Hi. I'm working on a tap for Chargify data. Looking for some tips to debug it. It's looping through pages of data and sending a data and state message after each request. The main thing it's reading is the events data from this endpoint. It's using the event id as the state. I ran it and it hit a timeout on my batch job. The next time it ran, I found that the id in its state was much lower than the last state written to my warehouse (Redshift). I'd like to know how Meltano handles state during a run (using Postgres state backend). I've noticed that the "run" table seems to have state values populated when taps complete. Does it store state during a run? Can it recover from this? Any documentation on this subject I can look at? BTW I'm writing this in TypeScript, in case that's relevant.

This section of the docs might be helpful: https://docs.meltano.com/guide/integration/#incremental-replication-state

Thanks. That's helpful.

Correct. Meltano reads the target's stdout for STATE message, which according to the spec the target is supposed to emit after it commits data to the dest system