This defect has been around a long time and continues to be a problem in the 10.1.3 release. This problem ONLY affects systems that use Oracle for the database. This is a problem with Primary and Aux tables but you seem to notice it more with Aux tables by the way you search.
LONG RAW had to be chosen because the Microsoft Oracle ODBC driver couldn't identify/use BLOB types (which were new types at the time, and it was the MS driver that everybody thought we should use). Note that Microsoft hasn't changed or improved their driver since that time, and we no longer consider the MS driver part of our "supported" TeamTrack configuration.
Upgrading to Oracle 11.2.0.3 appears to resolve the issue. This may be related to an Oracle bug having to do with the performance of hash joins in Oracle 11.2.0.2 running on IBM AIX.
it is possible that not all scenarios for upgrade of TeamTrack databases to SBM have been covered. While there was a defect fixed in 10.1.4 to handle this, you may still find issues when attempting upgrade of TT 661 databases.
The Administrator portion of the upgrade moves data back from temporary locations to their final storage columns. It does this by creating more columns in the destination table, converts the data, then drops the old columns. By doing this, it runs into the 1000 column limit in Oracle.
If you have long Workflow names and long Orchestration workflow names, you can end up with a name that is too long for one of the fields in the ODE database. If this happens (or you see the above message in the Event Manager logs), then contact Support for a workaround if this Defect does not show as resolved.
