hey @U050S5ZET we haven't documented the exact protocol requirements, but you can basically look at any of the implementations (JDBC, filesystem, Kafka) to get a feel for what's involved. You can also run isolated tests for those protocols to validate your implementations. What backend storage are you considering?

Thanks. Mostly just curious. But i’m also curious why Dynamodb isn’t an option for the tx log or index, since S3 is supported for doc storage.

My assumption is that it’s been looked at but its characteristics do not fit the requirements?

Ah just saw the other thread Martynas that answers my question 😅