Sigv4 Extension configuration failure for prometheusremotewrite

[jacobwalters]

I am using the opentelemetryoperator version v0.54.0. I am attempting to configure a prometheus collector that will push metrics to Amazon Managed Prometheus. I am trying to use the mechanism of the sigv4 extension with the prometheusremotewrite as described in this document.

https://github.com/open-telemetry/opentelemetry-collector-contrib/blob/main/extension/sigv4authextension/README.md

The collector pod has this error in the logs

2022/07/26 20:12:00 AWS OTel Collector version: v0.19.0
2022/07/26 20:12:00 found no extra config, skip it, err: open /opt/aws/aws-otel-collector/etc/extracfg.txt: no such file or directory
Error: failed to get config: invalid configuration: extension "sigv4auth" has invalid configuration: bad AWS credentials
2022/07/26 20:12:00 application run finished with error: failed to get config: invalid configuration: extension "sigv4auth" has invalid configuration: bad AWS credentials

[pavolloffay]

@jacobwalters is this related to collector or target allocator?

For a better understanding of the issue could you please share the full CR?

[bryan-aguilar]

We have seen this error before when IRSA is not properly setup. Docs here should point you in the correct direction.

[ThaiDangFr]

Same error here, and I'm sure IRSA is correctly configured as it works when EKS and AMP are in the same account. With different accounts, when I use the assume_role parameter, I end with the bad AWS credentials error.

[bryan-aguilar]

@jacobwalters @ThaiDangFr can you please share the collector configurations you were using? And the full CR as @pavolloffay requested.

[ThaiDangFr]

finally issue solved for me this week end, the trust policy was not set properly

[jaronoff97]

Closing this as it seems the issue is solved, please let me know if that's not the case.