[PHOENIX-5319] Document: A note on lower case table/schema name for Bulkload

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 5.0.0
• Fix Version/s: 4.15.0, 5.1.0
• Assignee: Karthik Palanisamy
• Reporter: Karthik Palanisamy
• Created: 06/Jun/19 22:14
• Updated: 21/Dec/19 00:54
• Resolved: 11/Jul/19 02:32

Description

The problem with bulk loading CSV/JSON data with lower case table name. This will be fixed as part of PHOENIX-3541.

Created this separate doc Jira because Phoenix Website is maintained in svn. 

NOTE:

Table names in Phoenix are case insensitive( generally uppercase). but sometimes user may require to do mapping of existing HBase table with lowercase name into Phoenix table, In this case, Double quotes around table name i.e "tablename" can be used to preserve case sensitivity. The same was extended to the bulkload options, but due to the way Apache Commons CLI library parse command line options(Ref CLI-275), we need to pass the argument as \"\"tablename\"\" instead of just "tablename" for CsvBulkLoadTool.

Example:

hadoop jar phoenix-<version>-client.jar org.apache.phoenix.mapreduce.CsvBulkLoadTool --table \"\"t\"\" --input /data/example.csv

 

Issue Links

• requires: PHOENIX-3541 Bulk Data Loading - Can't use table name by small letter