Java FileJournalManager.EditLogFile方法代码示例

本文整理汇总了Java中org.apache.hadoop.hdfs.server.namenode.FileJournalManager.EditLogFile方法的典型用法代码示例。如果您正苦于以下问题：Java FileJournalManager.EditLogFile方法的具体用法？Java FileJournalManager.EditLogFile怎么用？Java FileJournalManager.EditLogFile使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类org.apache.hadoop.hdfs.server.namenode.FileJournalManager的用法示例。

在下文中一共展示了FileJournalManager.EditLogFile方法的4个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Java代码示例。

示例1: finalizeLogSegment

import org.apache.hadoop.hdfs.server.namenode.FileJournalManager; //导入方法依赖的package包/类
/**
 * Finalize the log segment at the given transaction ID.
 */
public synchronized void finalizeLogSegment(RequestInfo reqInfo, long startTxId,
    long endTxId) throws IOException {
  checkFormatted();
  checkRequest(reqInfo);

  boolean needsValidation = true;

  // Finalizing the log that the writer was just writing.
  if (startTxId == curSegmentTxId) {
    if (curSegment != null) {
      curSegment.close();
      curSegment = null;
      curSegmentTxId = HdfsConstants.INVALID_TXID;
    }
    
    checkSync(nextTxId == endTxId + 1,
        "Trying to finalize in-progress log segment %s to end at " +
        "txid %s but only written up to txid %s",
        startTxId, endTxId, nextTxId - 1);
    // No need to validate the edit log if the client is finalizing
    // the log segment that it was just writing to.
    needsValidation = false;
  }
  
  FileJournalManager.EditLogFile elf = fjm.getLogFile(startTxId);
  if (elf == null) {
    throw new JournalOutOfSyncException("No log file to finalize at " +
        "transaction ID " + startTxId);
  }

  if (elf.isInProgress()) {
    if (needsValidation) {
      LOG.info("Validating log segment " + elf.getFile() + " about to be " +
          "finalized");
      elf.scanLog();

      checkSync(elf.getLastTxId() == endTxId,
          "Trying to finalize in-progress log segment %s to end at " +
          "txid %s but log %s on disk only contains up to txid %s",
          startTxId, endTxId, elf.getFile(), elf.getLastTxId());
    }
    fjm.finalizeLogSegment(startTxId, endTxId);
  } else {
    Preconditions.checkArgument(endTxId == elf.getLastTxId(),
        "Trying to re-finalize already finalized log " +
            elf + " with different endTxId " + endTxId);
  }

  // Once logs are finalized, a different length will never be decided.
  // During recovery, we treat a finalized segment the same as an accepted
  // recovery. Thus, we no longer need to keep track of the previously-
  // accepted decision. The existence of the finalized log segment is enough.
  purgePaxosDecision(elf.getFirstTxId());
}

开发者ID:naver，项目名称:hadoop，代码行数:58，代码来源:Journal.java

示例2: finalizeLogSegment

import org.apache.hadoop.hdfs.server.namenode.FileJournalManager; //导入方法依赖的package包/类
/**
 * Finalize the log segment at the given transaction ID.
 */
public synchronized void finalizeLogSegment(RequestInfo reqInfo, long startTxId,
    long endTxId) throws IOException {
  checkFormatted();
  checkRequest(reqInfo);

  boolean needsValidation = true;

  // Finalizing the log that the writer was just writing.
  if (startTxId == curSegmentTxId) {
    if (curSegment != null) {
      curSegment.close();
      curSegment = null;
      curSegmentTxId = HdfsServerConstants.INVALID_TXID;
    }
    
    checkSync(nextTxId == endTxId + 1,
        "Trying to finalize in-progress log segment %s to end at " +
        "txid %s but only written up to txid %s",
        startTxId, endTxId, nextTxId - 1);
    // No need to validate the edit log if the client is finalizing
    // the log segment that it was just writing to.
    needsValidation = false;
  }
  
  FileJournalManager.EditLogFile elf = fjm.getLogFile(startTxId);
  if (elf == null) {
    throw new JournalOutOfSyncException("No log file to finalize at " +
        "transaction ID " + startTxId);
  }

  if (elf.isInProgress()) {
    if (needsValidation) {
      LOG.info("Validating log segment " + elf.getFile() + " about to be " +
          "finalized");
      elf.scanLog(Long.MAX_VALUE, false);

      checkSync(elf.getLastTxId() == endTxId,
          "Trying to finalize in-progress log segment %s to end at " +
          "txid %s but log %s on disk only contains up to txid %s",
          startTxId, endTxId, elf.getFile(), elf.getLastTxId());
    }
    fjm.finalizeLogSegment(startTxId, endTxId);
  } else {
    Preconditions.checkArgument(endTxId == elf.getLastTxId(),
        "Trying to re-finalize already finalized log " +
            elf + " with different endTxId " + endTxId);
  }

  // Once logs are finalized, a different length will never be decided.
  // During recovery, we treat a finalized segment the same as an accepted
  // recovery. Thus, we no longer need to keep track of the previously-
  // accepted decision. The existence of the finalized log segment is enough.
  purgePaxosDecision(elf.getFirstTxId());
}

开发者ID:aliyun-beta，项目名称:aliyun-oss-hadoop-fs，代码行数:58，代码来源:Journal.java

示例3: finalizeLogSegment

import org.apache.hadoop.hdfs.server.namenode.FileJournalManager; //导入方法依赖的package包/类
/**
 * Finalize the log segment at the given transaction ID.
 */
public synchronized void finalizeLogSegment(RequestInfo reqInfo, long startTxId,
    long endTxId) throws IOException {
  checkJournalStorageFormatted();
  checkRequest(reqInfo);
  
  boolean needsValidation = true;

  // Finalizing the log that the writer was just writing.
  if (startTxId == curSegmentTxId) {
    if (curSegment != null) {
      curSegment.close();
      curSegment = null;
      curSegmentTxId = HdfsConstants.INVALID_TXID;
      currentSegmentWrittenBytes = 0L;
    }
    
    checkSync(nextTxId == endTxId + 1,
        "Trying to finalize in-progress log segment %s to end at " +
        "txid %s but only written up to txid %s",
        startTxId, endTxId, nextTxId - 1);
    // No need to validate the edit log if the client is finalizing
    // the log segment that it was just writing to.
    needsValidation = false;
  }
  
  FileJournalManager.EditLogFile elf = fjm.getLogFile(startTxId);
  if (elf == null) {
    throw new JournalOutOfSyncException("No log file to finalize at " +
        "transaction ID " + startTxId);
  }

  if (elf.isInProgress()) {
    if (needsValidation) {
      LOG.info("Validating log segment " + elf.getFile() + " about to be " +
          "finalized");
      elf.validateLog();

      checkSync(elf.getLastTxId() == endTxId,
          "Trying to finalize in-progress log segment %s to end at " +
          "txid %s but log %s on disk only contains up to txid %s",
          startTxId, endTxId, elf.getFile(), elf.getLastTxId());
    }
    fjm.finalizeLogSegment(startTxId, endTxId);
  } else {
    Preconditions.checkArgument(endTxId == elf.getLastTxId(),
        "Trying to re-finalize already finalized log " +
            elf + " with different endTxId " + endTxId);
  }

  // Once logs are finalized, a different length will never be decided.
  // During recovery, we treat a finalized segment the same as an accepted
  // recovery. Thus, we no longer need to keep track of the previously-
  // accepted decision. The existence of the finalized log segment is enough.
  purgePaxosDecision(elf.getFirstTxId());
}

开发者ID:rhli，项目名称:hadoop-EAR，代码行数:59，代码来源:Journal.java

示例4: finalizeLogSegment

import org.apache.hadoop.hdfs.server.namenode.FileJournalManager; //导入方法依赖的package包/类
/**
 * Finalize the log segment at the given transaction ID.
 */
public synchronized void finalizeLogSegment(RequestInfo reqInfo, long startTxId,
    long endTxId) throws IOException {
  checkFormatted();
  checkRequest(reqInfo);

  boolean needsValidation = true;

  // Finalizing the log that the writer was just writing.
  if (startTxId == curSegmentTxId) {
    if (curSegment != null) {
      curSegment.close();
      curSegment = null;
      curSegmentTxId = HdfsConstants.INVALID_TXID;
    }
    
    checkSync(nextTxId == endTxId + 1,
        "Trying to finalize in-progress log segment %s to end at " +
        "txid %s but only written up to txid %s",
        startTxId, endTxId, nextTxId - 1);
    // No need to validate the edit log if the client is finalizing
    // the log segment that it was just writing to.
    needsValidation = false;
  }
  
  FileJournalManager.EditLogFile elf = fjm.getLogFile(startTxId);
  if (elf == null) {
    throw new JournalOutOfSyncException("No log file to finalize at " +
        "transaction ID " + startTxId);
  }

  if (elf.isInProgress()) {
    if (needsValidation) {
      LOG.info("Validating log segment " + elf.getFile() + " about to be " +
          "finalized");
      elf.validateLog();

      checkSync(elf.getLastTxId() == endTxId,
          "Trying to finalize in-progress log segment %s to end at " +
          "txid %s but log %s on disk only contains up to txid %s",
          startTxId, endTxId, elf.getFile(), elf.getLastTxId());
    }
    fjm.finalizeLogSegment(startTxId, endTxId);
  } else {
    Preconditions.checkArgument(endTxId == elf.getLastTxId(),
        "Trying to re-finalize already finalized log " +
            elf + " with different endTxId " + endTxId);
  }

  // Once logs are finalized, a different length will never be decided.
  // During recovery, we treat a finalized segment the same as an accepted
  // recovery. Thus, we no longer need to keep track of the previously-
  // accepted decision. The existence of the finalized log segment is enough.
  purgePaxosDecision(elf.getFirstTxId());
}

开发者ID:ict-carch，项目名称:hadoop-plus，代码行数:58，代码来源:Journal.java

注：本文中的org.apache.hadoop.hdfs.server.namenode.FileJournalManager.EditLogFile方法示例由纯净天空整理自Github/MSDocs等开源代码及文档管理平台，相关代码片段筛选自各路编程大神贡献的开源项目，源码版权归原作者所有，传播和使用请参考对应项目的License；未经允许，请勿转载。