Managing Replication Slot when Disabling, Enabling, or Deleting a Pipeline

Deleting a Pipeline does not remove its replication slot. To prevent WAL accumulation, manually drop the replication slot after deleting the Pipeline.

Perform the following steps to drop the replication slot:

1. Connect to your PostgreSQL database instance as a user with the REPLICATION privilege using any SQL client tool, such as psql.

2. Run the following command to drop the replication slot:

   SELECT pg_drop_replication_slot('<configured_slot_name>')
   

   Note: Replace <configured_slot_name> with the replication slot configured for your Source. The slot name is displayed in the Delete Pipeline dialog.

   

   You can also find the replication slot in the Pipeline Setup tab of the Pipeline Detailed View.

When you disable a Pipeline, its replication slot remains active and continues retaining WAL files. Over time, this can consume significant disk space and affect the performance of your Source database.

The action you need to take depends on how long you intend to keep the Pipeline disabled:

• If you are disabling the Pipeline for a short period, no action is needed.

• If you are disabling the Pipeline for an extended period, perform the following steps:

  1. Connect to your PostgreSQL database instance as a user with the REPLICATION privilege using any SQL client tool, such as psql.

  2. Run the following command to drop the replication slot:

     SELECT pg_drop_replication_slot('<configured_slot_name>')
     

     Note: Replace <configured_slot_name> with the replication slot configured for your Source. The slot name is displayed in the Disable Pipeline dialog.

     

     You can also find the replication slot in the Pipeline Setup tab of the Pipeline Detailed View.

Note: Before re-enabling the Pipeline, follow the steps in the Enabling a Pipeline section to recreate the replication slot to ensure uninterrupted data ingestion.

Before enabling a Pipeline, ensure that the replication slot configured for your Source exists in your PostgreSQL database. If the slot was dropped, perform the following steps to recreate it before enabling the Pipeline:

1. Connect to your PostgreSQL database instance as a user with the REPLICATION privilege using any SQL client tool, such as psql.

2. Run the following command to create a replication slot:

   SELECT * FROM pg_create_logical_replication_slot('<configured_slot_name>', 'pgoutput')
   

   Note:

   • Replace <configured_slot_name> with the replication slot name configured for your Source. The slot name is displayed in the Enable Pipeline dialog.

     

     You can also find the replication slot in the Pipeline Setup tab of the Pipeline Detailed View.

   • If you create the slot with a different name, modify the Source configuration with the new slot before enabling the Pipeline.

A replication slot tracks WAL changes only from the point of its creation. Data changes that occurred in the Source database while the slot did not exist are not captured automatically. If you recreated the slot before enabling the Pipeline, you must resync the Pipeline to replicate any data changes that occurred while it was disabled.

Refer to the following table for the list of key updates made to this page: