Install Replay schema

Before gathering snapshots, a Replay schema must be created, where collected data will be stored.
The following is a step-by-step procedure to install this Replay schema.

1. Check requirements #

Tablespace #

First of all, a dedicated tablespace must exist, with an approximative size of 5GB per collected day, and per instance, in case of a RAC cluster. For example, if data are collected on a 2-nodes RAC cluster, then consider allocating 10GB per day.
These figures are very pessimistic and the actual disk usage can easily been monitored with the “Replay admin” feature.

Privileges #

The created Replay user will be granted few privileges. But installation procedure requires specific privileges to be able to create Replay user and grant access to catalog views and packages. That’s why installation must be performed from the database server, in order to connect / AS SYSDBA to the Oracle instance to be monitored by Replay. That means all required d.side scripts must be copied into a same location on the database server.

2. Ensure all d.side scripts are located in the same directory #

Installation scripts must be copied on Oracle server, where Replay schema creation will be run.
These scripts are part of d.side_home/d.side_scripts directory:

dscreate.sql
dscreate_pdb.sql
dsreplay.sql
dsprofiles.sql
dsalert.sql
dsreport.sql
dsreplaydiag.sql
create_console_refdb_schema.sql

All these scripts must be copied into the same location, as they are all called during installation process.

3. Go to this scripts directory, on database server, and connect #

cd /complete_path_to/d.side_scripts
sqlplus / as sysdba

4. Run schema creation script #

Prepare a Replay schema, ready to capture Oracle activity.

Choose the right script for the appropriate situation:

In multitenant environment, if you need to focus on a specific PDB and capture only the Oracle activity of this PDB, run:	Any other situation: non-multitenant, or CDB level of a multitenant environment, run:
@dscreate_pdb.sql	@dscreate.sql

The only difference between these two scripts is related to the connection string (requiring a container if PDB).

That’s it!
Replay schema is installed and ready to run a first capture.

For a global view of d.side Replay implementation, please refer to Replay Getting Started documentation.