Something strange is happening. I recently added a column to a table in Oracle9i, and when I am refreshing the structure in the universe, a designer error says ‘selected tables were not found’. In the table browser I can see this newly added column. Also, I could insert a new table in the universe. So what am I missing? :?
Perhaps you are missing the “owner” qualifier… If the"owner" of the table isn’t the same as that which is defined in the connection, that will explain your difference.
As long as the connection defined for the universe has select rights on the table, you can “rename” the table and add the owner on the appropriate line.
The owner of the table is the same as in the connection. I also tested the connection on a dummy universe, and inserted the changed table. In this universe, the table with the new column is inserted correcly. The owner qualifier is blank in both the universes.
Do you get this message on the table where the column was added or others? What happens if you try to parse an object from one of these “missing table”, do you get an error message? Are the objects defined as /Unknown/.column?
Without seeing it in action, it’s hard to describe what could be happening. It could be something as simple as having a lock on the table, or not committing the database changes.
Are you still having issues with this, or was it a one time occurence?
I found out the problem. I knew there was some problem with the connection. But couldnot figure it out :? . I was right, the connection I was using was ODBC instead of Oracle client. Once I used the new connection I could refresh the structure.
Got the same problem… Asked the DBA to check whether he granted select privilege to the changed table to enduser or not! DBA granted the privilege and everything worked fine
I know this topic actually is solved for the topic starter, but…
We do have a similar problem as described above, except that the ODBC connection is fine and it doesn’t concern an Oracle db.
Details:
BO 5.1.7
Progress 9.1d repository database
progress 10.0b (openEdge) datawarehouse/database
Progress connectivity files for BO 5.x (modified by progress for 10.0b)
2 universes on 2 different OpenEdge databases
the first universe we CAN import with Designer (or reporter for that matter). After import we receive the message that the universe was succesfully imported, but directly followed by:
[DataDirect][ODBC OPENEDGE driver][OPENEDGE]Internal error -20000 (see dsmret.h) in SQL from subsystem DSM SERVICES function dsmCursorFind called from ddm_get_flds on for . Save log for Progress technical support.
This message is displayed as many times as there are tables in the universe, in our case 42 times.
Also, we can’t refresh structure of insert tables: designer cannot find the selected tables.
But we can display table values and parse objects!!
The reports that were build on this universe also are working fine!
The second universe CANNOT be imported. The Designer hangs when the statusbar indicated that the joins are importing.
We have to end the Designer process with Windows Task Manager to end this.
We checked:
ODBC connections, middleware, drivers etc. > no problems
integrity of the universes > scan & repair OK
table owners all right? > same as in datasource
other universes on other databases > works fine
…
this problem never occurred before, when the same configuration was already working.
Is is possible it has to do with BO, or should we focus on Progress db?