Umami - Failed container start with environment variables

[echokos]

Adding any environment variables to the env file and then restarting the container causes the app to never start.

For example - on a new install, add FORCE_SSL = 1 to the env file and restart the container. The error log contains:

May 24 20:30:40 /app/data/env: line 5: FORCE_SSL: command not found
May 24 20:31:41 ==> Changing ownership
May 24 20:31:41 => Building database
May 24 20:31:41 /app/data/env: line 5: FORCE_SSL: command not found
May 24 20:32:41 ==> Changing ownership
May 24 20:32:41 => Building database
May 24 20:32:41 /app/data/env: line 5: FORCE_SSL: command not found

And it will continually stay in this state creating the errors forever until the container is stopped.

[robi]

@echokos Sounds like a packaging bug. @staff will find it.

[Neiluj]

@echokos said in Umami - Failed container start with environment variables:

May 24 20:32:41 /app/data/env: line

I had something similar with IGNORE_IP when I lazily copied the variable from the umami doc (https://umami.is/docs/environment-variables)

try this FORCE_SSL=1 (without spaces)

and restart your app - This seems to work.

[echokos]

@Neiluj That sure is the issue. @staff - maybe just adding a note in the Cloudron docs would be enough for this issue.

[nebulon]

For background, that file is source /app/data/env in the start.sh so it has to be a valid shell script. Such a behavior is happening in many apps, not sure how to document this or if we can add some extra trimming/filtering for such cases?

[3246]

@Neiluj thank you! That fixed it for me