The cloud-native orchestrator for the whole development lifecycle, with integrated lineage and observability.

dagster

Hi guys, I've just reconfigured my instance to use postgre for storage, using env var references to supply the credentials in dagster.yaml. Everything went very smoothly, but now my schedules are failing silently. I'm 99% sure the issue that these new variables dont exist for the cron user, but I'm unsure on how to configure this properly and could use some guidance.

there is an `environment_vars` arg on `ScheduleDefinition` to get the env vars in to the cron user env

&gt; are failing silently
ya this is not great- there should be a link to the log file that the bash script in the cron tab is set to dump in to somewhere in the scheduler UI, buried somewhere

but since we cant load the instance we can’t write any error events out

maybe this all could be handled by some sort of credential manager?

or for a more quick fix, copying all vars in dagster.yaml to schedules by default?

ran into a bug in the generated .sh files

```/opt/dagster/app$ /opt/dagster/dagster_home/schedules/scripts/511db1d235c5857dfaab1d6480dd0b919ee356f3.sh
/opt/dagster/dagster_home/schedules/scripts/511db1d235c5857dfaab1d6480dd0b919ee356f3.sh: line 7: syntax error near unexpected token `('
/opt/dagster/dagster_home/schedules/scripts/511db1d235c5857dfaab1d6480dd0b919ee356f3.sh: line 7: `export DAGSTER_PG_PW=*****(*****'```

workaround by doing `{"DAGSTER_PG_PW": '"blahlbah(blah"'}`?