dyladan
commented
1 year ago You're correct that this is because only strings are supported. If you want to use non-string attributes you will need to convert them to string first. In the case of a resource detector you could for example create a wrapper around the resource detector which converts all attribute values to string. Something like the following:
class StringConversionResourceDetector(ResourceDetector):
def __init__(self, detector, raise_on_error=False):
super().__init__(raise_on_error)
self.detector = detector
def detect(self) -> "Resource":
resource = self.detector.detect(self)
# convert resources to string here
for key in resource.attributes:
resource.attributes[key] = str(resource.attributes[key])
return resource
RangelReale
arminru
Is your feature request related to a problem? Please describe. I'm getting this error message on each of my metrics:
Skipping unsupported dimension with value type 'int'. This seems to be because only string dimension values are supported, but I'm using cloud resource detectors, like opentelemetry-resourcedetector-gcp, and it fetches some of its attributes from a metadata url, and python converts them to int automatically, so I have no way of pre-converting it before sending to the dynatrace exporter.Describe the solution you'd like That attribute values are automatically converted to string using
str.Describe alternatives you've considered There's no place where I can put a callback so I could convert them to string myself.
Additional context