= = = = in the preface

After Hive is successfully installed, if you have permission to log in to the cluster, directly enter Hive in the cluster shell to enter the Hive shell environment, as shown in Figure 1. However, for heavy workload projects, this kind of interactive platform with no intelligent hints, copy and paste query results, export data and poor, naturally very weak, here I recommend a few excellent interactive JDBC connection tools;

Official sons Hue and Zeppelin

Hue is usually only suitable for Hive. Zeppelin can also be used as spark’s interactive interface. Both versions are the same, and require a certain port for the cluster to open, download and configure. This is not the point;

== Dbeaver client ==

Dbeaver is one of my personal favorites to connect to Hive because:

  • Free, support a variety of common DB connection, network automatically download suitable JDBC driver package;
  • Style is like navICat and SSMS combination, support all kinds of data import and export, copy table header, generate data of all kinds of operation statements and other daily operations, but also can have a hint grammar, and generate and view e-R figure, perform shortcut keys and SSMS like Alt + X, very convenient;
  • Disadvantages: If you do not connect to Hive after half an hour, you will receive an error, as shown in Figure 2, and need to click disconnect/reconnect (only hive has this problem, hope to fix it soon).

Dbeaver download: dbeaver. IO /download/

Dbeaver configuration: First, find the hive-jdbC-23.5-standalone. Jar package in the Hive directory of your cluster installation and download it to a location on your machine. Then open up the installed DBeaver as shown, open up the data, create a new connection,

Add the standalone jar to the hive- JdbC-23.5-standalone. Click OK and click Finish.

==DG(Datagrip) Client ==

Datagrip is also a very fashionable big data component connection client.

  • Support a variety of common DB connection, network automatically download suitable JDBC driver package;
  • Style is like navICat and SSMS combination, support all kinds of data import and export, copy table header, generate data of all kinds of operation statements and other daily operations, but also can have prompt grammar, and generate and view e-R;
  • Hive automatically refreshes and disconnects or reconnects after about half an hour.
  • If no network, as long as there is network machine installation in a DB driver JDBC package, and then find the C: \ Users \ liuxiaowei DataGrip2019.3 \ config \ JDBC drivers – file, restart the client can copy in the past, as shown in figure 6.
  • Cons: Charges, free 30-day trial if you don’t pay.

  

Download link: www.jetbrains.com/datagrip/

Note: after the connection is configured, you need to edit the configuration again and select Schema and then All Schema. Otherwise, you may not see the database you want, as shown in Figure 7.

The final interface display is shown in Figure 8:

== One last word of advice ==

When an exception occurs, check whether the Hive JDBC configuration is started properly. The tool itself is not perfect. When dozens or hundreds of join queries are performed, the tool may not produce results. Don’t forget to run in the HIVE CLI environment. The CLI is only black box, no intelligent prompt, but is the most hive friendly interface, remember, remember!