Timeout should be longer or at least configurable

[dgolombek]

Describe the bug

We're seeing occasional timeouts while trying to login to ECR. It's not super frequent, but given enough builds, it ends up being noticeable. It feels like 15s is a short default -- I'd expect 30s for robustness, or at least for this to be configurable -- I'd rather our runs be robust even if they are sometimes a little bit slower.

Thu, 09 Oct 2025 21:49:44 GMT
Run aws-actions/amazon-ecr-login@062b18b96a7aff071d4dc91bc00c4c1a7945b076
Thu, 09 Oct 2025 21:49:44 GMT
Logging into registry <ACCOUNT>.dkr.ecr.***.amazonaws.com
Thu, 09 Oct 2025 21:50:02 GMT
Error: Could not login to registry <ACCOUNT>.dkr.ecr.***.amazonaws.com: WARNING! Using --password via the CLI is insecure. Use --password-stdin.
Error response from daemon: Get "[https://<ACCOUNT>.dkr.ecr.***.amazonaws.com/v2/](https://<ACCOUNT>.dkr.ecr.%2A%2A%2A.amazonaws.com/v2/)": net/http: request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)

To Reproduce

      - name: Login to Amazon ECR
        id: login-ecr
        uses: aws-actions/amazon-ecr-login@062b18b96a7aff071d4dc91bc00c4c1a7945b076 # v2.0.1

Run 1000 times, sometimes at peak times.

Expected behavior

This succeeds 99.99+% of the time.