Problem in v.0.6.0 versus v0.5.0

[phijzen]

When I run docker image cloudlena/s3manager version v0.5.0, i can successfully use the tool to interact with my buckets, but when i upgrade to version v0.6.0 the application fails to start and shows this error:

2026/02/09 12:28:20 no S3 instances configured. Please provide numbered environment variables like 1_NAME, 1_ENDPOINT, etc.

I have not changed any config when i upgraded from v0.5.0 to v0.6.0. The env vars that i apply to the instance are:

USE_IAM=True
AWS_ROLE_ARN=***
AWS_WEB_IDENTITY_TOKEN_FILE=***
AWS_STS_REGIONAL_ENDPOINTS=***
AWS_DEFAULT_REGION=***                                                                             
AWS_REGION=***                                                                                      

[phijzen]

Even when i add 1_NAME with a bucket name, it fails the same way.
Same result when i change all AWS_... env vars to 1_AWS_.. as well.

[luucasrb]

Same problem here. Besides that, we also have a problem that our Kubernetes orchestrator (Rancher) does not accept Environment Variables with the names 1_ since they use a validation rule to check the names ('[-._a-zA-Z][-._a-zA-Z0-9]*'), which makes a little bit harder to configure the new version of s3manager.