IssueTracker

The other use case here is to do historical reporting on Incidents.

Lacking an API the only option would be to have every policy send a notification to a pub/sub or webhook to be saved into a database. There's no option to (for example) send incidents directly to BigQuery (or some other GCP Database).

In any case this has two limitations:

1. People have to remember to explicitly configure every policy (even if using Terraform or some other Infrastructure-as-code option)
2. You can't retroactively do reporting on the period before you realized there was no API, or on a policy that accidentally omitted the reporting endpoint.

Minimally, incident statuses should be available like other metrics so that they can be queried and linked to via existing systems. It would be nice to have an api specifically for fetching the additional incident details, but I feel like this could be achieved incrementally, with the incidents having their own metric descriptor (name, status, etc.).

We're in the process of moving to Grafana based dashboards integrated with cloud monitoring and cloud logging plugins, and while we're able to handle most of our dashboard and logging access needs this way, observability of incidents doesn't appear to have an easy way to integrate currently. Due to the storage of the data within GCP's infrastructure, it doesn't seem feasible to move to Grafana's alerting system. Having some way to retrieve basic data on incidents via metrics would address our primary and most present concern at least.

If/when an incident management API will be made public, this could allow plugin development specifically for that purpose like the existing Cloud Logging, Monitoring, and Tracing plugins currently available, which would be ideal. If getting metrics on alert incidents is a lower and easier bar to achieve, would it be something that would fit in with this feature request, or should it be made separately?