Hypertable 文档架构.docx

Architecture(/documentation/architecture/)架构Google ArchitectureGoolge的架构Hypertable is a massively scalable database modeled after Googles Bigtable database. Bigtable is part of a group of scalable computing technologies developed by Google which is depicted in the following diagram.Hypertable是取材于Google Bigtable的一个高度可伸缩的数据库。Bigtable是Google开发的一系列可伸缩计算技术中的一个，如下图所示。Google File System (GFS) - This is the lowest layer of the Google scalable computing stack. It is a filesystem much like any other and allows for the creation of files and directories. The primary innovation of the Google filesystem is that it is massively scalable and highly available. It achieves high availability by replicating file data across three physical machines which means that it can lose up to two of the machines holding replicas and the data is still available. Hadoop provides an open source implementation of the GFS called HDFS.Google File System(GFS)-这是Google可伸缩计算层次中最底层。它是一个文件系统，同其他文件系统很像，考虑了文件和目录的创建。GFS的创新在于它是高可伸缩和高可靠的。它的高可靠性通过文件数据复制到3台物理机器中来实现，这样，即使有2台复制的数据失效，数据仍然可用。Hadoop提供了GFS的一个开源实现，叫做HDFS。MapReduce - This is a parallel computation framework designed to efficiently process data in the GFS. It provides a way to run a large amount of data through a piece of code (map) in parallel by pushing the code out to the machines where the data resides. It also includes a final aggregation step (reduce) which provides a way to re-order the data based on any arbitrary field. Hadoop provides an open source implementation of MapReduce.MapReduce–这是一个平行计算架构，目的是高效地处理GFS中的数据。它将一段程序（map）放到存放数据的计算机上运行，从而提供了一种平行处理大容量数据的方法。它也包含最终的聚集步骤（reduce），这一步骤将各节点数据再次排序。Hadoop提供了一个MapReduce的开源实现。Bigtable - This is Googles scalable database. It provides a way to create massive tables of information indexed by a primary key. As of this writing, over 90% of Googles web services are built on top of Bigtable, including Search, Google Earth, Google Analytics, Google Maps, Gmail, Orkut, YouTube, and many more. Hypertable