Request: Include S3 Storage Module

[scooke]

Since I'm on a Minio/object storage kick, I tried setting this up. But it didn't work using the linked info. I tried setting it up just as path style, rather than domain style (as in, I didn't make a new A Record using the bucket name in use). I'll try later.

[nichu42]

Has anyone made this work? Care to share your steps?
I made the entries for IDrive e2 in homeserver.yaml, but I keep getting an "invalid enpoint" error:

Jan 08 13:18:56 File "/app/code/env/lib/python3.10/site-packages/botocore/endpoint.py", line 402, in create_endpoint
Jan 08 13:18:56 raise ValueError("Invalid endpoint: %s" % endpoint_url)
Jan 08 13:18:56 ValueError: Invalid endpoint: xxxx.fra.idrivee2-42.com

I double-checked with rclone: All data was entered correctly.

[nichu42]

@nichu42
OK, I found the error. The config file is expecting the endpoint URL to start with https://
I have the S3 storage up and running now. If anyone needs help, please feel free to reach out to me.

[jdaviescoates]

@nichu42 great, thanks!

If you document your process/ set-up here, you or someone else might also be able to add the details to the docs via https://git.cloudron.io/cloudron/docs

[nichu42]

@jdaviescoates

Sure! It is quite easy, once you know what you are doing.

Take the following code and replace the <S3...> boxes with your S3 compatible storage's data.

media_storage_providers:
- module: s3_storage_provider.S3StorageProviderBackend
  store_local: True
  store_remote: True
  store_synchronous: True
  config:
    bucket: <S3_BUCKET_NAME>
    region_name: <S3_REGION_NAME>
    endpoint_url: <S3_LIKE_SERVICE_ENDPOINT_URL>
    access_key_id: <S3_ACCESS_KEY_ID>
    secret_access_key: <S3_SECRET_ACCESS_KEY>

Make sure the endpoint URL starts with https://
Copy the code to your clipboard.

Open the file manager of your Synapse App in Cloudron Dashboard. Navigate to configs and open the homeserver.yaml file. Paste the code from your clipboard at the end of the configuration file.
Save and restart the app.

[jdaviescoates]

@nichu42 great, thanks!

[andreasdueren]

@nichu42 What about the cleanup script and pushing existing media to s3? Did you ever set that up?

[nichu42]

@andreasdueren Synapse is on S3, yes.
My other project was Mastodon. It's still to be done.

[andreasdueren]

@nichu42 But what did you do with these? Are these scripts already installed in the package somewhere and can be called with a cron job?

[andreasdueren]

Also a bit confused as this just revealed to me that matrix is storing my thumbnails unencrypted?!