I have a three-step pipeline I run with

meltano run

- extractor, mapper, loader. A few days ago it failed due to some bad data. No problem, easy fix. However, now I can't run the pipeline without including the --force switch because, according to the DB, the job that failed is still running. Shouldn't the --force switch clean that up? I thought that's how things worked with

meltano elt

but maybe I'm just misremembering. Regardless, "you need to tweak the data in the Meltano DB" can't be the real answer here, can it? This feels like a standard enough scenario that manually editing the DB shouldn't be required to address it... at least to me.

That sounds like a bug.. can you make an issue so we can document and pull some engineers in. It’d be helpful too if you can share the DB row as well if possible.

Will do. I'm curious tough re: the state command - since job_ids are auto-generated by the run command, how will we know which job_id to pass to the state command? 🤔

The current format for job_id is

<env>:<tap-name>-to-<target-name>

. we have an option issue to make it easy to override though

Thanks. I guess I conflated "auto-generated" with "don't try to understand" 😛