ibm_cis_healthcheck | TCP Monitor bugs with default values

[MaAd12345]

Description

When creating a TCP health check monitor (type = "tcp"), the provider causes constant drift for two settings that are not applicable to TCP monitors but have default values in the provider:

1. path defaults to "/" but is not applicable for TCP checks
2. method defaults to "GET" but TCP checks use "connection_established" internally

  ~ resource "ibm_cis_healthcheck" "healthcheck" {
        id               = "<redacted>"
      ~ method           = "connection_established" -> "GET"
      + path             = "/"

This results in every terraform plan showing changes that can never be applied successfully. Unfortunately, this makes a lot of “noise” if you want to apply a real changes.

Technical Details

I guess the issue has two root causes:

1. The schema defines defaults that only make sense for HTTP(S) monitors but not for TCP:

cisGLBHealthCheckPath: {
    Type:         schema.TypeString,
    Description:  "path",
    Optional:     true,
    Default:      "/",  // This default causes drift for TCP
    ValidateFunc: validate.ValidateURLPath,
},
cisGLBHealthCheckMethod: {
    Type:         schema.TypeString,
    Description:  "method",
    Optional:     true,
    Default:      "GET",  // This default causes drift for TCP
    ValidateFunc: validate.InvokeValidator(ibmCISHealthCheck, cisGLBHealthCheckMethod),
},

Expected Behavior

For TCP monitors:

• path should not have a default value when type = "tcp"
• method should either accept "connection_established" or be completely optional for TCP monitors

The schema should be conditional based on the monitor type to avoid setting HTTP(S)-specific defaults for TCP monitors.