set up a status page for registry

[matifali]

Problem Description

The registry’s status page at coder.instatus.com needs enhancements for SaaS-grade reliability and transparency. Current monitoring relies on a basic 404-checking script. With the registry now deployed on Google Cloud Run and Instatus supporting Prometheus and Google Cloud integrations, this issue focuses on modernizing monitoring, alerting, and user access.

---

Desired Solution

Integration

• Add a prominent link to the status page in the registry footer.
• Update status categories (e.g., uptime, latency, etc) for better user insights. (Nice to have)

Monitoring

• Use Google Cloud Run metrics to track container health, CPU and memory usage, latency, etc. feeding data into Prometheus.
• Integrate Prometheus with Instatus for real-time metrics (e.g., success rates, errors, latency). (Nice to have)

Alerting

• Configure multi-channel alerts (email, Slack, PagerDuty) with thresholds for high(outage)/low(e.g. higher latency) priority issues.
• Ensure alerts are logged and visible on the status page for incident transparency.

Documentation

• Provide a playbook for handling alerts and troubleshooting incidents.
• Document Prometheus, Cloud Run, and Instatus setup steps.

Definition of Done

• Status page displays real-time, accurate health data and is linked in the registry UI.
• Monitoring (Prometheus + Cloud Run) and alerting pipelines are fully tested.
• All setup and maintenance documentation is complete and accessible.

[matifali]

I am setting one up. See coder/modules#342

[matifali]

instatus support Prometheus as a source.

[bpmct]

Looks like we're currently hosting this on https://coder.instatus.com/.

@matifali - can you work with Kira and Ben to create a good "definition of done" or product requirements for this issue? Some ideas, not well baked:

• The status page is clearly linked on the bottom of the registry
• We are confident in the "checker" script, and don't attribute it to any issues
• Alerting is properly set up

[matifali]

@Kira-Pilot, @bpmct , I Updated the issue Body to include the requirement.

[Parkreiner]

From the discussions last week, it sounds like we're okay with just updating the registry to link to the Instatus page. In which case, I can make a quick PR for that in a few minutes

[matifali]

Yes. We can link it to the footer. 🚀