Hi digital.ai,
We would like to send custom response-content without the mandatory JythonResponse format.
Currently, all plugins in digital.ai release with a rest-api interface respond with the payload constructed of a predefined format of this: {"entity": { .. }, "exception": null, "stderr": "", "stdout": ""}. We would like to send a custom content type like plain times series directly from release which are not compliant to this predefined struct.
For fetching those it might be necessary to provide the @Timed annotation at code level for the metrics API calls.
It would be a great opportunity providing customers the necessary interface to integrate digital.ai release into an existing monitoring ecosystem.
br,
Daniel
by: Daniel R. | 10 months ago | API & CLI tools
Comments
The extensive customization options in calls with third-party systems might lead to resource-intensive endpoint requests, potentially impacting system performance, which could be challenging to ensure. Do you have specific or more concrete use cases in mind that you would like to discuss?
I totally agree that this can be potentially impacting system performance. But I assume that this still be the fault of the Product Consumers using or providing those capabilities without any mind, control or regulations?
We would like to implement custom interfaces to for our operators and product teams to provide visualizations for release and task specific metrics like durations and delays that impact their responsibilities.
We believe this request is valid. Currently, we are experiencing a high volume of requests, so we will wait for broader customer upvotes to prioritize its development.