ora2pgWorkshop

Useful links:

Workshop: https://github.com/MicrosoftDocs/azure-docs/blob/master/articles/dms/tutorial-oracle-azure-postgresql-online.md

ora2pg image: https://hub.docker.com/r/georgmoser/ora2pg-docker

Oracle image: https://hub.docker.com/r/jaspeen/oracle-11g

HR DDLScript - Import the HR sample: https://www.oracle.com/database/technologies/appdev/datamodeler-samples.html

Examples: https://www.darold.net/confs/ora2pg_the_hard_way.pdf

The workshop was checked against Centos 7.

Setting up ora2pg:

sudo -i

From root user:

yum update -y
yum install docker -y

systemctl start docker
systemctl status docker

docker pull georgmoser/ora2pg-docker

mkdir /data /config

Get inside container:

docker run -it --privileged -v /config/:/config -v /data:/data georgmoser/ora2pg-docker /bin/bash

ora2pg --version

apt-get update -y

apt-get install vim

Generate a migration project:

ora2pg --project_base /data --init_project myproject

change ORACLE_DSN in config file:

vi /data/myproject/config/ora2pg.conf

cd data/myproject/

Discovery:

ora2pg -c config/ora2pg.conf -t SHOW_VERSION

ora2pg -c config/ora2pg.conf -t SHOW_SCHEMA 

Check which tables contain HR schema:

ora2pg -c config/ora2pg.conf -t SHOW_TABLE -n HR

List columns of table JOBS:

ora2pg -c config/ora2pg.conf -t SHOW_COLUMN -a 'TABLE[JOBS]' -n HR

Generate report

ora2pg -c config/ora2pg.conf -t SHOW_REPORT --estimate_cost --dump_as_html -n HR > reports/report.html

SCHEMA Migration

Offline migration

Inside /data/myproject directory create a new directory:

mkdir offline
cd offline

Create a new file with following content:

vi countries.sql

CREATE TABLE COUNTRIES 
    ( 
     COUNTRY_ID CHAR (2 BYTE)  NOT NULL , 
     COUNTRY_NAME VARCHAR2 (40 BYTE) , 
     REGION_ID NUMBER 
    ) LOGGING 
;

now run ora2pg against this file:

root@13a8720887da:/data/myproject/offline# ora2pg -i countries.sql -t TABLE -c ../config/ora2pg.conf 
[========================>] 1/1 tables (100.0%) end of table export.

root@13a8720887da:/data/myproject/offline# ls
CONSTRAINTS_output.sql	INDEXES_output.sql  countries.sql  output.sql

Investigate the content of 3 files that has been created.

Online migration

Create online directory inside your project:

root@13a8720887da:/data/myproject/offline# mkdir /data/myproject/online
root@13a8720887da:/data/myproject/offline# cd /data/myproject/online

Get the sources of oracle procedures:

root@13a8720887da:/data/myproject/online# ora2pg -t PROCEDURE -o procedure.sql -c ../config/ora2pg.conf
[========================>] 2/2 procedures (100.0%) end of procedures export.
root@13a8720887da:/data/myproject/online# ls
ADD_JOB_HISTORY_procedure.sql  SECURE_DML_procedure.sql  procedure.sql

Investigate the content of generated files

Add -p to the command above to convert procedures to plpgsql:

root@13a8720887da:/data/myproject/online# ora2pg -p -t PROCEDURE -o procedure.sql -c ../config/ora2pg.conf
[========================>] 2/2 procedures (100.0%) end of procedures export.
root@13a8720887da:/data/myproject/online# ls
ADD_JOB_HISTORY_procedure.sql  SECURE_DML_procedure.sql  procedure.sql

The previous content has been overwritten. Check if conversion went well by exploring the content of the files.

Schema export

Let run the export then!

change the name of the schema in ora2pg.conf file:

# Oracle schema/owner to use
SCHEMA  CHANGE_THIS_SCHEMA_NAME

So it looks as follows:

# Oracle schema/owner to use
SCHEMA  HR

Run the export. Make sure you are in the /data/myproject directory:

root@13a8720887da:/data/myproject# pwd
/data/myproject

root@13a8720887da:/data/myproject# ./export_schema.sh

Check the content of two directories:

• sources - where oracle sources are kept
• schema - with objects converted to PostgreSQL

DATA Export

From /data/myproject run the following command:

ora2pg -t COPY -o data.sql -b ./data -c ./config/ora2pg.conf

Investigate the content of the data directory:

root@13a8720887da:/data/myproject# cd data/
root@13a8720887da:/data/myproject/data# ls
COUNTRIES_data.sql  DEPARTMENTS_data.sql  EMPLOYEES_data.sql  JOBS_data.sql  JOB_HISTORY_data.sql  LOCATIONS_data.sql  REGIONS_data.sql  data.sql

PostgreSQL PaaS

• Create an instance in Azure Database for PostgreSQL.
• Connect to the instance and create a database using the instruction in this document.

Import to PostgreSQL PaaS

Create .pg_azure file with the credentials to your PaaS database:

vi .pg_azure

export PGDATABASE=ora2pg
export PGHOST=name.postgres.database.azure.com
export PGUSER=user@host
export PGPASSWORD=VeryBadPractice;)

Save the .pg_azure file and load it in the session:

source .pg_azure

Create a new database:

createdb ora2pg

From the /data/myproject directory load the files:

psql -f schema/tables/table.sql

psql -f schema/sequences/sequence.sql
psql -f schema/views/view.sql 
psql -f schema/procedures/procedure.sql
psql -f schema/triggers/trigger.sql

Check if all objects were correctly created in your new postgres database.

Import data:

psql -f data/data.sql

Create constraints:

psql -f schema/tables/INDEXES_table.sql
psql -f schema/tables/CONSTRAINTS_table.sql
psql -f schema/tables/FKEYS_table.sql

Test the migration

In order to lack of DBD::Pg perl module we are not able count rows. If you install the module the command to check migration looks as follows:

ora2pg -c config/ora2pg.conf -t TEST

Online data migration

For online data migration approach please refer here

Issues

• DBD::Pg perl module isn't provided in the container. You might want to install it or use psql for import to PostgreSQL
• import_all.sh script isn't working in current version of ora2pg because of following lines: pgdsn_defined=$(grep "^PG_DSN" config/ora2pg.conf | sed 's/.*dbi:Pg/dbi:Pg/') if [ "a$pgdsn_defined" = "a" ]; then if [ "a$DB_HOST" != "a" ]; then