[
https://issues.apache.org/jira/browse/IMPALA-3137?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Tim Armstrong resolved IMPALA-3137.
-----------------------------------
Resolution: Invalid
We don't track bugs with this driver on Apache JIRA.
> Switching between databases give a wrong result
> -----------------------------------------------
>
> Key: IMPALA-3137
> URL: https://issues.apache.org/jira/browse/IMPALA-3137
> Project: IMPALA
> Issue Type: Bug
> Components: Clients
> Affects Versions: Impala 2.2, Impala 2.3.0
> Reporter: Jiri Novak
> Priority: Minor
> Attachments: ImpalaSwitchDB.zip
>
>
> Setting default database in ODBC driver setup is case sensitive and affects
> the result of select in a very strange way.
> Use standard C# (.Net version is 4.5) and ADO.Net classes from
> System.Data.Odbc namespace to access ODBC data source, ODBC driver
> v2.5.30.1011 (32 bit)
> Test case -
> 1. Create two databases TESTDB1 and TESTDB2
> 2. Create two table TESTDB1.TABX and TESTDB2.TABX
> 3. Insert in table TESTDB1.TABX 5 rows
> 4. Set database on “Cloudera ODBC Driver for Impala DSN Setup” as TESTDB1
> (uppercase!)
> 5. Execute (both of them without statement terminator ';')
> {code:sql}
> USE TESTDB2
> SELECT COUNT(*) FROM TABX -- you will get the expected result – 0
> {code}
> 6. Repeat the step # 4 but in this case set the default database as
> testdb1 (lower case)
> 7. Execute the same script as in step # 5. Now you will get the result –
> 5!
> The table TABX has been resolved as TESTDB1.TABX and it’s wrong.
> There is a simple .Net application attached which can be used to simulate the
> issue. It is needed only to change a connection string directly in code.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
