Prometheus agent upgrade to v2

When upgrading your Prometheus agent to v2, you'll need to account for its reliance on Prometheus v3. This means any breaking changes introduced in Prometheus v3 may also be present in your agent's v2. Specifically, Prometheus v3 implements strict validation for the Content-Type header received during scraping. Unlike Prometheus v2, which would default to the standard Prometheus text protocol if the header was missing or invalid, Prometheus v3 will fail the scrape under these conditions.

As a result, you must implement the following changes:

• Both default Kubernetes jobs require fallback_scrape_protocol: "PrometheusText0.0.4" to be added to your specification. The default jobs for config.kubernetes.jobs should be configured as follows:

- job_name_prefix: default
  fallback_scrape_protocol: "PrometheusText0.0.4"
  target_discovery:
    pod: true
    endpoints: true
    filter:
      annotations:
        prometheus.io/scrape: true
- job_name_prefix: newrelic
  fallback_scrape_protocol: "PrometheusText0.0.4"
  integrations_filter:
    enabled: false
  target_discovery:
    pod: true
    endpoints: true
    filter:
      annotations:
        newrelic.io/scrape: true

• If you have any custom config.static_targets.jobs that target a Prometheus endpoint with missing or invalid headers, you must fix these to provide a valid header. As a temporary solution, you can add fallback_scrape_protocol: "PrometheusText0.0.4" to your specification.
• Similarly, any custom config.extra_scrape_configs that target a Prometheus endpoint with missing or invalid headers must be fixed to provide a valid header. A temporary solution is to add fallback_scrape_protocol: "PrometheusText0.0.4" to your specification.