thanks @lucasbradstreet. I may have to consume from kafka, so resume points would let me pick up where i left off in the queue? if so may need that

ZK and job would be in the same jar though, yes.

Yeah, it would be pretty difficult to make it be able to recover kafka checkpoints when the processes die then. Seems like a no go

ah ok

i suppose i can do initial tests with an embedded ZK and then move to an actual cluster if onyx looks like a good fit

> When you’re ready to stand your jar up, simply uberjar and start executing on the target machine. And i could theoretically push an uberjar that, on startup, kills all current jobs, migrates the database, and resumes all jobs? Thanks.