Replicate data from PostgreSQL to Snowflake

Learn to replicate data from PostgreSQL to Snowflake using OCI GoldenGate.

Before you begin

To successfully complete this quickstart, you must have the following:

A PostgreSQL installation to serve as the source database (See Environment setup)
Open port 5432 in your VCN's security list
A Snowflake database to serve as the target database.

Environment set up: PostgreSQL

To set up the environment for this Quickstart:

Run the following commands to install PostgreSQL.
1. Install PostgreSQL server:
```
sudo yum install postgresql-server
```
2. Install postgresql-contrib module to avoid this SQL exception:
```
sudo yum install postgresql-contrib
```
3. Create a new PostgreSQL database cluster:
```
sudo postgresql-setup --initdb
```
4. Enable the postgresql.service:
```
sudo systemctl enable postgresql.service
```
5. Start the postgresql.service:
```
sudo systemctl start postgresql.service
```
By default, PostgreSQL only allows local connections. Allow remote connectivity to PostgreSQL.
1. In /var/lib/pgsql/data/postgresql.conf, prepare the database for replication.
2. Locate and uncomment listen_addresses = 'localhost' and change localhost to an asterisk (`):
```
listen_addresses = '*'
```
3. Set the following parameters as follows:
  - wal_level = logical
  - max_replication_slots = 1
  - max_wal_senders = 1
  - track_commit_timestamp = on
  Note:
  
  Configure /var/lib/pgsql/data/pg_hba.conf to ensure that client authentication is set to allow connections from an Oracle GoldenGate host. For example, add the following:
```
#Allow connections from remote hosts

host all all 0.0.0.0/0 md5
```
  See The pg_hba.conf File for more information.
  1. Restart PostgreSQL server:
```
sudo systemctl restart postgresql.service
```
  1. If using Oracle Cloud Compute to host PostgreSQL, open port 5432:
```
sudo firewall-cmd --permanent --add-port=5432/tcp
sudo firewall-cmd --reload
sudo firewall-cmd --list-all
```
4. Open port 5432 in your VCN's security list.
5. Connect to PostgreSQL.
```
> sudo su - postgres
> psql
```
  Note: Alternatively, you can enter sudo su - postgres psql if the above example doesn’t work.
6. Set up PostgreSQL.
  1. Download and run seedSRCOCIGGLL_PostgreSQL.sql to set up the database and load the sample data.
  2. Run the following commands to set up the user (ensure you replace <password> with an actual password):
```
create user ggadmin with password '<password>';
alter user ggadmin with SUPERUSER;
GRANT ALL PRIVILEGES ON DATABASE ociggll TO ggadmin;
```

Environment set up: Snowflake

Create a GoldenGate user in Snowflake with appropriate privileges.
Create target tables using sample schema.

Ensure the tables and user have been successfully created.

Task 1: Create the OCI GoldenGate resources

This quickstart example requires deployments and connections for both the source and target.

Create a deployment for the source PostgreSQL database.
Create a Big Data deployment for the target Snowflake database.
Create a PostgreSQL connection with the following values:
1. For Type, select PostgreSQL Server from the dropdown.
2. For Database name, enter ociggll.
3. For Host, enter the public IP of the Compute instance that PostgreSQL runs on.
4. For Port, enter 5432.
5. For Username, enter ggadmin.
6. For Password, enter your password.
7. For Security Protocol, select Plain from the dropdown.
Create a Snowflake connection with the following values:
1. For Connection URL, enter jdbc:snowflake://<account_identifier>.snowflakecomputing.com/?warehouse=<warehouse name>&db=OCIGGLL.
  
  Note: Ensure you replace <account_identifier> and <warehouse name> with the appropriate values.
2. For Authentication Type, select Basic authentication from the dropdown.
3. For Username, enter a name.
4. For Password, enter a password.
(Optional) If your Big Data deployment doesn't have a public endpoint, then create a connection to GoldenGate, and then assign this connection to the source PostgreSQL deployment.
Assign the source PostgreSQL connection to the PostgreSQL deployment.
Assign the Snowflake connection to the target Big Data deployment.

Task 2: Enable supplemental logging

To enable supplemental logging:

Launch the PostgreSQL GoldenGate deployment console:
1. From the Deployments page, select the PostgreSQL deployment to view its details.
2. On the PostgreSQL deployment details page, select Launch console.
3. On the deployment console sign in page, enter the GoldenGate admin credentials provided in Task 1, step 1.
  
  Note: Sign in is required if IAM wasn’t selected as the credential store when creating a deployment.
In the deployment console, select DB Connections in the left navigation, then the source PostgreSQL database, and then Trandata.
On the Trandata page, next to TRANDATA Information, select Add Trandata (plus icon).
In the Trandata panel, for Schema Name, enter src_ociggll, and then select Submit.

Note: Use the search field to search for src_ociggll and verify the tables were added.

Task 3: Create the Extracts

Add the Change Data Capture Extract:
1. In the left navigation, select Extracts,
2. On the Extracts page, select Add Extract (plus icon), and then complete the fields as follows:
  - On the Extract Information page:
    1. For Extract type, select Change Data Capture Extract.
    2. For Process Name, enter a name for the Extract, such as ECDC.
    3. Select Next.
  - On the Extract Options page:
    1. For Source credentials, select Oracle GoldenGate from the Domain dropdown
    2. Select the source PostgreSQL database from the Alias dropdown.
    3. For Extract Trail Name, enter a two-character trail name, such as C1.
    4. Select Next.
  - On the Extract Parameters page, replace MAP *.*, TARGET *.*; with the following:
```
TABLE SRC_OCIGGLL.*;
```
3. Select Create and Run.
Add the Initial Load Extract:
1. On the Extracts page, select Add Extract, and then complete the Add Extract form as follows:
  - On the Extract Information page:
    1. For Extract type, select Initial Load Extract.
    2. For Process Name, enter a name, such as EIL.
    3. Select Next.
  - On the Extract Options page:
    1. For Source credentials, select Oracle GoldenGate from the Domain dropdown.
    2. Select the PostgreSQL database from the Alias, dropdown.
    3. For Extract Trail Name, enter a two-character trail name, such as I1.
    4. Select Next.
  - On the Extract Parameters page, replace the contents of the textarea with the following:
```
EXTRACT EIL
USERIDALIAS PostgreSQL_Compute, DOMAIN OracleGoldenGate
EXTFILE I1, PURGE
TABLE src_ociggll.*;
```
    Note: Ensure that you remove the SOURCEDB parameter in front of USERIDALIAS before you move on.
2. Select Create and Run.

You return to the Extracts page, where you can observe the Extract starting.

Task 4: Create the Distribution Path for Initial Load Extract

To create a Distribution Path for Initial Load Extract:

In the Oracle Cloud console, on the Deployments page, select the target Big Data deployment.
On the deployment details page, select Launch Console, and then log in as the admin user.
If using IAM credential store, proceed to the Create a Distribution Path step. If using GoldenGate credential store, create a user with which the source GoldenGate uses to connect to the target GoldenGate.
1. In the navigation menu, select User Administration.
2. Select Add New User (plus icon), complete the fields as follows, and then select Submit:
  - For Username, enter ggsnet.
  - For Role, select Operator.
  - Enter the password twice for verification.
In the source PostgreSQL deployment console, create a Path Connection for the user created in the previous step.
1. In the navigation menu, select Path Connections.
2. Select Add Path Connection (plus icon), complete the fields as follows, and then select Submit:
  - For Credential Alias, enter dpuser.
  - For User ID, enter ggsnet
  - For Password, enter the same password used in the previous step.
Create a Distribution Path.
1. In the service menu bar, select Paths, and then select Add Distribution Path (plus icon).
2. Complete the Add Path form as follows:
  - On the Path Information page:
    1. For Path Name, enter a name for this path.
    2. Select Next.
  - On the Source Options page:
    1. For Source Extract, leave blank.
    2. For Trail Name, enter the Initial Load Extract trail name (I1).
    3. Select Next.
  - On the Target Options page:
    1. For Target, select wss.
    2. For Target Host, enter the target deployment URL, without the https:// or any trailing slashes.
    3. For Port Number, enter 443.
    4. For Trail Name, enter I1.
    5. For Target Authentication Method, select UserID Alias.
    6. For Domain, enter the domain name created in the previous step.
    7. For Alias, enter the alias created in the previous step (dpuser).
    8. Select Next.
3. Select Create and Run.
You return to the Distribution Service page where you can review the path created.
In the target deployment console, review the Receiver path created as a result of the Distribution path:
1. Select Paths.
2. Review the Path details.

Task 5: Add the Replicat for Initial Load

In the target Big Data deployment console, add the Initial Load Replicat.
1. In the navigation menu, select Replicats, and then select Add Replicat (plus icon).
2. On the Replicats page, then complete the Add Replicat fields as follows:
  1. On the Replication Information page:
    1. For Replicat type, select Classic or Coordinated.
    2. For Process Name, enter a name, such as RIL.
    3. Select Next.
  2. On the Replicat Options page:
    1. For Replicat Trail Name, enter the name of the Trail from Task 2 (I1).
    2. For Target Credentials, select the Domain and Alias for the Snowflake connection.
    3. For Available aliases, select an alias from the dropdown, such as Snowflake.
    4. (Optional) Enable external storage to select an available staging location from the dropdown.
    5. Select Next.
  3. On the Parameter File page, add the following mapping:
```
INSERTALLRECORDS
MAP src_ociggll.src_city, TARGET SRCMIRROR_OCIGGLL.SRC_CITY;
MAP src_ociggll.src_region, TARGET SRCMIRROR_OCIGGLL.SRC_REGION;
MAP src_ociggll.src_customer, TARGET SRCMIRROR_OCIGGLL.SRC_CUSTOMER;
MAP src_ociggll.src_orders, TARGET SRCMIRROR_OCIGGLL.SRC_ORDERS;
MAP src_ociggll.src_order_lines, TARGET SRCMIRROR_OCIGGLL.SRC_ORDER_LINES;
MAP src_ociggll.src_product, TARGET SRCMIRROR_OCIGGLL.SRC_PRODUCT;
```
  4. On the Properties page, review the properties, and then select Create and Run.
    
    You return to the Replicats page, where you can review the Replicat details.

To verify the Initial Load, connect to Snowflake database and run following queries:

select * from SRCMIRROR_OCIGGLL.SRC_CITY;
select * from SRCMIRROR_OCIGGLL.SRC_CUSTOMER;

The output should return the data that was loaded into the target database tables as a result of the Initial Load.

Task 6: Create the Distribution Path for Change Data Capture

In the source PostgreSQL deployment console, select Distribution Service.
Select Add Distribution Path.
Complete the Add Path form as follows:
1. On the Path Information page:
  1. For Path Name, enter a name.
  2. Select Next.
2. On the Source Options page:
  1. For Source Extract, select the Change Data Capture Extract (ECDC).
  2. For Trail Name, select the Change Data Capture trail file (C1).
  3. Select Next.
3. On the Target Options page:
  1. For Target, select wss.
  2. For Target Host, enter the target deployment console URL (you can find this on the deployment details page, without the https:// or any trailing slashes).
  3. For Port Number, enter 443.
  4. For Trail Name, enter C1.
  5. For Target Authentication Method, select UserID Alias.
  6. For Domain, enter the domain name.
  7. For Alias, enter the alias.
4. Select Create Path and Run.
In the target Big Data deployment console, select Receiver Service, and then review the Receiver path created.

Task 7: Add a Replicat for Change Data Capture

Perform updates to the source PostgreSQL database to verify replication to Snowflake.

Add the Replicat.
1. In the target Big Data deployment console, select Administration Service, and then in the navigation menu, select Replicats.
2. On the Replicats page, select Add Replicat (plus icon), and then complete the Add Replicat form as follows:
  - On the Replicat Information page:
    1. For Replicat type, select Classic or Coordinated.
    2. For Process Name, enter a name, such as RCDC.
    3. Select Next.
  - On the Replicat Options page:
    1. For Replicat Trail Name, enter the name of the Trail from Task 3 (C1).
    2. For Target Credentials, select the Domain and Alias for the Snowflake connection.
    3. (Optional) Enable external storage to select an available staging location from the dropdown.
  - On the Parameter Files page, add the following mapping, and then select Next:
```
MAP src_ociggll.src_city, TARGET SRCMIRROR_OCIGGLL.SRC_CITY;
MAP src_ociggll.src_region, TARGET SRCMIRROR_OCIGGLL.SRC_REGION;
MAP src_ociggll.src_customer, TARGET SRCMIRROR_OCIGGLL.SRC_CUSTOMER;
MAP src_ociggll.src_orders, TARGET SRCMIRROR_OCIGGLL.SRC_ORDERS;
MAP src_ociggll.src_order_lines, TARGET SRCMIRROR_OCIGGLL.SRC_ORDER_LINES;
MAP src_ociggll.src_product, TARGET SRCMIRROR_OCIGGLL.SRC_PRODUCT;
```
  - On the Properties page, review the properties, and then select Create and Run.
    
    You return to the Replicats page, where you can review the Replicat details.

Verify Change Data Capture:

Perform updates to the source PostgreSQL database to verify replication to Snowflake. Run the following script to perform inserts into the PostgreSQL database:

Insert into src_ociggll.src_city (CITY_ID,CITY,REGION_ID,POPULATION) values (1000,'Houston',20,743113);
Insert into src_ociggll.src_city (CITY_ID,CITY,REGION_ID,POPULATION) values (1001,'Dallas',20,822416);
Insert into src_ociggll.src_city (CITY_ID,CITY,REGION_ID,POPULATION) values (1002,'San Francisco',21,157574);
Insert into src_ociggll.src_city (CITY_ID,CITY,REGION_ID,POPULATION) values (1003,'Los Angeles',21,743878);
Insert into src_ociggll.src_city (CITY_ID,CITY,REGION_ID,POPULATION) values (1004,'San Diego',21,840689);
Insert into src_ociggll.src_city (CITY_ID,CITY,REGION_ID,POPULATION) values (1005,'Chicago',23,616472);
Insert into src_ociggll.src_city (CITY_ID,CITY,REGION_ID,POPULATION) values (1006,'Memphis',23,580075);
Insert into src_ociggll.src_city (CITY_ID,CITY,REGION_ID,POPULATION) values (1007,'New York City',22,124434);
Insert into src_ociggll.src_city (CITY_ID,CITY,REGION_ID,POPULATION) values (1008,'Boston',22,275581);
Insert into src_ociggll.src_city (CITY_ID,CITY,REGION_ID,POPULATION) values (1009,'Washington D.C.',22,688002);

In the source PostgreSQL deployment console, select the RCDC, and then select Statistics. Verify that src_ociggll.src_city has 10 inserts.

Note: If the Extract captured no inserts, then restart the ECDC Extract.
In the target Big Data deployment console, select the RCDC, review its Details and Statistics to verify the number of Inserts.