If both the primary database and the backup database are updated and users are prevented from logging on, (for example, users are specifically revoked, the SETROPTS INACTIVE is set to one day, and a weekend is passed and all users revoked.), and if you prefer to fix your primary, instead of using the most recent offline backup of your database (which might need a "replay" of all valid commands since that backup was made), you might do the following.

1. Use an emergency ICHRDSNT, which names as the emergency primary a new RACF® database, and, which names the production primary database as the emergency backup database. Make sure that you set the flag to duplicate updates for the backup.
2. Re-IPL.
3. Log on to IBMUSER. For example, all of your users in your production database are revoked, however, on the emergency primary, IBMUSER is not revoked, and is special.
   Note:

   If IBMUSER on your primary production database is revoked, you might unrevoke IBMUSER right away, or after step 4.

4. Issue RVARY SWITCH and then the emergency primary is the production primary (and IBMUSER's ACEE is still special). Continue with whatever is appropriate for your situation (for example, unrevoke your users, change SETROPTS INACTIVE back to a reasonable value, and so on).
5. Resync your offline production backup database from your updated primary database with IRRUT200 or IRRUT400.
   Note:

   Since the production backup database is offline, IRRUT200's PARM=ACTIVATE is not applicable.

6. IPL with your normal ICHRDSNT to activate the real production primary and backup databases.