SAP System Refresh Phases

Initialization

• Initialize the refresh
• Get the SID of source and target SAP systems and their databases
• Prepare file system on the target SAP system

Pre-processing

Extract data from the target SAP system. This data is preserved such that after the database copy it can be put back into the target SAP system.

• Save SAP system data from the target SAP system
  • Logical systems
  • Logon groups
  • RFC destinations
  • Users
  • SLD configuration
  • TMS
  • ALE configuration
  • …
• SAP data is extracted from SAP tables
  • Table names can be customized
  • Additional standard and custom tables can be defined
• SAP data is extracted using
  • R3trans
  • Direct database access for data that cannot be accessed through SAP means

Database Copy

The database copy itself is not part of the templates, but the templates can be used to drive the database copy.

In general there are two options to copy the database:

• Export and Import
• Backup and Restore

The standard template checks if custom chains are available to drive the database copy and calls them. Otherwise the refresh chain stops at the database copy and waits on an operator message for the confirmation that the database was laid over.

There are template chains for performing a standard set of steps for restoring Oracle database. These steps include:

• Retrieving the control file from the source system
• Checking and adapting file ownership of the database file structure
• Adapting the control file to rename the target database
• Renaming and applying redologs
• Creating OPS$ users
• Starting the listener

Export and Import

This database copy option employs either SAP tools, mainly R3load, or database tools. This option is generally slower than Backup and Restore, but allows simultaneous database reorganization, conversion, and even database release changes.

Backup and Restore

The database is backed up on the source SAP system and restored on the target SAP system. This option is generally faster than Export and Import, but it does not provide any reorganization opportunities and requires additional work in the database itself, eg. SID change.

Oracle Templates

For Oracle on UNIX there are a number of process definitions and chains that can be used to carry out database related steps like:

• Retrieval of the control file from the source Oracle database
• Rename of the target Oracle database
• Creation of the OPS$ users

The typical procedure here follows the following schema:

• Copy Source phase:
  • retrieve the control file from the source Oracle database
  • copy the control file from the source to the target system
• Copy Target phase:
  • get the archivelog mode
  • preserve SAP database users
  • wait for the actual database copy
  • change ownership - calls a custom process definition
  • check ownership of the database files
  • adjust the control file retrieved from the source database
  • rename the target database
  • start the target database
  • restore SAP database users
  • create / modify OPS$ users
  • ensure the listener for the target database is running
  • check the connection to the database via R3Trans

Post-Processing

In this step the data exported from the target SAP system is re-imported and additional customization is carried out, such that the target SAP system is ready to be used.

• Import preserved data
• Delete contents of some SAP tables (customizable)
  • ALCONSEG
  • ALSYSTEMS
  • DBSNP
  • MONI
  • OSMON
  • PAHI
  • SDBAD
  • SDBAP
  • SDBAR
  • TPFET
  • TPHFT
  • BDLS
  • Prepare the database for BDLS
  • Test logical systems in parallel
  • Convert logical systems sequentially
  • Reconfigure database after BDLS
• Cleanup the target SAP system
  • Remove old batch processes, reschedule processes
  • Joblogs
  • TemSe
  • Dumps
  • Printers
  • …
• Close clients
• Set login screen
• Create system message
• Notify users

Finalize the refresh

• Gather system status information after the refresh for comparison
• Test RFC destinations

See Also

System Copy

sscpTopic