We are very pleased to announce version 0.3.2 of the Snowplow Postgres Loader
The biggest change is that the loader is better behaved at checkpointing outstanding events when it terminates. This means if you ever stop and restart the loader (or if it ever crashes and restarts) then it will pick up again much closer to where it left off processing in the enriched stream. This mitigates the problem of duplicate events landing in your database.
The Snowplow postgres loader is shaping up to be a very credible loading solution for low-volume pipelines (<500 events per second). We don’t use it for high-volume pipelines because it was never designed for horizontal scaling, but we are pleased at how it provides an entry-level solution for newcomers to the Snowplow ecosystem.
Upgrading to 0.3.2
If you are already using a 0.3.x version then you can simply start using the new docker image without making any other changes
docker pull snowplow/snowplow-postgres-loader:0.3.2
Our docs site has a complete guide for running and configuring the loader.