Golden Signals for Numaflow Pipeline

[vigith]

how do we quantify the reliability, availability, and performance of a Numaflow pipeline? While pipelines differ in their business objectives, what are the common denominators of core / key metrics that Numaflow (as a platform) needs to measure, monitor, and alert on?

We need to capture for 2 personas

• [ ] Numaflow health for platform components
  • [ ] Controller
  • [ ] ISB
  • [ ] numa container
• [x] #1822
  • [ ] backpressure
  • [ ] UDF container health