We are running large scale 18.104.22.168 databases on Sun E10K platforms. We have noted an occasional problem with a locally managed TEMP tablespace, asserting an ORA-01652 error.
Our research indicates that there may be an Oracle bug at play here, but we have not been able to produce a test case to demonstrate that LMT TEMP is inherently flawed as compared to DMT TEMP behaviour. Can such a test case be provided?
There is a bug in Oracle 8i relating to LMT TEMP tablespaces. The problem is that the database is not reusing temporary sort segments. If you are receiving the ORA-1652 errors with your application, then your test case is to create a DMT TEMP tablespace and then rerun your application. If you are running into the bug, then you should not be receiving the ORA-1652 error any more. If you are still receiving the error, then your TEMP tablespace clearly is not large enough to support your application. If it is the bug, then your options are to use a DMT TEMP tablespace or upgrade and patch to Oracle 22.214.171.124 or 10g.
Dig Deeper on Oracle database design and architecture
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.