# Egress Private Service Connect Endpoint Setup: MongoDB CDC Source on Google Cloud for Confluent Cloud
The MongoDB CDC Source connector for Confluent Cloud supports Google Cloud Private
Service Connect through Egress Private Service Connect Endpoints. Use
this guide to configure the connector and set up the required endpoints.
## Prerequisites
The following is a list of prerequisites for configuring the MongoDB CDC Source
connector with an Egress Private Service Connect Endpoint:
* A Confluent Cloud Dedicated cluster is running with Google Cloud Private Service Connect.
* A Dedicated MongoDB Atlas database is running within the same region and cloud as
the Confluent Cloud cluster.
* [Confluent CLI](https://docs.confluent.io/confluent-cli/current/install.html#) is
installed.
* [jq](https://jqlang.github.io/jq/download/) is installed.
## Step 1. Create a MongoDB private endpoint
In MongoDB, create a private endpoint for the Dedicated MongoDB Atlas database
you want to access:
1. Obtain the Confluent Cloud Project ID.
1. In the Confluent Cloud Console, select the associated Private Service Connect
network you want to use in the **Network manage** menu in your
environment.
2. Click **+ Create endpoint** in the **Egress connections** tab.
3. Click the service you want to connect to, specifically, **MongoDB**.
4. Make a note of Confluent’s Google Cloud Account ID associated with the Private
Service Connect Endpoint.
5. Close the workflow for now by clicking **Cancel**. Complete the remaining
configuration in a later step.
2. In the MongoDB Atlas console, go to **Network Access** → **Private Endpoint**.
3. Click **Add Private Endpoint**.
4. Select the associated cloud provider **GCP**, and click **Next**.
5. Select the associated region where your Confluent Cloud cluster resides, and click
**Next**.
6. In the **Google Cloud Project ID** field, specify the Confluent Cloud Project ID
that you retrieved in Step 1 of this section.
7. To enable the **Next** button, populate the other fields, **VPC Name**,
**Subnet Name**, and **Private Service Connect Endpoint Prefix**, with
placeholder text, for example, “x”.
8. Click **Next**, and retrieve the Target Service Attachment URI.
## Step 2. Create an Egress Private Service Connect Endpoint
MongoDB Atlas requires the creation of 50 Private Service Connect endpoints to
support its internal sharding and load-balancing architecture. Each endpoint
maps to a potential MongoDB node. This is a MongoDB Atlas requirement, not a
Confluent Cloud limitation. Due to this high amount, the Confluent CLI is used
to automate the creation of these endpoints instead of manually creating each
one through the console.
1. Retrieve the gateway ID associated with the Confluent Cloud network
1. Log in to Confluent Cloud using Confluent CLI:
```bash
confluent login
```
2. Get the network ID of associated the Confluent Cloud network you are using:
```bash
confluent network list
```
3. Describe the Confluent Cloud network to get the gateway ID of the network ID you retrieved
in the previous step:
```bash
confluent network describe
```
2. Create the endpoints with the following command. Use the Confluent Cloud
gateway ID you retrieved in the previous step and the MongoDB target service
attachment URI you retrieved in the last step in
[Step 1. Create a MongoDB private endpoint](#cc-mongodb-cdc-source-eap-gcp-create-endpoint). The URI is in the
`projects<...>-"$i"` pattern.
```bash
for i in {0..49}
do
confluent network access-point private-link egress-endpoint \
create mongo-"$i" --cloud gcp \
--gateway \
--service
sleep 1
done
```
3. Generate a JSON file containing a list of the IP addresses and Private Service
Connect endpoint names to be uploaded to MongoDB Atlas. Make sure to update
the gateway ID.
#### NOTE
The filtering is done with jq because the CLI does not support filtering
by gateway.
```bash
confluent network access-point private-link egress-endpoint list --output json \
| jq '.[] | \
select ( (.name | contains("mongo")) and select (.gateway | contains("")) ) | \
{IPAddress: .gcp_private_service_connect_endpoint_ip_address, name: .gcp_private_service_connect_endpoint_name}' | \
jq -s > atlas.json
```
## Step 3. Configure the MongoDB private endpoint
Finish creating the private endpoint for your Dedicated MongoDB Atlas database.
1. Go back to the MongoDB console and click on **Next** to move to the
**Finalize** step of the private endpoint creation workflow.
2. In the **Finalize** step, upload the JSON file containing a list of IP
addresses and Private Service Connect endpoint names. You created this file
in the last step of
[Step 2. Create an Egress Private Service Connect Endpoint](#cc-mongodb-cdc-source-eap-gcp-create-egress-psc-endpoint) section.
3. When the Endpoint Status changes to “Available”, you are ready to create the
Confluent Cloud MongoDB CDC Source Connector.
This can take several minutes as endpoint creation must be completed on the
Confluent Cloud side and then get accepted on the MongoDB Atlas side.
## Step 4. Create the MongoDB CDC Source connector
1. Specify the authentication details for MongoDB.
1. In the MongoDB Atlas console, click **Database** under **Deployment**.
2. Click **Connect** for the associated database.
3. Select the **Private Endpoint** connection type, and select the endpoint
you created in Step 3.
4. Click **Choose a connection method**.
5. Select **Shell** in the **Access your data through tools** section.
6. Get the endpoint.
7. In Confluent Cloud, specify the MongoDB credentials during the connector
creation steps.
In the **MongoDB Connection String** field, specify the endpoint you retrieved in
the previous step.
2. Follow the steps to create the [MongoDB CDC Source connector](cc-mongodb-cdc-source-debezium.md#cc-mongodb-source-cdc-debezium-configure-connector)
in Confluent Cloud.