Pipeline failure due to replication slot errors

Last updated on Feb 19, 2024
Applies To PostgreSQL Source (all variants) Pipelines with Logical Replication mode.
Error Message Text(s) - Logical replication failed due to WAL expiry. Click FIX NOW to resume the data ingestion.
- Unable to read data. Replication slot is stuck at IO event WALRead. Click FIX NOW to resume the data ingestion.
- The PostgreSQL replication slot used by the Pipeline has become corrupt. Click FIX NOW to resume the data ingestion.

Potential Causes

  • Hevo is unable to read the Write-Ahead Logs (WALs). This can happen due to the following reasons:

    • The WAL files have expired. This can happen if you have a sudden spike in Events which causes the WALs to expire sooner than expected, and Hevo is unable to replicate data to the Destination.

    • The PostgreSQL replication slot used by the Pipeline is not able to complete the IO event WALRead, and is waiting for data to be read from the WAL files.

    • The PostgreSQL replication slot has become corrupt.

Suggested Action(s)

  1. In the error displayed in the Pipelines Detailed View, click FIX NOW.

    Click FIX NOW

  2. Do one of the following:

    • Recover the lost Events by ingesting them as historical data. To do this:

      1. Enable the Run Historical Load option.

      2. Select the objects you want to restart the historical load for.

        Run Historical Load enabled

      3. Click OK, FIX IT.

    • Skip the expired Events and read the latest available WAL files. To do this:

      1. Disable the Run Historical Load option.

        Run Historical Load disabled

      2. Click OK, FIX IT.


Revision History

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

Date Release Description of Change
Sep-05-2022 NA Brought this page under its respective Source documentation folder.
Jun-21-2022 1.91 New document.

Tell us what went wrong