Skills / Competency: to install hive and associated derby database in Linux environment
Essential Steps:
- Follow the guides links attached, you may install with a pre-built/pre-compiled stable release, or source files with your own compilation. Follow the first if you are in-experienced
- Hive needs to connect to either an "Embedded/Local" or "Remote " database for storage of metadata
- Derby is the built-in or embedded DB that get downloaded, and can be configured as either mode
- Modification to relevant xml and environmental files (~/.bashrc, hive-site.xml, hive-env.sh) is required
- Java class path need to be properly modified to access the compiled jar files
Installation Guides:
Related Information:
Points to note in installation:
a) I realized in hive: hive-site.xml and in the installation steps to create /user/hive/warehouse:
"**user**" can be so easily confused with "**usr**" without any alert at all. so remember:
under hadoop fs, this is /user/hive/warehouse, which points to metadata_db folder, to create (NOT /usr/hive/warehouse)
b) It can be confusing as there are more than one way to set those parameters
There are two modes - embedded/local mode vs remote mode of database