[RFE] Refactor main.rs to remove exposed goalstate sections and use configurable IMDS timeout and retry settings

[peytonr18]

References Issue #137 with custom configuration settings for IMDS timeout and retry settings.

Remove the vm_goalstate and goalstate::report_health sections in main.rs, as they do not need direct exposure to the user, and we should strive to keep main.rs as basic as possible. Also update the hardcoded IMDS retry and timeout values to leverage the ability to have custom configuration settings as is created by config.rs.

Put this logic in libazureinit:

    let vm_goalstate = goalstate::get_goalstate(
        &client,
        Duration::from_secs(imds_http_retry_interval_sec),
        Duration::from_secs(imds_http_timeout_sec),
        None, // default wireserver goalstate URL
    )
    .await
    .with_context(|| "Failed to get desired goalstate.")?;

    goalstate::report_health(
        &client,
        vm_goalstate,
        Duration::from_secs(imds_http_retry_interval_sec),
        Duration::from_secs(imds_http_timeout_sec),
        None, // default wireserver health URL
    )
    .await
    .with_context(|| "Failed to report VM health.")?;

Use the IMDS struct here, instead:

    let imds_http_timeout_sec: u64 = 5 * 60;
    let imds_http_retry_interval_sec: u64 = 2;