Migrate dynamic scripted Grafana Dashboards to Scenes

[ntheanh201]

1. Upgrade to Grafana v10.2.3
2. Migrate Scripted Dashboards to Grafana Scenes

• I've transitioned our existing scripted dashboards to the new Grafana Scenes. The migration also ensures compatibility with future Grafana updates and aligns with Grafana's recommended best practices. Check it here: New in Grafana 10: Grafana Scenes for building dynamic dashboarding experiences
  

3. Update documentation: the way to configure Grafana and the current Traffic Portal doesn't use directly the Grafana dashboards

---

Which Traffic Control components are affected by this PR?

• Traffic Stats
• CDN in a Box

What is the best way to verify this PR?

• Build Traffic Stats
• Run CDN in a Box with Traffic Stats, Grafana
• Visit Grafana server, go to Apps -> TrafficControl Grafana Scenes
• See the dashboards

Check the images below

If this is a bugfix, which Traffic Control versions contained the bug?

PR submission checklist

• This PR has tests
• This PR has documentation
• This PR has a CHANGELOG.md entry
• This PR DOES NOT FIX A SERIOUS SECURITY VULNERABILITY (see the Apache Software Foundation's security guidelines for details)

[ntheanh201]

Cache Group dashboard

[ntheanh201]

Delivery Service dashboard

[ntheanh201]

Server dashboard

[ntheanh201]

@jhg03a I also updated the Dockerfile of Grafana, I think you can give me some reviews

[ocket8888]

This all looks fine on GitHub. I'll carve out some time tomorrow to try to spin it up and see how it actually works (#5939 permitting).

[ntheanh201]

This all looks fine on GitHub. I'll carve out some time tomorrow to try to spin it up and see how it actually works (#5939 permitting).

Did you finish with checking?

[ocket8888]

Did you finish with checking?

Sorry, no, I got busy. And then... I forgot 😅

[ocket8888]

This appears to work in its own right, but even after rebasing I can't get Docker to realize the builder on this branch is newer than the one it can pull. So I'm hoping merging it to master just solves that by publishing a new builder image. If it doesn't, for some reason, CiaB runs may be broken until it's resolved (locally, in all likelihood CI would be unaffected in this scenario) - although the project itself is still buildable if done from actual scratch. I think that's unlikely to happen, but excited to find out for sure.