Product: Cloud SQL for PostgreSQL
Issue Type: Service Availability / Capacity Planning
Background:
Per GCP documentation on Patterns for scalable and resilient apps | Cloud Architecture Center | Google Cloud Documentation , Google recommends deploying across multiple zones and using managed services with
built-in redundancy. We have followed these guidelines by enabling High Availability (multi-zone) for our Cloud SQL instance.
However, we are now blocked from performing updates due to regional capacity constraints, with no available mitigation path since:
- Reservations are not supported for Cloud SQL (per About reservations | Compute Engine | Google Cloud Documentation - only Compute Engine, GKE, Batch, Dataflow, Dataproc, and
Vertex AI support reservations) - No advance warning system exists for Cloud SQL capacity constraints
- No priority queue for existing production workloads
Error Message:
The zone or region does not have sufficient resources to handle the request at the moment.
Questions for Google Cloud:
- Capacity Timeline: What is the current capacity status and estimated restoration timeline for db-perf-optimized-N-4 in europe-west6?
- Priority Allocation: Can Google prioritize capacity for existing production instances over new deployments?
- Alternatives: Are there equivalent machine types with current availability? Specifically:
- Can we use a different N-series configuration?
- Is db-custom-4-32768 available as a workaround?
- Feature Request: Given that reservations don’t support Cloud SQL, what mechanisms exist (or are planned) for capacity assurance on managed database services?
- Monitoring: Is there a way to receive proactive notifications about capacity constraints affecting our instance types/regions?