Google Cloud - Error Reporting Client Libraries

0
votes

I'd like to use the Google Error Reporting Client library (from google.cloud import error_reporting).

Basically, you instantiate a client: client = error_reporting.Client(service="my_script", version="my_version") and then you can raise error using:

  1. client.report("my message") or
  2. client.report_exception() when an exception is caught

I have 3 environments (prod, staging and dev). They are each setup on their own Kubernetes cluster (with their own namespace). When I look at Google Cloud Error Reporting dashboard, I would to quickly locate on which environment and which class/script the error was raised.

Using service is a natural choice to describe the class/script but what about the environment? What is the best practice? Should I use the version to store that, e.g. version="staging_0.0.2"?

Many thanks in advance Cheers, Lamp'

1
pythonkuberneteserror-reportinggoogle-cloud-logginggoogle-cloud-error-reporting
Can you explain in more details what you've tried so far ? – Wojtek_B
I think you should provide your feedback to Google using the Issue Tracker. Two things surprise me about this. First, that you can't apply an arbitrary set of labels to errors. Second, that the service doesn't instrument known Resource Types like GKE containers. – DazWilkin
See: issuetracker.google.com/issues/181167949 – DazWilkin

1 Answers

0
votes

I think the Error Reporting service is deficient (see comment above).

Since you're using Kubernetes, how about naming your Error Reporting services to reflect Kubernetes Service names: ${service}.${namespace}.svc.cluster.local?

You could|should replace the internal cluster.local domain part with some unique external specifier (FQDN) to your cluster: $[service}.${namespace}.${cluster}

NOTE These needn't be actual Kubernetes Services but some way for you to uniquely identify the thing within a Kubernetes cluster my_script.errorreporting.${namespace}.${cluster}