Determine the location and set up a directory where the captured workload will be stored. Before starting the workload capture, ensure that the directory is empty and has ample disk space to store the workload. If the directory runs out of disk space during a workload capture, the capture will stop. To estimate the amount of disk space that is required, you can run a test capture on your workload for a short duration (such as a few minutes) to extrapolate how much space you will need for a full capture. To avoid potential performance issues, you should also ensure that the target replay directory is mounted on a separate file system.
For Oracle RAC, consider using a shared file system. Alternatively, you can set up one capture directory path that resolves to separate physical directories on each instance, but you will need to consolidate the files created in each of these directories into a single directory. For captures on an Oracle RAC database, Enterprise Manager only supports Oracle RAC configured with a shared file system. The entire content of the local capture directories on each instance (not only the capture files) must be copied to the shared directory before it can be used for preprocessing or data masking. For example, assume that you are:
Running an Oracle RAC environment in Linux with two database instances named host1 and host2
Using a capture directory object named CAPDIR that resolves to /$ORACLE_HOME/rdbms/capture on both instances
Using a shared directory that resides in /nfs/rac_capture
You will need to login into each host and run the following command:
cp -r /$ORACLE_HOME/rdbms/capture/* /nfs/rac_capture
After this is done for both instances, the /nfs/rac_capture shared directory is ready to be preprocessed or masked.