Red Hat Bugzilla – Bug 366431
unixODBC connection error with postgres when client_min_messages set to debug5, debug4 or debug3
Last modified: 2013-07-02 23:15:51 EDT
Description of problem:
In the configuration file of postgresql, postgresql.conf, if the
client_min_messages set to debug5, debug4 or debug3 then unixODBC throws error
message "Failed to SQLTables" while checking it via DataManager or if they check
it via isql they got the error message "[unixODBC]DEBUG: name: unnamed;
blockState: STARTED; state: INPROGR, xid/subid/cid: 777/1/0, nestlvl: 1,
children:[ISQL]ERROR: Could not SQLTables".
For all other (debug2, debug1, notice, error, log, warning) configuration, it
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1.Install unixODBC and postgresql
2. Configure /etc/odbc.ini and /etc/odbcinst.ini
3. Configure postgresql and set the "client_min_messages set = debug5" in
4. Create a Database.
5. Test the connectivity via DataManager or isql.
Throws error message for querying sql tables. Failed to SQLTables.
It should properly access the database and show the tables.
when there is a activity from the clinet, to debug any error, client can ask for
the debug level 5. For development environments, can be asked by the customer.
This report has no more information than you previously gave by email. What is the unixODBC version?
Which Postgres ODBC driver are you using, and what's the version of that? What's the full contents of the
two ODBC configuration files?
Created attachment 266851 [details]
The odbc.ini file has attached.
Created attachment 266861 [details]
(In reply to comment #1)
> This report has no more information than you previously gave by email. What
is the unixODBC version?
> Which Postgres ODBC driver are you using, and what's the version of that?
What's the full contents of the
> two ODBC configuration files?
Unix ODBC Version
unixODBC-2.2.11-7.1 -- both 32 bit and 64 bit
unixODBC-2.2.12-1.el5s2 -- both 32 bit and 64 bit
Postgresql ODBC version
RHEL-5 has now reached production phase 2, in which only security and other critical bugs will be addressed. This is neither, so I'm afraid it won't be fixed in RHEL-5. I believe the issue has been addressed in the newer postgresql-odbc code in RHEL-6. (Also, RHEL-5 users might be able to migrate to unixODBC64 and postgresql-odbc64, which should resolve the problem as well.)