为何自定义指标在写入时设置资源类型而非描述符阶段？能否强制绑定？

文档提到手动创建指标描述符时需要选择资源类型，但创建MetricDescriptor的代码中没有设置资源类型的字段：

client = monitoring_v3.MetricServiceClient()
project_name = f"projects/{project_id}"
descriptor = ga_metric.MetricDescriptor()
descriptor.type = "custom.googleapis.com/my_metric" + str(uuid.uuid4())
descriptor.metric_kind = ga_metric.MetricDescriptor.MetricKind.GAUGE
descriptor.value_type = ga_metric.MetricDescriptor.ValueType.DOUBLE
descriptor.description = "This is a simple example of a custom metric."

labels = ga_label.LabelDescriptor()
labels.key = "TestLabel"
labels.value_type = ga_label.LabelDescriptor.ValueType.STRING
labels.description = "This is a test label"
descriptor.labels.append(labels)

descriptor = client.create_metric_descriptor(
    name=project_name, metric_descriptor=descriptor
)
print("Created {}.".format(descriptor.name))

只有在写入自定义指标时才设置资源类型：

series = monitoring_v3.TimeSeries()
series.metric.type = "custom.googleapis.com/my_metric" + str(uuid.uuid4())

###### SET TYPE HERE ######
series.resource.type = "gce_instance"

series.resource.labels["instance_id"] = "1234567890123456789"
series.resource.labels["zone"] = "us-central1-c"
series.metric.labels["TestLabel"] = "My Label Data"
now = time.time()
seconds = int(now)
nanos = int((now - seconds) * 10**9)
interval = monitoring_v3.TimeInterval(
    {"end_time": {"seconds": seconds, "nanos": nanos}}
)
point = monitoring_v3.Point({"interval": interval, "value": {"double_value": 3.14}})
series.points = [point]
client.create_time_series(name=project_name, time_series=[series])

尝试用gce_instance、generic_task和global三种资源类型写入该指标，结果生成了三个不同的指标。原本期望能将自定义指标绑定到特定资源类型，这样写入错误类型时会抛出错误。想问是否有办法为自定义指标强制绑定资源类型？

可以在创建MetricDescriptor时显式设置resource_type字段，以此强制绑定特定资源类型。修改后的创建代码如下：

client = monitoring_v3.MetricServiceClient()
project_name = f"projects/{project_id}"
descriptor = ga_metric.MetricDescriptor()
descriptor.type = "custom.googleapis.com/my_metric" + str(uuid.uuid4())
descriptor.metric_kind = ga_metric.MetricDescriptor.MetricKind.GAUGE
descriptor.value_type = ga_metric.MetricDescriptor.ValueType.DOUBLE
descriptor.description = "This is a simple example of a custom metric."
# 绑定指定资源类型
descriptor.resource_type = "gce_instance"

labels = ga_label.LabelDescriptor()
labels.key = "TestLabel"
labels.value_type = ga_label.LabelDescriptor.ValueType.STRING
labels.description = "This is a test label"
descriptor.labels.append(labels)

descriptor = client.create_metric_descriptor(
    name=project_name, metric_descriptor=descriptor
)
print("Created {}.".format(descriptor.name))

这样创建的指标描述符就和gce_instance资源类型绑定，后续写入指标时如果使用其他资源类型，调用create_time_series会直接抛出错误，符合预期的校验逻辑。

内容的提问来源于stack exchange，提问作者red888