The last line after all the redo logs and data files, is: ALTER DATABASE OPEN RESETLOGS Note the NOARCHIVELOG.

If you’re not sure of the syntax I’ll mail you a copy. Startup nomount Re-create the control file with the initial line – the database name in this case is DIFFERENT from the old database name – so I had to perform this extra step to rename everything: CREATE CONTROLFILE REUSE SET DATABASE “CS90PROD” RESETLOGS NOARCHIVELOG. Also include the standard – init.ora file, addition to the tnsnames.ora file, set up the spfile, orapw file. Hi, converting from the 10.2.0.5 database from 32-bit to 10.2.0.5 64-bit on another server: I performed an rsync copy of the entire database: data files, redo logs, control files – from one server to another, same exact directory setup. A recovery/restore is far more restrictive as it assumes same platform, same file locations, etc. The benefit of using expdp and empdp is that it facilitates crossing versions and platforms. This may mean changing DNS, tnsnames files, etc. – Migrate any client pointers to point to the new DB. Fix these, then rerun the import, choosing to overwrite the existing data. You will likely get some errors with permissions, etc.

– Create your users and tablespaces on the new system to exactly match the names and privileges on the old ones. This gets me the data without the system tablespace info (which will be populated as data is loaded). – What I did was to perform a database EXPORT using expdp of the database. Our setup: Old server: Windows Server 2005, Oracle 10g, APEX 3.2 New server: Windows Server 2008 64-bit, Oracle 11g, APEX 4.1 Having very recently done something very similar to this, I’ll tell you what we did.

Oracle 10g poster