Anyone who regularly works with query languages invariably develops personal preferences. For T-SQL, it may be a preference for which metadata is included with the results of a query, how quoted identifiers are handled, or what the default behavior should be for transactions. These types of settings can typically be configured at a session level, and Hive is no exception in allowing this. In fact, Hive provides users with an impressive number of configurable session properties. Honestly, you’ll probably never need to change the majority of these settings, and if/when you do, it’ll most likely apply to a specific Hive script (i.e. to improve performance). However, there are a handful of Hive settings that you may wish to always enable if they’re not already defaulted server-wide, such as displaying column headers. One option is to set these manually at the start of each session, using the SET command. But this can quickly get tedious if you have more than 1 or 2 settings to change. A better option in those scenarios, and the topic of this blog post, is to use a HiveRC file to configure your personal preferences for Hive’s default behavior.
For those of you not familiar with the concept, Linux commonly uses RC files — which I believe stands for “runtime configuration,” but don’t quote me on that 🙂 — for defining preferences, and various applications support these, typically in the format of .<app>rc. These will usually live in a user’s home directory, and some examples include .bashrc, .pythonrc, and .hiverc.
Now that we have a little context, let’s walk through how to create your personal .hiverc file. Note that all of these steps take place on the same server you use for connecting to Hive.
|
1 2 3 4 5 6 |
# bash # ensure you are in your personal home directory cd ~ # use Vim to create a personal Hive configuration file vim .hiverc |
Now, from inside Vim, do the following:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 |
# vim # enter "insert" mode in Vim i # type or copy/paste these values into your Vim file (right-click to paste) set hive.cli.print.current.db=true; set hive.cli.print.header=true; set hive.execution.engine=tez; set hive.auto.convert.join=true; # press the {ESC Key} to exit "insert" mode # write/save your changes :w # exit Vim :q |
You should be back at your bash prompt. Now run these commands to verify everything is working as expected.
|
1 2 3 4 5 6 7 8 9 10 |
# bash # view your HiveRC file to ensure everything looks good cat ~/.hiverc # enter Hive hive # verify settings set hive.cli.print.current.db; # etc. etc. |
That’s all there is to it! Not too hard, huh? But to make things even easier, I’ve posted an example of my personal HiveRC file on my Hadoopsie GitHub repo.
That’s all for now, you awesome nerds. 🙂