e2e: backup access preflight drops the port-forward log when the S3 client cannot be configured #4243

Describe the bug

hack/e2e-chainsaw/_lib/backup-access-preflight.sh drops the port-forward capture when mc alias set fails. The other two failure paths in the helper print it: the readiness check at line 88 does, and the retry timeout does as of #4242. This one is the exit that still throws it away.

mc alias set reaches the endpoint, so a tunnel that came up and then went sideways is a plausible cause of a failure here, and port-forward.log is the only place that shows it.

Environment

• Cozystack: main at 79602e9ca
• In-tree e2e lane

To Reproduce

Make mc alias set fail inside the preflight, for example by pointing COZY_E2E_S3_PREFLIGHT_PORT at a port where something answers but is not the S3 gateway.

Expected behaviour

The failure prints port-forward.log the way the readiness and timeout paths do.

Actual behaviour

Lines 93-97 at 79602e9ca:

  if ! mc alias set backup-preflight "https://127.0.0.1:${local_port}" \
      "${access_key}" "${secret_key}" --insecure >/dev/null; then
    echo "failed to configure S3 client for BucketAccess ${namespace}/${access_name}" >&2
    return 1
  fi

One correction worth recording, because it makes the gap smaller than it looks: only stdout goes to /dev/null here, so mc's own error message does reach the job log. What is missing is the port-forward capture, not everything.

Additional context

Lower priority than the timeout path, which fires on the S3 round trip itself and is far more common. A fix wants its own case in hack/backup-access-preflight.bats.

Checklist

• I have checked the documentation
• I have searched for similar issues
• I have included all required information
• I have provided clear steps to reproduce
• I have included relevant logs