Source Oracle pluggable database (PDB) #

This page contains instructions for setting up Oracle PDB as data source. Oracle PDB is a crucial part of Oracle multitenant architecture.

Arcion supports both snapshot and realtime replication for Oracle PDB.

Grant PDB permissions #

Ensure that you’re connected as a common user with privileges granted on both CDB$ROOT, the CDC container, and the PDB.

Provide following additional permissions:

GRANT SET CONTAINER TO <USERNAME> CONTAINER=ALL;
GRANT SELECT ON DBA_PDBS to <USERNAME> CONTAINER=ALL;

Change the open mode of the PDB to READ WRITE:

ALTER PLUGGABLE DATABASE $PDB_NAME OPEN READ WRITE FORCE;

Get the current SCN #

Take note of the current system change number (SCN) with the following SQL command:

select CURRENT_SCN from v$database;

The output is similar to the following:

CURRENT_SCN
-----------
    2401901

Configure snapshot replication #

For snapshot mode, follow these steps.

1. Set up connection configuration #

For connecting to the PDB database, provide its connection details to Replicant in the following format:

type: ORACLE

host: HOST
port: PORT_NUMBER

service-name: "SERVICE_NAME"

username: "USERNAME"
password: "PASSWORD"

max-connections: 30
max-retries: 10
retry-wait-duration-ms: 1000

Replace the following:

HOST: the Oracle PDB host
PORT_NUMBER: the port number of the PDB host
SERVICE_NAME: the service name for the PDB
USERNAME: the username that connects to the PDB
PASSWORD: the password associated with USERNAME

The following is a sample configuration:

type: ORACLE

host: 10.0.0.18
port: 1521

service-name: "ORCLPDB1"

username: 'alex'
password: 'alex1234'

max-connections: 30
max-retries: 10
retry-wait-duration-ms: 1000

2. Set up Extractor configuration #

For Extractor configuration, you can use the same parameters as a traditional Oracle database source. For more information, see Parameters related to snapshot mode and Snapshot mode in Extractor reference.

3. Run Replicant #

Run Replicant with the snapshot option, specifying the connection, Extractor, and Applier configuration files. For example:

./bin/replicant snapshot \
./conf/conn/oracle_multitenant_pdb.yaml ./conf/conn/yugabyte.yaml \
--extractor ./conf/src/oracle_common.yaml \
--applier ./conf/dst/yugabyte.yaml \
--filter ./filter/oracle19c_filter.yaml

For more information about running Replicant in snapshot mode, see Replicant snapshot mode.

Configure realtime replication #

For realtime mode, follow these steps.

1. Set up connection configuration #

For realtime replication, you need to connect to the multitenant container database (CDB). Using the pdb-connection field, you also need to pass the PDB connection details.

The following is the configuration structure:

type: ORACLE

host: CDB_HOST
port: CDB_PORT_NUMBER

service-name: "CDB_SERVICE_NAME"

username: "CDB_USERNAME"
password: "CDB_PASSWORD"

max-connections: 30
max-retries: 10
retry-wait-duration-ms: 1000

continuous-log-mining: {true|false}

pdb-connection:
  host: PDB_HOST
  port: PDB_PORT_NUMBER
  service-name: "PDB_SERVICE_NAME"
  username: "PDB_USERNAME"
  password: "PDB_PASSWORD"
  max-connections: 10
  pdb-name: PDB_NAME
  continuous-log-mining: {true|false}

Replace the following:

CDB_HOST: the Oracle CDB host
CDB_PORT_NUMBER: the port number of the CDB host
CDB_SERVICE_NAME: the service name for the CDB
CDB_USERNAME: the username that connects to the CDB
CDB_PASSWORD: the password associated with CDB_USERNAME
PDB_HOST: the Oracle PDB host
PDB_PORT_NUMBER: the port number of the PDB host
PDB_SERVICE_NAME: the service name for the PDB
PDB_USERNAME: the username that connects to the PDB
PDB_PASSWORD: the password associated with PDB_USERNAME

The following is a sample configuration:

type: ORACLE

host: 10.0.0.18
port: 1521

service-name: "ORCLCDB"

username: 'alex'
password: 'alex1234'

max-connections: 30
max-retries: 10
retry-wait-duration-ms: 1000
continuous-log-mining: false

pdb-connection:
  host: 10.0.0.18
  port: 1521
  service-name: "ORCLPDB1"
  username: 'alex'
  password: 'alex1234'
  max-connections: 10
  pdb-name: ORCLPDB1
  continuous-log-mining: false

2. Fetch schemas #

Before starting realtime replication, you need to fetch the schemas first by connecting to the PDB:

./bin/replicant fetch-schemas \
./conf/conn/oracle_multitenant_pdb.yaml \
--filter ./filter/oracle_filter.yaml \
--output-file ./oracle_schema.yaml

You can pass a different location for the output file.

3. Set up Extractor configuration #

For Extractor configuration, you need to specify your configuration under the realtime section of the Extractor configuration file. The following steps are specific to Oracle PDB:

I. Specify the heartbeat table and schema details #

Create the heartbeat table in the CDB and pass its details using the heartbeat parameter. For example:

realtime:
  heartbeat:
    enable: true
    schema: "C##REPLICANT"
    interval-ms: 10000

II. Specify the starting SCN #

Notice that we get the current SCN in the first section. For realtime replication, you need to specify that SCN as the starting SCN in the realtime section of the Extractor configuration file:

realtime:
  start-position:
    start-scn: 2393338

The reason for this is to apply anything that happens after the snapshot starts.

The rest of the Extractor parameters available to you are the same as a traditional Oracle database source. For more information, see Parameters related to realtime mode and Realtime mode Extractor reference.

4. Run Replicant #

As the last step, run Replicant with the necessary options and arguments. Remember to use the --src-schemas option to specify the schema we fetch in the second step.

./bin/replicant real-time \
./conf/conn/oracle_multitenant_pdb.yaml \ 
./conf/conn/yugabyte.yaml \ 
--extractor ./conf/src/oracle_common.yaml \ 
--applier ./conf/dst/yugabyte.yaml \ 
--filter ./filter/oracle_filter.yaml \
--src-schemas ./output/oracle_schema.yaml

In the preceding example, we run Replicant with the following options and arguments:

The connection configuration files of the source and the target
The Extractor configuration file with the --extractor option
The Applier configuration file with the --applier option
A filter file with the --filter option

For more information about running Replicant in realtime mode, see Replicant realtime mode.