Configurar una base de datos de AlloyDB para PostgreSQL para CDC
Organízate con las colecciones
Guarda y clasifica el contenido según tus preferencias.
En esta página se describe cómo configurar la captura de datos de cambios (CDC) para transmitir datos desde una base de datos de AlloyDB para PostgreSQL a un destino admitido, como BigQuery o Cloud Storage.
Configurar la conectividad entre Datastream y AlloyDB
Para que Datastream pueda conectarse a la instancia de AlloyDB, debe configurar un proxy TCP en el proyecto de consumidor. La imagen del proxy TCP está disponible públicamente para todos los clientes.
Inicia el contenedor proxy:
Activa Cloud Shell en el proyecto del consumidor.
Ejecuta la siguiente secuencia de comandos para crear el proxy TCP y deshabilitar la red de puente para evitar problemas de conectividad:
Habilita la decodificación lógica en la instancia principal de AlloyDB. En Google Cloud, asigna el valor ON a la marca alloydb.logical_decoding. Para obtener información sobre cómo actualizar las marcas de la base de datos, consulta Configurar las marcas de la base de datos de una instancia.
Conéctate a la instancia de AlloyDB mediante la dirección IP del proxy TCP ejecutando el siguiente comando desde cualquier VM de la misma VPC:
psql-hPROXY_IP\-UDB_USER\-dDB_NAME
Haz los cambios siguientes:
PROXY_IP: la dirección IP del proxy TCP.
DB_USER: nombre de usuario de la base de datos de AlloyDB.
DB_NAME: nombre de la base de datos de AlloyDB.
Ejecuta el siguiente comando para conceder privilegios de replicación a tu usuario de la base de datos:
ALTER USER DB_USER WITH REPLICATION;
Crea una publicación. Te recomendamos que crees una publicación solo para las tablas que quieras replicar. De esta forma, Datastream puede leer los datos relevantes y se reduce la carga de la base de datos y de Datastream:
CREATE PUBLICATION PUBLICATION_NAME
FOR TABLE SCHEMA1.TABLE1, SCHEMA2.TABLE2;
Haz los cambios siguientes:
PUBLICATION_NAME: el nombre de tu publicación. Deberá proporcionar este nombre cuando cree un flujo en el asistente de creación de flujos de Datastream.
SCHEMA: el nombre del esquema que contiene la tabla.
TABLE: el nombre de la tabla que quieres replicar.
También puede crear una publicación para todas las tablas de su base de datos. Tenga en cuenta que este enfoque aumenta la carga tanto en la base de datos de origen como en Datastream:
CREATE PUBLICATION PUBLICATION_NAME FOR ALL TABLES;
Crea un slot de replicación ejecutando el siguiente comando.
REPLICATION_SLOT_NAME: el nombre de tu ranura de replicación.
Deberá proporcionar este nombre cuando cree un flujo en el asistente de creación de flujos de Datastream.
Crear un usuario de Datastream
Para crear un usuario de Datastream, introduce el siguiente comando de PostgreSQL:
CREATE USER USER_NAME WITH REPLICATION LOGIN PASSWORD 'USER_PASSWORD';
Haz los cambios siguientes:
USER_NAME: nombre del usuario de Datastream que quieras crear.
USER_PASSWORD: la contraseña de inicio de sesión del usuario de Datastream que quieras crear.
Concede los siguientes privilegios al usuario que has creado:
GRANT SELECT ON ALL TABLES IN SCHEMA SCHEMA_NAME TO USER_NAME;
GRANT USAGE ON SCHEMA SCHEMA_NAME TO USER_NAME;
ALTER DEFAULT PRIVILEGES IN SCHEMA SCHEMA_NAME GRANT SELECT ON TABLES TO USER_NAME;
Haz los cambios siguientes:
SCHEMA_NAME: el nombre del esquema al que quieres conceder los privilegios.
USER_NAME: el usuario al que quieres conceder los privilegios.
Siguientes pasos
Consulta más información sobre cómo funciona Datastream con las fuentes de PostgreSQL.
[[["Es fácil de entender","easyToUnderstand","thumb-up"],["Me ofreció una solución al problema","solvedMyProblem","thumb-up"],["Otro","otherUp","thumb-up"]],[["Es difícil de entender","hardToUnderstand","thumb-down"],["La información o el código de muestra no son correctos","incorrectInformationOrSampleCode","thumb-down"],["Me faltan las muestras o la información que necesito","missingTheInformationSamplesINeed","thumb-down"],["Problema de traducción","translationIssue","thumb-down"],["Otro","otherDown","thumb-down"]],["Última actualización: 2025-09-10 (UTC)."],[[["\u003cp\u003eThis guide explains how to configure an AlloyDB for PostgreSQL database to work with Datastream, beginning with setting up a TCP proxy for connectivity.\u003c/p\u003e\n"],["\u003cp\u003eTo enable Datastream's connection to the AlloyDB instance, a TCP proxy must be launched in the consumer project using a provided script, replacing placeholders with appropriate values.\u003c/p\u003e\n"],["\u003cp\u003eLogical decoding must be enabled on the AlloyDB primary instance, along with granting replication privileges to a designated database user via the provided command, and creating a publication for tables to be replicated.\u003c/p\u003e\n"],["\u003cp\u003eA dedicated Datastream user must be created with specific privileges, including replication login and select permissions on tables and schemas, using the provided commands.\u003c/p\u003e\n"],["\u003cp\u003eAfter completing the setup, you can use the TCP proxy IP address for the connection profile, and create a stream in the Datastream stream creation wizard using the publication and replication slot names.\u003c/p\u003e\n"]]],[],null,["# Configure an AlloyDB for PostgreSQL database\n\nThe following sections cover how to configure an AlloyDB for PostgreSQL database.\n\nSet up connectivity between Datastream and AlloyDB\n--------------------------------------------------\n\nTo enable Datastream to connect to the AlloyDB instance, you need to set up a TCP proxy in the consumer project. The [TCP proxy image](http://gcr.io/dms-images/tcp-proxy) is publicly available to all customers.\n\n1. Launch the proxy container:\n\n 1. Activate Cloud Shell in the consumer project.\n 2. Run the following script to create the TCP proxy and disable bridge networking to avoid connectivity issues:\n\n gcloud compute instances create-with-container \\\n --zone=\u003cvar translate=\"no\"\u003eREGION_ID\u003c/var\u003e \u003cvar translate=\"no\"\u003eVM_NAME\u003c/var\u003e \\\n --container-image gcr.io/dms-images/tcp-proxy \\\n --tags=dms-tcp-proxy \\\n --container-env=SOURCE_CONFIG=\u003cvar translate=\"no\"\u003eALLOYDB_IP\u003c/var\u003e:\u003cvar translate=\"no\"\u003eALLOYDB_PORT\u003c/var\u003e \\\n --can-ip-forward \\\n --network=\u003cvar translate=\"no\"\u003eSOURCE_AND_DEST_VPC\u003c/var\u003e \\\n --machine-type=\u003cvar translate=\"no\"\u003eVM_TIER\u003c/var\u003e \\\n --metadata=startup-script='#! /bin/bash\n mkdir -p /etc/docker\n cat \u003c\u003cEOF \u003e /etc/docker/daemon.json\n {\"bridge\":\"none\"}\n EOF\n systemctl restart docker'\n\n Replace the following:\n - \u003cvar translate=\"no\"\u003eREGION_ID\u003c/var\u003e: The region in which you want to create the TCP proxy.\n - \u003cvar translate=\"no\"\u003eVM_NAME\u003c/var\u003e: The name of the virtual machine.\n - \u003cvar translate=\"no\"\u003eALLOYDB_IP\u003c/var\u003e: The IP address of the AlloyDB for PostgreSQL instance.\n - \u003cvar translate=\"no\"\u003eALLOYDB_PORT\u003c/var\u003e: The port number of the AlloyDB for PostgreSQL instance.\n - \u003cvar translate=\"no\"\u003eSOURCE_AND_DEST_VPC\u003c/var\u003e: The VPC network to which the source and destination are connected.\n - \u003cvar translate=\"no\"\u003eVM_TIER\u003c/var\u003e: The type of your virtual machine.\n\n An example command with updated parameters: \n\n gcloud compute instances create-with-container \\\n --zone=us-central1-c ds-tcp-proxy \\\n --container-image gcr.io/dms-images/tcp-proxy \\\n --tags=ds-tcp-proxy \\\n --container-env=SOURCE_CONFIG=10.16.0.5:5432 \\\n --can-ip-forward \\\n --network=default \\\n --machine-type=e2-micro\n --metadata=startup-script='#! /bin/bash\n mkdir -p /etc/docker\n cat \u003c\u003cEOF \u003e /etc/docker/daemon.json\n {\"bridge\":\"none\"}\n EOF\n systemctl restart docker'\n\n | You may receive a prompt to authorize Cloud Shell. Click **AUTHORIZE**.\n2. Run the script to print the IP address of the TCP Proxy with the following inputs:\n\n gcloud compute instances describe \u003cvar translate=\"no\"\u003eVM_NAME\u003c/var\u003e \\\n --format=\"yaml(networkInterfaces[].networkIP)\" \\\n --zone=\u003cvar translate=\"no\"\u003eREGION_ID\u003c/var\u003e\n\n Replace the following:\n - \u003cvar translate=\"no\"\u003eVM_NAME\u003c/var\u003e: The name of your virtual machine.\n - \u003cvar translate=\"no\"\u003eREGION_ID\u003c/var\u003e: The region in which you created the TCP proxy.\n\n | Use this IP address in the Datastream [connection profile](/datastream/docs/create-connection-profiles#cp4postgresdb) configuration.\n3. (Optional) Create a firewall rule to limit ingress traffic to the TCP proxy:\n\n gcloud compute firewall-rules create \u003cvar translate=\"no\"\u003eFIREWALL_RULE_NAME\u003c/var\u003e \\\n --direction=INGRESS \\\n --priority=1000 \\\n --target-tags=dms-tcp-proxy \\\n --network=\u003cvar translate=\"no\"\u003eSOURCE_VPC\u003c/var\u003e \\\n --action=ALLOW \\\n --rules=tcp:\u003cvar translate=\"no\"\u003eALLOYDB_PORT\u003c/var\u003e \\\n --source-ranges=\u003cvar translate=\"no\"\u003eIP_RANGE\u003c/var\u003e\n\n Replace the following:\n - \u003cvar translate=\"no\"\u003eFIREWALL_RULE_NAME\u003c/var\u003e: The name of your firewall rule.\n - \u003cvar translate=\"no\"\u003eSOURCE_VPC\u003c/var\u003e: The VPC network to which your source is connected.\n - \u003cvar translate=\"no\"\u003eALLOYDB_PORT\u003c/var\u003e: The port number of the AlloyDB for PostgreSQL instance.\n - \u003cvar translate=\"no\"\u003eIP_RANGE\u003c/var\u003e: The IP address range to which you want to limit ingress traffic.\n\n An example command to allow ingress traffic from all IP addresses: \n\n gcloud compute firewall-rules create ds-proxy1 \\\n --direction=INGRESS \\\n --priority=1000 \\\n --target-tags=ds-tcp-proxy \\\n --network=default \\\n --action=ALLOW \\\n --rules=tcp:5432\n\nConfigure AlloyDB for replication\n---------------------------------\n\n1. Enable logical decoding for the AlloyDB primary instance. In Google Cloud, set\n the value for the `alloydb.logical_decoding` flag to **ON** . For information about\n updating the database flags, see [Configure an instance's database flags](/alloydb/docs/instance-configure-database-flags).\n\n | **Note:** Updating the `alloydb.logical_decoding` flag restarts the instance.\n2. Connect to the AlloyDB instance using the TCP proxy IP address by running the\n following command from any VM on the same VPC:\n\n psql -h \u003cvar translate=\"no\"\u003ePROXY_IP\u003c/var\u003e \\\n -U \u003cvar translate=\"no\"\u003eDB_USER\u003c/var\u003e \\\n -d \u003cvar translate=\"no\"\u003eDB_NAME\u003c/var\u003e\n\n Replace the following:\n - \u003cvar translate=\"no\"\u003ePROXY_IP\u003c/var\u003e: The IP address of the TCP proxy.\n - \u003cvar translate=\"no\"\u003eDB_USER\u003c/var\u003e: The username for the AlloyDB database.\n - \u003cvar translate=\"no\"\u003eDB_NAME\u003c/var\u003e: The AlloyDB database name.\n3. Run the following command to grant replication privileges to your database user:\n\n ```\n ALTER USER DB_USER WITH REPLICATION;\n ```\n4. Create a publication. We recommend that you create a publication only for the\n tables that you want to replicate. This allows Datastream to read-only\n the relevant data, and lowers the load on the database and Datastream:\n\n ```\n CREATE PUBLICATION PUBLICATION_NAME\n FOR TABLE SCHEMA1.TABLE1, SCHEMA2.TABLE2;\n ```\n\n Replace the following:\n - \u003cvar translate=\"no\"\u003ePUBLICATION_NAME\u003c/var\u003e: The name of your publication. You'll need to provide this name when you create a stream in the Datastream stream creation wizard.\n - \u003cvar translate=\"no\"\u003eSCHEMA\u003c/var\u003e: The name of the schema that contains the table.\n - \u003cvar translate=\"no\"\u003eTABLE\u003c/var\u003e: The name of the table that you want to replicate.\n\n You can also create a publication for all tables in your database. Note that\n this approach increases the load on both the source database and Datastream: \n\n ```\n CREATE PUBLICATION PUBLICATION_NAME FOR ALL TABLES;\n ```\n | For PostgreSQL 15 and later, you can create a publication for all tables in a schema. This approach lets you replicate changes for tables in the specified list of schemas, including tables that you create in the future: \n |\n | ```\n | CREATE PUBLICATION PUBLICATION_NAME\n | FOR TABLES IN SCHEMA SCHEMA1, SCHEMA2;\n | ```\n5. Create a replication slot by executing the following command.\n\n ```\n SELECT PG_CREATE_LOGICAL_REPLICATION_SLOT('REPLICATION_SLOT_NAME', 'pgoutput');\n ```\n\n Replace the following:\n - \u003cvar translate=\"no\"\u003eREPLICATION_SLOT_NAME\u003c/var\u003e: The name of your replication slot. You'll need to provide this name when you create a stream in the Datastream stream creation wizard.\n\nCreate a Datastream user\n------------------------\n\n1. To create a Datastream user, enter the following PostgreSQL command:\n\n ```\n CREATE USER USER_NAME WITH REPLICATION LOGIN PASSWORD 'USER_PASSWORD';\n ```\n\n Replace the following:\n - \u003cvar translate=\"no\"\u003eUSER_NAME\u003c/var\u003e: The name of the Datastream user that you want to create.\n - \u003cvar translate=\"no\"\u003eUSER_PASSWORD\u003c/var\u003e: The login password for the Datastream user that you want to create.\n2. Grant the following privileges to the user you created:\n\n ```\n GRANT SELECT ON ALL TABLES IN SCHEMA SCHEMA_NAME TO USER_NAME;\n GRANT USAGE ON SCHEMA SCHEMA_NAME TO USER_NAME;\n ALTER DEFAULT PRIVILEGES IN SCHEMA SCHEMA_NAME GRANT SELECT ON TABLES TO USER_NAME;\n ```\n\n Replace the following:\n - \u003cvar translate=\"no\"\u003eSCHEMA_NAME\u003c/var\u003e: The name of the schema to which you want to grant the privileges.\n - \u003cvar translate=\"no\"\u003eUSER_NAME\u003c/var\u003e: The user to whom you want to grant the privileges.\n\n | The first command might display warnings, however it is safe to ignore them.\n\nWhat's next\n-----------\n\n- Learn more about how Datastream works with [PostgreSQL sources](/datastream/docs/sources-postgresql)."]]