Java Scan.setId方法代码示例

import org.apache.hadoop.hbase.client.Scan; //导入方法依赖的package包/类
private ResultScanner[] splitScan(Table table, Scan originalScan, AbstractRowKeyDistributor
        rowKeyDistributor) throws IOException {
    Scan[] scans = rowKeyDistributor.getDistributedScans(originalScan);
    final int length = scans.length;
    for (int i = 0; i < length; i++) {
        Scan scan = scans[i];
        // other properties are already set upon construction
        scan.setId(scan.getId() + "-" + i);
    }

    ResultScanner[] scanners = new ResultScanner[length];
    boolean success = false;
    try {
        for (int i = 0; i < length; i++) {
            scanners[i] = table.getScanner(scans[i]);
        }
        success = true;
    } finally {
        if (!success) {
            closeScanner(scanners);
        }
    }
    return scanners;
}
 

import org.apache.hadoop.hbase.client.Scan; //导入方法依赖的package包/类
private void doRawScan() throws IOException {
  FilterList filterList = new FilterList();
  CompareFilter.CompareOp startOp = CompareFilter.CompareOp.GREATER_OR_EQUAL;
  CompareFilter.CompareOp stopOp = CompareFilter.CompareOp.LESS_OR_EQUAL;
  for (int i = 0; i < indexColumnNames.length && i < scanValues.length; i++) {
    filterList.addFilter(
        new SingleColumnValueFilter(familyName, Bytes.toBytes(indexColumnNames[i]), startOp,
            Bytes.toBytes(scanValues[i][0])));
    filterList.addFilter(
        new SingleColumnValueFilter(familyName, Bytes.toBytes(indexColumnNames[i]), stopOp,
            Bytes.toBytes(scanValues[i][1])));
  }
  Scan scan = new Scan();
  scan.setFilter(filterList);
  scan.setId("raw-scan");
  Table table = conn.getTable(tableName);
  ResultScanner scanner = table.getScanner(scan);
  Result result;
  int count = 0;
  while ((result = scanner.next()) != null) {
    ++count;
    if (PRINT_RESULT) printResult(result);
  }
  scanner.close();
  System.out.println("raw scan has " + count + " records");
}
 

import org.apache.hadoop.hbase.client.Scan; //导入方法依赖的package包/类
private ResultScanner getNextScanner() throws IOException {
  if (INIT_REGION_SIZE != getRegionNumber()) {
    throw new IOException(
        "region number changed from " + INIT_REGION_SIZE + " to " + getRegionNumber());
  }
  if (regionLocationQueue.isEmpty()) return null;
  HRegionLocation regionLocation = regionLocationQueue.poll();

  Scan newScan = new Scan(rawScan);
  byte[] key = regionLocation.getRegionInfo().getStartKey();
  if (key != null && key.length > 0) newScan.setStartRow(key);
  key = regionLocation.getRegionInfo().getEndKey();
  if (key != null && key.length > 0) newScan.setStopRow(key);
  newScan.setAttribute(IndexConstants.SCAN_WITH_INDEX, Bytes.toBytes("Hi"));
  newScan.setId(rawScan.getId());
  newScan.setCacheBlocks(rawScan.getCacheBlocks());
  newScan.setCaching(rawScan.getCaching());
  return table.getScanner(newScan);
}
 

import org.apache.hadoop.hbase.client.Scan; //导入方法依赖的package包/类
private ResultScanner getNextScanner() throws IOException {
  if (INIT_REGION_SIZE != getRegionNumber()) {
    throw new IOException(
        "region number changed from " + INIT_REGION_SIZE + " to " + getRegionNumber());
  }
  if (regionLocationQueue.isEmpty()) return null;
  HRegionLocation regionLocation = regionLocationQueue.poll();
  Scan newScan = new Scan(rawScan);
  if (regionLocation.getRegionInfo().getStartKey() != null)
    newScan.setStartRow(regionLocation.getRegionInfo().getStartKey());
  if (regionLocation.getRegionInfo().getEndKey() != null)
    newScan.setStopRow(regionLocation.getRegionInfo().getEndKey());
  newScan.setAttribute(IndexConstants.SCAN_WITH_INDEX, Bytes.toBytes("Hi"));
  newScan.setFilter(rangeList.toFilterList());
  newScan.setAttribute(IndexConstants.MAX_SCAN_SCALE, Bytes.toBytes(1.0f));
  newScan.setId(rawScan.getId());
  newScan.setCacheBlocks(rawScan.getCacheBlocks());
  newScan.setCaching(rawScan.getCaching());
  return table.getScanner(newScan);
}
 

import org.apache.hadoop.hbase.client.Scan; //导入方法依赖的package包/类
private Scan[] splitScans(Scan originalScan) throws IOException {
    Scan[] scans = this.keyDistributor.getDistributedScans(originalScan);
    for (int i = 0; i < scans.length; ++i) {
        Scan scan = scans[i];
        scan.setId(originalScan.getId() + "-" + i);
    }
    return scans;
}
 

import org.apache.hadoop.hbase.client.Scan; //导入方法依赖的package包/类
private void doIndexScan() throws IOException {
  ScanRange.ScanRangeList rangeList = new ScanRange.ScanRangeList();
  FilterList filterList = new FilterList();
  CompareFilter.CompareOp startOp = CompareFilter.CompareOp.GREATER_OR_EQUAL;
  CompareFilter.CompareOp stopOp = CompareFilter.CompareOp.LESS_OR_EQUAL;
  for (int i = 0; i < indexColumnNames.length && i < scanValues.length; i++) {
    rangeList.addScanRange(new ScanRange(familyName, Bytes.toBytes(indexColumnNames[i]),
        Bytes.toBytes(scanValues[i][0]), Bytes.toBytes(scanValues[i][1]), startOp, stopOp,
        DataType.INT));
    filterList.addFilter(
        new SingleColumnValueFilter(familyName, Bytes.toBytes(indexColumnNames[i]), startOp,
            Bytes.toBytes(scanValues[i][0])));
    filterList.addFilter(
        new SingleColumnValueFilter(familyName, Bytes.toBytes(indexColumnNames[i]), stopOp,
            Bytes.toBytes(scanValues[i][1])));
  }
  Scan scan = new Scan();
  scan.setFilter(filterList);
  if (rangeList.getRanges().size() > 0) {
    scan.setAttribute(ScanRange.SCAN_RANGE_ATTRIBUTE_STR, rangeList.toBytesAttribute());
  }
  scan.setId("LMD-scan");
  scan.setCaching(1);
  ResultScanner scanner = BaseIndexScanner.getIndexScanner(conn, relation, scan);
  Result result;
  int count = 0;
  while ((result = scanner.next()) != null) {
    count++;
    if (PRINT_RESULT) printResult(result);
  }
  scanner.close();
  System.out.println("LMDIndex scan has " + count + " records");
}
 

import org.apache.hadoop.hbase.client.Scan; //导入方法依赖的package包/类
/**
 * init selected range and scanner
 *
 * @throws IOException
 */
private void initScanner() throws IOException {
  ScanRange selectedRange = null;
  int selectedRegionNumber = Integer.MAX_VALUE;
  for (ScanRange range : rangeList.getRanges()) {
    int cover = countCoveringRegions(conn,
        relation.getIndexTableName(range.getFamily(), range.getQualifier()), range.getStart(),
        range.getStop());
    LOG.info("LCDBG, " + cover + " regions are covered by range " + range);
    if (selectedRegionNumber > cover) {
      selectedRegionNumber = cover;
      selectedRange = range;
    }
  }
  LOG.info("LCDBG, GC Scanner using range " + selectedRange + " with " + selectedRegionNumber
      + " regions for scan id= " + rawScan.getId());
  indexFamily = selectedRange.getFamily();
  indexQualifier = selectedRange.getQualifier();
  List<ScanRange> list = new ArrayList<>(rangeList.getRanges());
  list.remove(selectedRange);
  Scan scan = new Scan();
  scan.setStartRow(selectedRange.getStart());
  scan.setStopRow(selectedRange.getStop());
  scan.setFamilyMap(rawScan.getFamilyMap());
  scan.setCaching(rawScan.getCaching());
  scan.setCacheBlocks(rawScan.getCacheBlocks());
  scan.setId(rawScan.getId());
  scan.setFilter(new ScanRange.ScanRangeList(list).toFilterList());
  Table table = conn.getTable(
      relation.getIndexTableName(selectedRange.getFamily(), selectedRange.getQualifier()));
  scanner = table.getScanner(scan);
}
 

import org.apache.hadoop.hbase.client.Scan; //导入方法依赖的package包/类
private ScanResult runOneTime(IndexType indexType, String fileName, Configuration conf,
    int cacheSize, AbstractWorkload workload) throws IOException {
  ScanRange.ScanRangeList rangeList = ScanRange.ScanRangeList.getScanRangeList(fileName);
  System.out.println("scan range in file " + fileName + " is: " + rangeList);
  DITBScanBase scanner;
  if (indexType == IndexType.MDIndex) {
    scanner = new DITBMDIndexScanner(workload.getTableName(), indexType, workload);
  } else {
    scanner = new DITBNormalScanner(workload.getTableName(), indexType, workload);
  }
  Scan scan = ScanRange.ScanRangeList.getScan(fileName);
  scan.setCacheBlocks(false);
  scan.setCaching(cacheSize);
  scan.setId(fileName);
  System.out.println("scan filter " + scan.getFilter());
  //    ScanResult scanResult = new ScanResult(fileName + ": " + rangeList);
  ScanResult scanResult = new ScanResult();
  int reportCount = 0;
  Result[] results;
  long timeStart = System.currentTimeMillis();
  ResultScanner resultScanner = scanner.getScanner(scan);
  try {
    while (true) {
      results = resultScanner.next(cacheSize);
      if (scanResult.firstLatency == -1)
        scanResult.firstLatency = System.currentTimeMillis() - timeStart;
      if (results == null || results.length == 0) break;
      if (PRINT_RESULT) {
        for (Result result : results)
          System.out.println(workload.parseResult(result));
      }
      scanResult.nbResults += results.length;
      reportCount += results.length;
      if (reportCount >= reportInterval) {
        System.out.println(
            "finish scan for " + results.length + " records, " + scanResult.nbResults
                + " in total");
        reportCount = 0;
      }
    }
  } catch (IOException e) {
    e.printStackTrace();
  }
  scanResult.totalTime = System.currentTimeMillis() - timeStart;
  if (scanResult.nbResults == 0) scanResult.nbResults = 1;
  return scanResult;
}
 

import org.apache.hadoop.hbase.client.Scan; //导入方法依赖的package包/类
/**
 * scan all index tables, common rowkeys will be saved in rowkeySet
 * can be optimized in 2 ways:
 * 1. scan index tables by the order of #CandidateRowkeys, similar to CCIndex
 * 2. scan index tables in parallel
 *
 * @throws IOException
 */
public static Queue<byte[]> createRowkeyQueueBySecondaryIndex(Connection conn,
    IndexTableRelation relation, Map<byte[], NavigableSet<byte[]>> familyMap,
    ScanRange.ScanRangeList rangeList, Scan rawScan) throws IOException {
  TreeSet<byte[]> rowkeySet = null;
  long timeToMerge = 0;
  for (ScanRange range : rangeList.getRanges()) {
    Scan scan = new Scan();
    scan.setStartRow(range.getStart());
    scan.setStopRow(range.getStop());
    scan.setFamilyMap(familyMap);
    scan.setCaching(rawScan.getCaching());
    scan.setCacheBlocks(rawScan.getCacheBlocks());
    scan.setId(rawScan.getId());
    if (range.getStartTs() != -1 && range.getStopTs() != -1) {
      scan.setTimeRange(range.getStartTs(), range.getStopTs());
    }
    TableName tableName = relation.getIndexTableName(range.getFamily(), range.getQualifier());
    Table table = conn.getTable(tableName);
    ResultScanner scanner = table.getScanner(scan);
    Result res;
    long timeStart = System.currentTimeMillis();
    TreeSet<byte[]> candidateSet = new TreeSet<>(Bytes.BYTES_COMPARATOR);
    while ((res = scanner.next()) != null) {
      candidateSet.add(IndexPutParser.parseIndexRowKey(res.getRow())[0]);
    }
    System.out.println(String
        .format("get %d candidate rowkeys from %s in scan %s, cost %.2f seconds",
            candidateSet.size(), range.toString(), scan.getId(),
            (System.currentTimeMillis() - timeStart) / 1000.0));
    if (rowkeySet == null) {
      rowkeySet = candidateSet;
    } else {
      timeStart = System.currentTimeMillis();
      rowkeySet = getCommonSet(rowkeySet, candidateSet);
      timeToMerge += (System.currentTimeMillis() - timeStart);
    }
    System.out.println(
        "common key set size " + rowkeySet.size() + " after " + range + " in scan " + scan
            .getId());
    if (rowkeySet.isEmpty()) { // no commons keys at all, can ignore the rest index tables
      break;
    }
  }
  System.out.println(String
      .format("get %d result rowkeys in scan %s, cost %.2f seconds", rowkeySet.size(),
          rawScan.getId(), timeToMerge / 1000.0));
  if (rowkeySet != null && !rowkeySet.isEmpty()) {
    Queue<byte[]> rowkeyQueue = new LinkedList<>();
    for (byte[] rowkey : rowkeySet)
      rowkeyQueue.add(rowkey);
    return rowkeyQueue;
  }
  return null;
}