Wednesday, September 04, 2013

answer for minimal downtime to replatform a large database

conclusions of testing with incremental updated image copy.

1. if using current control file for recovery, the CURRENT online log are also needed. then open without RESETLOGS


2. If using backup controlfile , then needs to recover with "recover database noredo", followed by open RESETLOGS.

3. If using backup controlfile, once restored from datafilecopy, the COPY's info is no longer insided controlfile.  needs to re-create level0 datafilecopy after recovery, open RESETLOGS.

4. to minimize the downtime for large (TB level) database, can pre-copy a baseline datafiles via NFS for example, followed by rolling forward with incremental archived log files and current control files plus CURRENT online log files.