• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>2017 International Conference on Inventive Computing and Informatics >HDFS memory usage analysis
【24h】

HDFS memory usage analysis

机译：HDFS内存使用情况分析

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Due to high use of online activities, there will be lot of data which is getting generated. To handle this data there should be an efficient system which can process the data effectively. One such system is Hadoop Distributed File System. HDFS consists of number of nodes, and one is the master among them while all others are slave nodes. Master node is called as NameNode and slave nodes are called as DataNodes. System will record the activities in the log file. Whatever the transactions are happening at the front end everything will get recorded at backend. We can capture this data and analyze to find out the system behavior in the failure situation. Since we are using the system for large transactions we need to have an idea on usage of the resources in the HDFS such as memory usage. Analyzers can get the stats of the operating system and the other devices separately i.e, for each service we need to have one analyzer to collect the stats. This paper presents the analyzer or shipper for collecting the metrics from operating system, memory and the other installed services.
机译：由于大量使用在线活动，因此将生成大量数据。为了处理这些数据，应该有一个有效的系统可以有效地处理数据。一种这样的系统是Hadoop分布式文件系统。 HDFS由多个节点组成，其中一个是主节点，而其他所有节点都是从节点。主节点称为NameNode，从节点称为DataNode。系统将活动记录在日志文件中。无论前端发生什么事务，所有内容都会在后端记录下来。我们可以捕获这些数据并进行分析以找出故障情况下的系统行为。由于我们将系统用于大型事务，因此我们需要了解HDFS中资源的使用情况，例如内存使用情况。分析器可以分别获取操作系统和其他设备的统计信息，即，对于每种服务，我们需要一个分析器来收集统计信息。本文介绍了用于从操作系统，内存和其他已安装服务中收集指标的分析器或托运人。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号