Can't get S3-compatible storage working

[beatboxa]

I'm trying to use an S3 bucket on RustFS (which is AWS S3-compatible). I'm using all-in-one and docker-compose.

RustFS and my S3 buckets are working; and I am able to connect to them from other applications, using the same access key. For example, I am able to connect to external storage in NextCloud, use the AWS S3 connector and connect just fine--and within NextCloud, I can also interact with the objects in the bucket, such as creating a folder or file.

However, I cannot get it working in hi.events. I followed the documentation here: https://hi.events/docs/getting-started/deploying#aws-configuration

But elsewhere, I see additional environment variables, that are not listed in that documentation. For example, here: #95

Here is the latest iteration of some of my related environment variables:

•   APP_CDN_URL: (https://myS3url.com)
  

•   AWS_ACCESS_KEY_ID: (myaccesskey)
  

•   AWS_DEFAULT_REGION: (https://myS3url.com)
  

•   AWS_ENDPOINT: (https://myS3url.com)
  

•   AWS_PRIVATE_BUCKET: (myprivatebucket)
  

•   AWS_PUBLIC_BUCKET: (mypublicbucket)
  

•   AWS_SECRET_ACCESS_KEY: (mykey)
  

•   AWS_URL: (https://myS3url.com)
  

•   AWS_USE_PATH_STYLE_ENDPOINT: True
  

(I had previously tried a few of the above with (https://myS3url.com/mypublicbucket) as well.

When I attempt to upload a cover image through the hi.events UI, I get a "server error."

And the log shows this:

2026-01-05 04:13:09.692778+00:00WARNING:  database "hi-events" has a collation version mismatch
2026-01-05 04:13:09.692827+00:00DETAIL:  The database was created using collation version 2.36, but the operating system provides version 2.41.
2026-01-05 04:13:09.692841+00:00HINT:  Rebuild all objects in this database that use the default collation and run ALTER DATABASE "hi-events" REFRESH COLLATION VERSION, or build PostgreSQL with the right library version.
2026-01-05 04:13:09.780111+00:00172.16.9.1 - - [05/Jan/2026:04:13:09 +0000] "POST /api/images HTTP/1.1" 500 44 "https://(myurl)/manage/organizer/1/organizer-homepage-designer" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/142.0.0.0 Safari/537.36"
2026-01-05 04:13:09.780242+00:00127.0.0.1 -  05/Jan/2026:04:13:09 +0000 "POST /index.php/images" 500

I tried to reindex and alter the database; but I think that's a different issue. The main issue seems to be the next ones: the "500" messages..

Any suggestions?

If I cannot get this working, what is the volume/path that images are uploaded to, so that I can get some persistent file storage when upgrading the container? (ie. define a volume to map that container directory to a directory external to the docker container)?

[beatboxa]

I was able to get one step further: one issue was:

AWS_DEFAULT_REGION: (https://myS3url.com)

I the solution was:

AWS_DEFAULT_REGION: (myS3url.com)

(remove "https://")

This gets rid of the "server error" when uploading an image file.

However, now the issue is that the image file does not appear to make it into the S3 bucket; nor does an image appear (just a dead link image frame)...

[beatboxa]

Resolved the above. It now works as expected.

The second issue was:

APP_CDN_URL: (https://myS3url.com)

The solution was:

APP_CDN_URL: (https://myS3url.com/(mypublicbucket))

(added "/(mypublicbucket)" to the end of the URL)

I'm guessing this is in conjunction with the variable:

AWS_USE_PATH_STYLE_ENDPOINT: True

Otherwise, I'd guess the correct value would be: (https://(mypublicbucket).myS3url.com)

I'm also guessing I have some extraneous environment variables; but I don't want to rock the boat.

Hope this helps someone else;
and that the documentation is clarified for S3-compatible object storage on non-AWS services.