitionInfo() throws IOException, InterruptedException { + if (col != null) { + // output written to baseDir/colName/colName.pci-r-00000 (etc) + String partitionFileName = col.getName() + "/" + col.getName() + PARTITION_COL_INFO_FILE_POSTFIX; - try { - String msg; + mos.write(BatchConstants.CFG_OUTPUT_PARTITION, NullWritable.get(), new LongWritable(timeMinValue), partitionFileName); + mos.write(BatchConstants.CFG_OUTPUT_PARTITION, NullWritable.get(), new LongWritable(timeMaxValue), partitionFileName); + logger.info("write partition info for col : " + col.getName() + " minValue:" + timeMinValue + " maxValue:" + timeMaxValue); + } + } - List<Long> allCuboids = Lists.newArrayList(); - allCuboids.addAll(cuboidHLLMap.keySet()); - Collections.sort(allCuboids); + private void outputDict(TblColRef col, Dictionary<String> dict) throws IOException, InterruptedException { + // output written to baseDir/colName/colName.rldict-r-00000 (etc) + String dictFileName = col.getName() + "/" + col.getName() + DICT_FILE_POSTFIX; - msg = "Total cuboid number: \t" + allCuboids.size(); - writeLine(out, msg); - msg = "Samping percentage: \t" + samplingPercentage; - writeLine(out, msg); - - writeLine(out, "The following statistics are collected based on sampling data."); - writeLine(out, "Number of Mappers: " + baseCuboidRowCountInMappers.size()); - for (int i = 0; i < baseCuboidRowCountInMappers.size(); i++) { - if (baseCuboidRowCountInMappers.get(i) > 0) { - msg = "Base Cuboid in Mapper " + i + " row count: \t " + baseCuboidRowCountInMappers.get(i); - writeLine(out, msg); - } - } + try (ByteArrayOutputStream baos = new ByteArrayOutputStream(); DataOutputStream outputStream = new DataOutputStream(baos);) { + outputStream.writeUTF(dict.getClass().getName()); + dict.write(outputStream); - long grantTotal = 0; - for (long i : allCuboids) { - grantTotal += cuboidHLLMap.get(i).getCountEstimate(); - msg = "Cuboid " + i + " row count is: \t " + cuboidHLLMap.get(i).getCountEstimate(); - writeLine(out, msg); - } + mos.write(BatchConstants.CFG_OUTPUT_DICT, NullWritable.get(), new BytesWritable(baos.toByteArray()), dictFileName); + } + } - msg = "Sum of all the cube segments (before merge) is: \t " + totalRowsBeforeMerge; - writeLine(out, msg); + private void outputStatistics(List<Long> allCuboids) throws IOException, InterruptedException { + // output written to baseDir/statistics/statistics-r-00000 (etc) + String statisticsFileName = BatchConstants.CFG_OUTPUT_STATISTICS + "/" + BatchConstants.CFG_OUTPUT_STATISTICS; - msg = "After merge, the cube has row count: \t " + grantTotal; - writeLine(out, msg); + ByteBuffer valueBuf = ByteBuffer.allocate(BufferedMeasureCodec.DEFAULT_BUFFER_SIZE); - if (grantTotal > 0) { - msg = "The mapper overlap ratio is: \t" + totalRowsBeforeMerge / grantTotal; - writeLine(out, msg); - } + // mapper overlap ratio at key -1 + long grandTotal = 0; + for (HLLCounter hll : cuboidHLLMap.values()) { + grandTotal += hll.getCountEstimate(); + } + double mapperOverlapRatio = grandTotal == 0 ? 0 : (double) totalRowsBeforeMerge / grandTotal; + mos.write(BatchConstants.CFG_OUTPUT_STATISTICS, new LongWritable(-1), new BytesWritable(Bytes.toBytes(mapperOverlapRatio)), statisticsFileName); + + // mapper number at key -2 + mos.write(BatchConstants.CFG_OUTPUT_STATISTICS, new LongWritable(-2), new BytesWritable(Bytes.toBytes(baseCuboidRowCountInMappers.size())), statisticsFileName); + + // sampling percentage at key 0 + mos.write(BatchConstants.CFG_OUTPUT_STATISTICS, new LongWritable(0L), new BytesWritable(Bytes.toBytes(samplingPercentage)), statisticsFileName); - } finally { - IOUtils.closeQuietly(out); + for (long i : allCuboids) { + valueBuf.clear(); + cuboidHLLMap.get(i).writeRegisters(valueBuf); + valueBuf.flip(); + mos.write(BatchConstants.CFG_OUTPUT_STATISTICS, new LongWritable(i), new BytesWritable(valueBuf.array(), valueBuf.limit()), statisticsFileName); } } - private void writeLine(FSDataOutputStream out, String msg) throws IOException { - out.write(msg.getBytes()); - out.write('\n'); + private void logMapperAndCuboidStatistics(List<Long> allCuboids) throws IOException { + logger.info("Total cuboid number: \t" + allCuboids.size()); + logger.info("Samping percentage: \t" + samplingPercentage); + logger.info("The following statistics are collected based on sampling data."); + logger.info("Number of Mappers: " + baseCuboidRowCountInMappers.size()); + for (int i = 0; i < baseCuboidRowCountInMappers.size(); i++) { + if (baseCuboidRowCountInMappers.get(i) > 0) { + logger.info("Base Cuboid in Mapper " + i + " row count: \t " + baseCuboidRowCountInMappers.get(i)); + } + } + + long grantTotal = 0; + for (long i : allCuboids) { + grantTotal += cuboidHLLMap.get(i).getCountEstimate(); + logger.info("Cuboid " + i + " row count is: \t " + cuboidHLLMap.get(i).getCountEstimate()); + } + + logger.info("Sum of all the cube segments (before merge) is: \t " + totalRowsBeforeMerge); + logger.info("After merge, the cube has row count: \t " + grantTotal); + if (grantTotal > 0) { + logger.info("The mapper overlap ratio is: \t" + totalRowsBeforeMerge / grantTotal); + } } } http://git-wip-us.apache.org/repos/asf/kylin/blob/f6dda4fe/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/SaveStatisticsStep.java ---------------------------------------------------------------------- diff --git a/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/SaveStatisticsStep.java b/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/SaveStatisticsStep.java index 2671042..28f99fb 100644 --- a/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/SaveStatisticsStep.java +++ b/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/SaveStatisticsStep.java @@ -60,9 +60,11 @@ public class SaveStatisticsStep extends AbstractExecutable { ResourceStore rs = ResourceStore.getStore(kylinConf); try { FileSystem fs = HadoopUtil.getWorkingFileSystem(); - Path statisticsFilePath = new Path(CubingExecutableUtil.getStatisticsPath(this.getParams()), BatchConstants.CFG_STATISTICS_CUBOID_ESTIMATION_FILENAME); - if (!fs.exists(statisticsFilePath)) - throw new IOException("File " + statisticsFilePath + " does not exists"); + Path statisticsDir = new Path(CubingExecutableUtil.getStatisticsPath(this.getParams())); + Path statisticsFilePath = HadoopUtil.getFilterOnlyPath(fs, statisticsDir, BatchConstants.CFG_OUTPUT_STATISTICS); + if (statisticsFilePath == null) { + throw new IOException("fail to find the statistics file in base dir: " + statisticsDir); + } FSDataInputStream is = fs.open(statisticsFilePath); try { @@ -110,7 +112,7 @@ public class SaveStatisticsStep extends AbstractExecutable { double overlapThreshold = kylinConf.getCubeAlgorithmAutoThreshold(); logger.info("mapperNumber for " + seg + " is " + mapperNumber + " and threshold is " + mapperNumLimit); logger.info("mapperOverlapRatio for " + seg + " is " + mapperOverlapRatio + " and threshold is " + overlapThreshold); - + // in-mem cubing is good when // 1) the cluster has enough mapper slots to run in parallel // 2) the mapper overlap ratio is small, meaning the shuffle of in-mem MR has advantage http://git-wip-us.apache.org/repos/asf/kylin/blob/f6dda4fe/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/UpdateCubeInfoAfterBuildStep.java ---------------------------------------------------------------------- diff --git a/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/UpdateCubeInfoAfterBuildStep.java b/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/UpdateCubeInfoAfterBuildStep.java index dc80399..81d5c42 100644 --- a/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/UpdateCubeInfoAfterBuildStep.java +++ b/engine-mr/src/main/java/org/apache/kylin/engine/mr/steps/UpdateCubeInfoAfterBuildStep.java @@ -80,8 +80,13 @@ public class UpdateCubeInfoAfterBuildStep extends AbstractExecutable { private void updateTimeRange(CubeSegment segment) throws IOException { final TblColRef partitionCol = segment.getCubeDesc().getModel().getPartitionDesc().getPartitionDateColumnRef(); final String factColumnsInputPath = this.getParams().get(BatchConstants.CFG_OUTPUT_PATH); - Path outputFile = new Path(factColumnsInputPath, partitionCol.getIdentity() + FactDistinctColumnsReducer.PARTITION_COL_INFO_FILE_POSTFIX); - FileSystem fs = HadoopUtil.getFileSystem(outputFile.toString()); + Path colDir = new Path(factColumnsInputPath, partitionCol.getName()); + FileSystem fs = HadoopUtil.getFileSystem(colDir.toString()); + Path outputFile = HadoopUtil.getFilterOnlyPath(fs, colDir, partitionCol.getName() + FactDistinctColumnsReducer.PARTITION_COL_INFO_FILE_POSTFIX); + if (outputFile == null) { + throw new IOException("fail to find the partition file in base dir: " + colDir); + } + FSDataInputStream is = null; long minValue = Long.MAX_VALUE, maxValue = Long.MIN_VALUE; try { @@ -97,5 +102,4 @@ public class UpdateCubeInfoAfterBuildStep extends AbstractExecutable { segment.setDateRangeStart(minValue); segment.setDateRangeEnd(maxValue); } - }
