digma-ai / opentelemetry-instrumentation-digma Goto Github PK

Hi,
Thank you for building this library. I tried to use @Instrument decorator in my class and got to a problem that it seems like the decorator tries to push 'self' into static methods args.

Basic example:

@instrument
class Foo:
    @staticmethod
    def foo():
        pass

f = Foo()
f.foo()

Returns:

TypeError: foo() takes 0 positional arguments but 1 was given

Any suggestions?

I was interested in using the decorator you have built for tracing, but this package bundles a hardcoded opentelemetry version (1.12.0) that is old and contains bugs that have been fixed.

In order to avoid conflicts with different customer configurations, Digma should use its own deployment attribute
"digma.environment"

please note that the backend already supports it

need to issue a new version of the instrumentation and update our samples

Modify the exporter code to include two optional environment variables if they exist:

1. ENVIRONMENT
2. COMMIT_ID
   Both these values should be passed on to the server along with the error information

Hello,
I have a project full of opentracing/ opentelemetry span. I was looking on how to implement decorators and found your excelent article @doppleware

Since I have a lot of rework for moving opentracing -> opentelemetry, I wonder if I can do a step by step with this implementation.
And why not using Digma when I have some times.

regards

Write custom formatter for traceback that looks at the _locals and more intelligently decides what to do instead of automatically converting to string.

1. If value is str and big (over 64 characters), record length [Done]
2. If value is array and big (over 10 members), record length [Done]
3. If value is object, don't serialize to str -- ignore [Done]
4. If value is numeric, record value
5. If value is boolean, record value