Java Code Examples for org.apache.kylin.cube.CubeInstance#getOriginalSegmentToOptimize()

The following examples show how to use org.apache.kylin.cube.CubeInstance#getOriginalSegmentToOptimize() . You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may check out the related API usage on the sidebar.

Example 1

Source File: FilterRecommendCuboidDataMapper.java From kylin-on-parquet-v2 with Apache License 2.0

6 votes

@Override
protected void doSetup(Context context) throws IOException {
    super.bindCurrentConfiguration(context.getConfiguration());

    String cubeName = context.getConfiguration().get(BatchConstants.CFG_CUBE_NAME);
    String segmentID = context.getConfiguration().get(BatchConstants.CFG_CUBE_SEGMENT_ID);

    KylinConfig config = AbstractHadoopJob.loadKylinPropsAndMetadata();

    CubeManager cubeManager = CubeManager.getInstance(config);
    CubeInstance cube = cubeManager.getCube(cubeName);
    CubeSegment optSegment = cube.getSegmentById(segmentID);
    CubeSegment originalSegment = cube.getOriginalSegmentToOptimize(optSegment);

    enableSharding = originalSegment.isEnableSharding();
    baseCuboid = cube.getCuboidScheduler().getBaseCuboidId();

    recommendCuboids = cube.getCuboidsRecommend();
    Preconditions.checkNotNull(recommendCuboids, "The recommend cuboid map could not be null");
}

Example 2

Source File: ConvergeCuboidDataReducer.java From kylin-on-parquet-v2 with Apache License 2.0

6 votes

@Override
protected void doSetup(Context context) throws IOException {
    super.bindCurrentConfiguration(context.getConfiguration());
    mos = new MultipleOutputs(context);

    String cubeName = context.getConfiguration().get(BatchConstants.CFG_CUBE_NAME);
    String segmentID = context.getConfiguration().get(BatchConstants.CFG_CUBE_SEGMENT_ID);

    KylinConfig config = AbstractHadoopJob.loadKylinPropsAndMetadata();

    CubeInstance cube = CubeManager.getInstance(config).getCube(cubeName);
    CubeSegment cubeSegment = cube.getSegmentById(segmentID);
    CubeSegment oldSegment = cube.getOriginalSegmentToOptimize(cubeSegment);

    this.enableSharding = oldSegment.isEnableSharding();
    this.baseCuboid = cube.getCuboidScheduler().getBaseCuboidId();
}

Example 3

Source File: UpdateOldCuboidShardMapper.java From kylin-on-parquet-v2 with Apache License 2.0

6 votes

@Override
protected void doSetup(Context context) throws IOException {
    super.bindCurrentConfiguration(context.getConfiguration());

    String cubeName = context.getConfiguration().get(BatchConstants.CFG_CUBE_NAME);
    String segmentID = context.getConfiguration().get(BatchConstants.CFG_CUBE_SEGMENT_ID);

    KylinConfig config = AbstractHadoopJob.loadKylinPropsAndMetadata();

    CubeInstance cube = CubeManager.getInstance(config).getCube(cubeName);
    CubeSegment cubeSegment = cube.getSegmentById(segmentID);
    CubeSegment oldSegment = cube.getOriginalSegmentToOptimize(cubeSegment);

    cubeDesc = cube.getDescriptor();

    rowKeySplitter = new RowKeySplitter(oldSegment);
    rowKeyEncoderProvider = new RowKeyEncoderProvider(cubeSegment);
}

Example 4

Source File: FilterRecommendCuboidDataMapper.java From kylin with Apache License 2.0

6 votes

@Override
protected void doSetup(Context context) throws IOException {
    super.bindCurrentConfiguration(context.getConfiguration());

    String cubeName = context.getConfiguration().get(BatchConstants.CFG_CUBE_NAME);
    String segmentID = context.getConfiguration().get(BatchConstants.CFG_CUBE_SEGMENT_ID);

    KylinConfig config = AbstractHadoopJob.loadKylinPropsAndMetadata();

    CubeManager cubeManager = CubeManager.getInstance(config);
    CubeInstance cube = cubeManager.getCube(cubeName);
    CubeSegment optSegment = cube.getSegmentById(segmentID);
    CubeSegment originalSegment = cube.getOriginalSegmentToOptimize(optSegment);

    enableSharding = originalSegment.isEnableSharding();
    baseCuboid = cube.getCuboidScheduler().getBaseCuboidId();

    recommendCuboids = cube.getCuboidsRecommend();
    Preconditions.checkNotNull(recommendCuboids, "The recommend cuboid map could not be null");
}

Example 5

Source File: ConvergeCuboidDataReducer.java From kylin with Apache License 2.0

6 votes

@Override
protected void doSetup(Context context) throws IOException {
    super.bindCurrentConfiguration(context.getConfiguration());
    mos = new MultipleOutputs(context);

    String cubeName = context.getConfiguration().get(BatchConstants.CFG_CUBE_NAME);
    String segmentID = context.getConfiguration().get(BatchConstants.CFG_CUBE_SEGMENT_ID);

    KylinConfig config = AbstractHadoopJob.loadKylinPropsAndMetadata();

    CubeInstance cube = CubeManager.getInstance(config).getCube(cubeName);
    CubeSegment cubeSegment = cube.getSegmentById(segmentID);
    CubeSegment oldSegment = cube.getOriginalSegmentToOptimize(cubeSegment);

    this.enableSharding = oldSegment.isEnableSharding();
    this.baseCuboid = cube.getCuboidScheduler().getBaseCuboidId();
}

Example 6

Source File: UpdateOldCuboidShardMapper.java From kylin with Apache License 2.0

6 votes

@Override
protected void doSetup(Context context) throws IOException {
    super.bindCurrentConfiguration(context.getConfiguration());

    String cubeName = context.getConfiguration().get(BatchConstants.CFG_CUBE_NAME);
    String segmentID = context.getConfiguration().get(BatchConstants.CFG_CUBE_SEGMENT_ID);

    KylinConfig config = AbstractHadoopJob.loadKylinPropsAndMetadata();

    CubeInstance cube = CubeManager.getInstance(config).getCube(cubeName);
    CubeSegment cubeSegment = cube.getSegmentById(segmentID);
    CubeSegment oldSegment = cube.getOriginalSegmentToOptimize(cubeSegment);

    cubeDesc = cube.getDescriptor();

    rowKeySplitter = new RowKeySplitter(oldSegment);
    rowKeyEncoderProvider = new RowKeyEncoderProvider(cubeSegment);
}

Example 7

Source File: UpdateCubeInfoAfterOptimizeStep.java From kylin-on-parquet-v2 with Apache License 2.0

5 votes

@Override
protected ExecuteResult doWork(ExecutableContext context) throws ExecuteException {
    final CubeManager cubeManager = CubeManager.getInstance(context.getConfig());
    final CubeInstance cube = cubeManager.getCube(CubingExecutableUtil.getCubeName(this.getParams()));
    final CubeSegment segment = cube.getSegmentById(CubingExecutableUtil.getSegmentId(this.getParams()));

    CubeSegment originalSegment = cube.getOriginalSegmentToOptimize(segment);
    long sourceCount = originalSegment.getInputRecords();
    long sourceSizeBytes = originalSegment.getInputRecordsSize();

    CubingJob cubingJob = (CubingJob) getManager().getJob(CubingExecutableUtil.getCubingJobId(this.getParams()));
    long cubeSizeBytes = cubingJob.findCubeSizeBytes();

    segment.setLastBuildJobID(CubingExecutableUtil.getCubingJobId(this.getParams()));
    segment.setLastBuildTime(System.currentTimeMillis());
    segment.setSizeKB(cubeSizeBytes / 1024);
    segment.setInputRecords(sourceCount);
    segment.setInputRecordsSize(sourceSizeBytes);
    segment.setDimensionRangeInfoMap(originalSegment.getDimensionRangeInfoMap());

    try {
        cubeManager.promoteNewlyOptimizeSegments(cube, segment);
        return new ExecuteResult();
    } catch (IOException e) {
        logger.error("fail to update cube after build", e);
        return ExecuteResult.createError(e);
    }
}

Example 8

Source File: UpdateCubeInfoAfterOptimizeStep.java From kylin with Apache License 2.0

5 votes

@Override
protected ExecuteResult doWork(ExecutableContext context) throws ExecuteException {
    final CubeManager cubeManager = CubeManager.getInstance(context.getConfig());
    final CubeInstance cube = cubeManager.getCube(CubingExecutableUtil.getCubeName(this.getParams()));
    final CubeSegment segment = cube.getSegmentById(CubingExecutableUtil.getSegmentId(this.getParams()));

    CubeSegment originalSegment = cube.getOriginalSegmentToOptimize(segment);
    long sourceCount = originalSegment.getInputRecords();
    long sourceSizeBytes = originalSegment.getInputRecordsSize();

    CubingJob cubingJob = (CubingJob) getManager().getJob(CubingExecutableUtil.getCubingJobId(this.getParams()));
    long cubeSizeBytes = cubingJob.findCubeSizeBytes();

    segment.setLastBuildJobID(CubingExecutableUtil.getCubingJobId(this.getParams()));
    segment.setLastBuildTime(System.currentTimeMillis());
    segment.setSizeKB(cubeSizeBytes / 1024);
    segment.setInputRecords(sourceCount);
    segment.setInputRecordsSize(sourceSizeBytes);
    segment.setDimensionRangeInfoMap(originalSegment.getDimensionRangeInfoMap());

    try {
        cubeManager.promoteNewlyOptimizeSegments(cube, segment);
        return new ExecuteResult();
    } catch (IOException e) {
        logger.error("fail to update cube after build", e);
        return ExecuteResult.createError(e);
    }
}

Example 9

Source File: FilterRecommendCuboidDataJob.java From kylin-on-parquet-v2 with Apache License 2.0

4 votes

@Override
public int run(String[] args) throws Exception {
    Options options = new Options();
    try {
        options.addOption(OPTION_JOB_NAME);
        options.addOption(OPTION_CUBE_NAME);
        options.addOption(OPTION_SEGMENT_ID);
        options.addOption(OPTION_INPUT_PATH);
        options.addOption(OPTION_OUTPUT_PATH);
        parseOptions(options, args);

        job = Job.getInstance(getConf(), getOptionValue(OPTION_JOB_NAME));
        String cubeName = getOptionValue(OPTION_CUBE_NAME).toUpperCase(Locale.ROOT);
        String segmentID = getOptionValue(OPTION_SEGMENT_ID);
        Path input = new Path(getOptionValue(OPTION_INPUT_PATH));
        Path output = new Path(getOptionValue(OPTION_OUTPUT_PATH));

        CubeManager cubeMgr = CubeManager.getInstance(KylinConfig.getInstanceFromEnv());
        CubeInstance cube = cubeMgr.getCube(cubeName);
        CubeSegment optSegment = cube.getSegmentById(segmentID);
        CubeSegment originalSegment = cube.getOriginalSegmentToOptimize(optSegment);

        logger.info("Starting: " + job.getJobName());

        setJobClasspath(job, cube.getConfig());

        // Mapper
        job.setMapperClass(FilterRecommendCuboidDataMapper.class);
        job.setMapOutputKeyClass(Text.class);
        job.setMapOutputValueClass(Text.class);

        // Input
        job.setInputFormatClass(SequenceFileInputFormat.class);
        FileInputFormat.setInputPaths(job, input);

        // Reducer
        ConvergeCuboidDataUtil.setupReducer(job, originalSegment, output);

        // set job configuration
        job.getConfiguration().set(BatchConstants.CFG_CUBE_NAME, cubeName);
        job.getConfiguration().set(BatchConstants.CFG_CUBE_SEGMENT_ID, segmentID);
        // add metadata to distributed cache
        attachSegmentMetadata(originalSegment, job.getConfiguration(), false, false);

        this.deletePath(job.getConfiguration(), output);

        return waitForCompletion(job);
    } catch (Exception e) {
        logger.error("error in CuboidJob", e);
        printUsage(options);
        throw e;
    } finally {
        if (job != null)
            cleanupTempConfFile(job.getConfiguration());
    }
}

Example 10

Source File: UpdateOldCuboidShardJob.java From kylin-on-parquet-v2 with Apache License 2.0

4 votes

@Override
public int run(String[] args) throws Exception {
    Options options = new Options();
    try {
        options.addOption(OPTION_JOB_NAME);
        options.addOption(OPTION_CUBE_NAME);
        options.addOption(OPTION_SEGMENT_ID);
        options.addOption(OPTION_INPUT_PATH);
        options.addOption(OPTION_OUTPUT_PATH);
        parseOptions(options, args);

        job = Job.getInstance(getConf(), getOptionValue(OPTION_JOB_NAME));
        String cubeName = getOptionValue(OPTION_CUBE_NAME).toUpperCase(Locale.ROOT);
        String segmentID = getOptionValue(OPTION_SEGMENT_ID);
        Path input = new Path(getOptionValue(OPTION_INPUT_PATH));
        Path output = new Path(getOptionValue(OPTION_OUTPUT_PATH));

        CubeManager cubeMgr = CubeManager.getInstance(KylinConfig.getInstanceFromEnv());
        CubeInstance cube = cubeMgr.getCube(cubeName);
        CubeSegment optSegment = cube.getSegmentById(segmentID);
        CubeSegment originalSegment = cube.getOriginalSegmentToOptimize(optSegment);

        logger.info("Starting: " + job.getJobName());

        setJobClasspath(job, cube.getConfig());

        // Mapper
        job.setMapperClass(UpdateOldCuboidShardMapper.class);
        job.setMapOutputKeyClass(Text.class);
        job.setMapOutputValueClass(Text.class);

        // Input
        job.setInputFormatClass(SequenceFileInputFormat.class);
        FileInputFormat.setInputPaths(job, input);

        // Reducer
        ConvergeCuboidDataUtil.setupReducer(job, originalSegment, output);

        // set job configuration
        job.getConfiguration().set(BatchConstants.CFG_CUBE_NAME, cubeName);
        job.getConfiguration().set(BatchConstants.CFG_CUBE_SEGMENT_ID, segmentID);
        // add metadata to distributed cache
        attachSegmentsMetadataWithDict(Lists.newArrayList(optSegment, originalSegment), job.getConfiguration());

        this.deletePath(job.getConfiguration(), output);

        return waitForCompletion(job);
    } catch (Exception e) {
        logger.error("error in CuboidJob", e);
        printUsage(options);
        throw e;
    } finally {
        if (job != null)
            cleanupTempConfFile(job.getConfiguration());
    }
}

Example 11

Source File: FilterRecommendCuboidDataJob.java From kylin with Apache License 2.0

4 votes

@Override
public int run(String[] args) throws Exception {
    Options options = new Options();
    try {
        options.addOption(OPTION_JOB_NAME);
        options.addOption(OPTION_CUBE_NAME);
        options.addOption(OPTION_SEGMENT_ID);
        options.addOption(OPTION_INPUT_PATH);
        options.addOption(OPTION_OUTPUT_PATH);
        parseOptions(options, args);

        job = Job.getInstance(getConf(), getOptionValue(OPTION_JOB_NAME));
        String cubeName = getOptionValue(OPTION_CUBE_NAME).toUpperCase(Locale.ROOT);
        String segmentID = getOptionValue(OPTION_SEGMENT_ID);
        Path input = new Path(getOptionValue(OPTION_INPUT_PATH));
        Path output = new Path(getOptionValue(OPTION_OUTPUT_PATH));

        CubeManager cubeMgr = CubeManager.getInstance(KylinConfig.getInstanceFromEnv());
        CubeInstance cube = cubeMgr.getCube(cubeName);
        CubeSegment optSegment = cube.getSegmentById(segmentID);
        CubeSegment originalSegment = cube.getOriginalSegmentToOptimize(optSegment);

        logger.info("Starting: " + job.getJobName());

        setJobClasspath(job, cube.getConfig());

        // Mapper
        job.setMapperClass(FilterRecommendCuboidDataMapper.class);
        job.setMapOutputKeyClass(Text.class);
        job.setMapOutputValueClass(Text.class);

        // Input
        job.setInputFormatClass(SequenceFileInputFormat.class);
        FileInputFormat.setInputPaths(job, input);

        // Reducer
        ConvergeCuboidDataUtil.setupReducer(job, originalSegment, output);

        // set job configuration
        job.getConfiguration().set(BatchConstants.CFG_CUBE_NAME, cubeName);
        job.getConfiguration().set(BatchConstants.CFG_CUBE_SEGMENT_ID, segmentID);
        // add metadata to distributed cache
        attachSegmentMetadata(originalSegment, job.getConfiguration(), false, false);

        this.deletePath(job.getConfiguration(), output);

        return waitForCompletion(job);
    } catch (Exception e) {
        logger.error("error in CuboidJob", e);
        printUsage(options);
        throw e;
    } finally {
        if (job != null)
            cleanupTempConfFile(job.getConfiguration());
    }
}

Example 12

Source File: UpdateOldCuboidShardJob.java From kylin with Apache License 2.0

4 votes

@Override
public int run(String[] args) throws Exception {
    Options options = new Options();
    try {
        options.addOption(OPTION_JOB_NAME);
        options.addOption(OPTION_CUBE_NAME);
        options.addOption(OPTION_SEGMENT_ID);
        options.addOption(OPTION_INPUT_PATH);
        options.addOption(OPTION_OUTPUT_PATH);
        parseOptions(options, args);

        job = Job.getInstance(getConf(), getOptionValue(OPTION_JOB_NAME));
        String cubeName = getOptionValue(OPTION_CUBE_NAME).toUpperCase(Locale.ROOT);
        String segmentID = getOptionValue(OPTION_SEGMENT_ID);
        Path input = new Path(getOptionValue(OPTION_INPUT_PATH));
        Path output = new Path(getOptionValue(OPTION_OUTPUT_PATH));

        CubeManager cubeMgr = CubeManager.getInstance(KylinConfig.getInstanceFromEnv());
        CubeInstance cube = cubeMgr.getCube(cubeName);
        CubeSegment optSegment = cube.getSegmentById(segmentID);
        CubeSegment originalSegment = cube.getOriginalSegmentToOptimize(optSegment);

        logger.info("Starting: " + job.getJobName());

        setJobClasspath(job, cube.getConfig());

        // Mapper
        job.setMapperClass(UpdateOldCuboidShardMapper.class);
        job.setMapOutputKeyClass(Text.class);
        job.setMapOutputValueClass(Text.class);

        // Input
        job.setInputFormatClass(SequenceFileInputFormat.class);
        FileInputFormat.setInputPaths(job, input);

        // Reducer
        ConvergeCuboidDataUtil.setupReducer(job, originalSegment, output);

        // set job configuration
        job.getConfiguration().set(BatchConstants.CFG_CUBE_NAME, cubeName);
        job.getConfiguration().set(BatchConstants.CFG_CUBE_SEGMENT_ID, segmentID);
        // add metadata to distributed cache
        attachSegmentsMetadataWithDict(Lists.newArrayList(optSegment, originalSegment), job.getConfiguration());

        this.deletePath(job.getConfiguration(), output);

        return waitForCompletion(job);
    } catch (Exception e) {
        logger.error("error in CuboidJob", e);
        printUsage(options);
        throw e;
    } finally {
        if (job != null)
            cleanupTempConfFile(job.getConfiguration());
    }
}