[
https://issues.apache.org/jira/browse/PHOENIX-3710?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15981088#comment-15981088
]
Csaba Skrabak commented on PHOENIX-3710:
----------------------------------------
Workaround should work:
{noformat}
--data-table \"\"my_lowcase_table\"\"
{noformat}
Backslashes prevent bash from interpreting the quotation marks. The outmost
pair of double quotes will be swallowed by the commons-cli but one pair of
double quotes remain there. Then Phoenix code comes to interpret the thing as a
case sensitive table name.
> Cannot use lowername data table name with indextool
> ---------------------------------------------------
>
> Key: PHOENIX-3710
> URL: https://issues.apache.org/jira/browse/PHOENIX-3710
> Project: Phoenix
> Issue Type: Bug
> Affects Versions: 4.8.0
> Reporter: Matthew Shipton
> Priority: Minor
>
> {code}
> hbase org.apache.phoenix.mapreduce.index.IndexTool --data-table
> \"my_lowcase_table\" --index-table INDEX_TABLE --output-path /tmp/some_path
> {code}
> results in:
> {code}
> java.lang.IllegalArgumentException: INDEX_TABLE is not an index table for
> MY_LOWCASE_TABLE
> {code}
> This is despite the data table being explictly lowercased.
> Appears to be referring to the lowcase table, not the uppercase version.
> Workaround exists by changing the tablename, but this is not always feasible.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
