We had Holidays in Israel and after that I have attended an Application Server 10g R2 Administration course in Barcelona.
Now back to work I started to recover and document a set of Crash Scenarios that were destroyed during a round of implementation.
The Crash Scenarios were developed to be run by DBA's that are not familiar with RAC 10g and Rman, so they can test and solve production situations later, with confidence.
The crash scenarios consist of a crash script, that actually executes the scenario, produces the crash, and a set of guidelines to recover from it.
Crash Scenario 1 is simple:
- The database crash,
- A tablespace is deleted from ASM while the database is down,
- The database fail to start
- The datafiles from the tablespace needs to be recovered using an rman backup.
- Startup mount
- Invoke rman
- Restore datafile
- Recover datafile
- Open the database
