Connect to Redis
Use the project-network endpoint or the external TLS endpoint.
Retrieve credentials
Open the instance's Connection page. It shows Project network and, when enabled, External TLS endpoint rows and the last rotation time. Each connection string is hidden initially. Use the eye button at the end of a string to show or hide only that row, or use its copy button to copy the full connection string. The username and password are included in the URI; separate credential fields are not shown.
Use the managed username and password together. The private endpoint requires connectivity to the instance's project network; it is not an internet address.
New instances support a maximum of 100 simultaneous connections, including connections used to operate and monitor the service. Reuse connections through an application connection pool. Check Overview metrics for your instance's current limit; older instances may have a different limit.
Test a private connection
Run redis-cli from a client that can reach the private endpoint. Replace the placeholders with the values from Connection:
redis-cli -h PRIVATE_HOST -p 6379 --user USERNAME --askpass PINGEnter the password when prompted. Use the actual port shown in the Console.
Test an external connection
Enable the external endpoint, then use a TLS-capable client with its hostname and port:
redis-cli --tls --sni EXTERNAL_HOST -h EXTERNAL_HOST -p EXTERNAL_PORT --user USERNAME --askpass PINGConfigure an appropriate CA trust source for certificate validation if your client does not already have one. Do not replace the hostname with a raw IP address or disable TLS to work around a connection failure.
A successful test returns PONG. The Redis CLI documentation describes client connection options.
Configure an application
Use the connection string for the intended endpoint. External connections use TLS; preserve that requirement in your client settings. Keep credentials in application secrets and update them after credential rotation.