When you create your first Aiven for PostgreSQL service you might be surprised to see the disk usage gradually increasing and increasing even though you are not yet inserting or updating much data.

This is completely normal within the first 24 hours of operation of an Aiven for PostgreSQL service because of our WAL archiving settings.

To prevent data loss due to node failure on Hobbyist and Startup plans we set the archive_timeout configuration to write WAL segments to disk at regular intervals which are backed up to cloud storage. Even if your service is idle, each WAL segment occupies the same amount of space on disk which is why the disk usage grows steadily over time.

After the first 24 hours of service, old WAL segments will begin to be archived and deleted from disk to free up space. From this point onward the disk usage will no longer be dominated by new WAL segments as it reaches a steady state for low traffic services.

You can read more about WAL archiving here.

Did this answer your question?