简体繁体 English

使用Shell远程查询Hive

[英]Query Hive remotely using shell

原文 2016-02-02 08:44:43 9 2 shell/ hadoop/ hive/ remote-server/ hiveql

Let's imagine I have access to an Hive datawarehouse, I can query it using some webservice. 假设我可以访问Hive数据仓库，可以使用某些Web服务对其进行查询。 The problem is that I cannot automate the query using this service, so I would like to be able to query Hive from an external script (that I would be able to automate). 问题是我无法使用此服务自动执行查询，因此我希望能够从外部脚本查询Hive（使我能够自动执行）。

For now, I've only seen people running Hive on their local machine and querying it, I was wondering if it was possible to do it remotely ? 现在，我只看到有人在本地计算机上运行Hive并对其进行查询，我想知道是否可以远程进行？ If yes, how ? 如果是，怎么办？

Thanks a lot ! 非常感谢！

2 个解决方案

As far as I understood, you are asking if there are ways to connect to hive from a remote machine? 据我了解，您在问是否有方法可以从远程计算机连接到配置单元？

You could install hive client (beeline) on any remote machine and connect to hive via jdbc. 您可以在任何远程计算机上安装hive客户端（直线）并通过jdbc连接到hive。

Take a look here: https://cwiki.apache.org/confluence/display/Hive/HiveServer2+Clients 在这里看看： https : //cwiki.apache.org/confluence/display/Hive/HiveServer2+Clients

An easy way to do this, is to deploy the client configuration of hadoop/yarn on the remote machine. 一种简单的方法是在远程计算机上部署hadoop / yarn的客户端配置。 If the remote cluster is secured with firewalls and kerberos, you will need access to those first. 如果远程群集已使用防火墙和kerberos保护，则需要首先访问它们。 After that it's just a matter of starting up a hive shell or committing a job submit to Yarn. 之后，只需要启动配置单元外壳或提交作业即可。

When you use Cloudera, you might be able to add the host to the cluster and install a "gateway" role for yarn and hive on the target machine. 使用Cloudera时，您可能可以将主机添加到群集，并在目标计算机上为yarn和hive安装“网关”角色。 This is very straight-forward and requires just a few minutes of work. 这非常简单，只需几分钟的工作。

Alternatively using the JDBC connector should also work, as stated in Facha's answer. 如Facha的回答所述，也可以使用JDBC连接器。