Add some generic debug logs

[sarayourfriend]

Problem

We sometimes get p99 response timing spikes up into the 20s of seconds. It would be nice to be able to have some idea where this is coming from.

Until we are able to ingest and use the Prometheus metrics for the frontend, we can rely on logs to gather general timings.

Description

Log server side render timings and API request response times. To make it possible to easily trace a requests logs, we should generate and log a request ID (similar to how we do this for the Django logs).

The goal is to identify whether:

1. We could eagerly timeout API requests and return a "skeleton" rendered page to improve perceived frontend performance
2. If there are specific types of SSR requests that cause long render times (or perhaps lengthen times in other parts of the Nuxt request/response lifecycle)

We will want to complete #536 first to add a logging utility, probably implementing the request ID there. And then for this issue (as a separate task) write a Nuxt plugin that adds logging to various hooks with the request ID put into the Nuxt context.

Alternatives

Someone could spend time focused on getting Grafana Agent wired up to consume Prometheus metrics from the Nuxt service. We need to change the load balancer rules for the Nuxt service to disallow external requests from hitting the metrics endpoint, only the Grafana Agent should be able to reach it.

Additional context

Implementation

• 🙋 I would be interested in implementing this feature.

[sarayourfriend]

This is technically solved by #3108 because we can now actually see per-request timing on Nuxt. I'll close this as solved. It would be nice to add more logging to the Nuxt service, but let's do so on an as-needed basis, rather than artificially trying to decide what logs are helpful.