Create/Modify upstream CI job

[dtfranz]

To enforce our performance goals, we need to create a CI job or modify an existing one which enforces certain performance thresholds.

The e2e job in the github CI is a good first candidate. It allows us to avoid creating another CI job and use the near-real-world runtime environment to look for performance issues. If we find that the overhead of having prometheus running on our e2e tests is too high, we can instead add a new CI job and perhaps make it periodic.