Job run failed to shutdown

[jhamman]

I ran a job over the weekend (run-1b450464 in the earthmover-devs workspace). This job was started with --interactive and was terminated locally but kept running on coiled. Fortunately, I checked the dashboard on and saw that it was still running.

Is there someway to put a max lifetime or idle cull behavior on jobs like this?

[jrbourbeau]

Thanks for the issue @jhamman (sorry for the slow response here).

was terminated

Joe mentioned offline that this was due to a blip in his network

@ntabris I'm wondering if you have any thoughts on this one?

[ntabris]

Hm, we could have something on the VM handle when then the SSH session exits. Not sure how trivial/hard this would be.