Ejemplo n.º 1
   * Returns statistics on this table as a tabular result set. Used for the SHOW TABLE STATS
   * statement. The schema of the returned TResultSet is set inside this method.
  public TResultSet getTableStats() {
    TResultSet result = new TResultSet();
    TResultSetMetadata resultSchema = new TResultSetMetadata();
    resultSchema.addToColumns(new TColumn("Region Location", ColumnType.STRING.toThrift()));
    resultSchema.addToColumns(new TColumn("Start RowKey", ColumnType.STRING.toThrift()));
    resultSchema.addToColumns(new TColumn("Est. #Rows", ColumnType.BIGINT.toThrift()));
    resultSchema.addToColumns(new TColumn("Size", ColumnType.STRING.toThrift()));

    // TODO: Consider fancier stats maintenance techniques for speeding up this process.
    // Currently, we list all regions and perform a mini-scan of each of them to
    // estimate the number of rows, the data size, etc., which is rather expensive.
    try {
      long totalNumRows = 0;
      long totalHdfsSize = 0;
      List<HRegionLocation> regions =
              hTable_, HConstants.EMPTY_END_ROW, HConstants.EMPTY_START_ROW);
      for (HRegionLocation region : regions) {
        TResultRowBuilder rowBuilder = new TResultRowBuilder();
        HRegionInfo regionInfo = region.getRegionInfo();
        Pair<Long, Long> estRowStats =
            getEstimatedRowStats(regionInfo.getStartKey(), regionInfo.getEndKey());

        long numRows = estRowStats.first.longValue();
        long hdfsSize = getHdfsSize(regionInfo);
        totalNumRows += numRows;
        totalHdfsSize += hdfsSize;

        // Add the region location, start rowkey, number of rows and raw Hdfs size.

      // Total num rows and raw Hdfs size.
      if (regions.size() > 1) {
        TResultRowBuilder rowBuilder = new TResultRowBuilder();
    } catch (IOException e) {
      throw new RuntimeException(e);
    return result;
Ejemplo n.º 2
 public void verify_string_column_type_has_code_of_2() {
   assertThat(ColumnType.STRING.value(), is(2));