菜单

本页面中的信息并不完全以您的首选语言展示,我们正在完善其他语言版本。想要以您的首选语言了解相关信息,可以点击这里下载PDF。

平台

Cloudera Impala

NOT ON THE CURRENT EDITION
This blip is not on the current edition of the radar. If it was on one of the last few editions it is likely that it is still relevant. If the blip is older it might no longer be relevant and our assessment might be different today. Unfortunately, we simply don't have the bandwidth to continuously review blips from previous editions of the radarUnderstand more
Nov 2015
试验?

For a while now the Hadoop community has been trying to bring low-latency, interactive SQL capability to the Hadoop platform (better known as SQL-on-Hadoop). This has led to a few open source systems such as Cloudera Impala, Apache Drill, Facebook’s Presto etc being developed actively through 2014. We think the SQL-on-Hadoop trend signals an important shift as it changes Hadoop's proposition from being a batch oriented technology that was complementary to databases into something that could compete with them.  Cloudera Impala was one of the first SQL-on-Hadoop platforms. It is a distributed, massively-parallel, C++ based query engine. The core component of this platform is the Impala daemon that coordinates the execution of the SQL query across one or more nodes of the Impala cluster. Impala is designed to read data from files stored on HDFS in all popular file formats. It leverages Hive's metadata catalog, in order to share databases and tables between the two database platforms. Impala comes with a shell as well as JDBC and ODBC drivers for applications to use. 

May 2015
试验?