Penumbra events pipeline
Here's an architecture diagram, describing how ABCI Events work in Penumbra.
Software components
There are several components in play:
- pd, the Penumbra daemon
- pindexer, used to enrich event data with Penumbra-specific AppViews
- penumbra-reindexer, used to collect historical event data
- Databases in both PostgreSQL and SQLite formats
How you host these services is up you. If the application you intend to support requires event data, then you'll need to provide pindexer. Running the reindexer is optional, but will provide a more resilient service over time.