When `meltano elt` is interrupted, leaves the job in the `RUNNING` state. Among other things, this causes the UI to always report the task as running. For example: ```bash meltano elt --job_id test tap-covid-19 target-jsonl --transform=skip & sleep 30 kill $(ps ux | grep meltano | grep test | awk '{print $2}') ``` | id | job\_id | state | started\_at | ended\_at | payload | payload\_flags | run\_id | trigger | | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | | 10 | test | RUNNING | 2020-07-17 22:50:13.724255 | NULL | {} | 0 | 74cf1aa5-50c2-4215-a2c2-19f72718e4a5 | cli | I tried several signals, including `SIGTERM`, `SIGHUP`, `SIGINT`, `SIGQUIT`, `SIGUSR1` and they all behaved the same way. Also, somewhat related to #2147, I never saw the state update on the running job. This may be the current intentional behavior, but I might expect Meltano to store the state for running jobs as well as successful and failed jobs.