Three ARM deployments of a PostgreSQL Flexible Server in West US 2 have
failed with the same terminal error on the Postgres resource only. Every
other resource in the same template succeeded (Container Apps
environment, container app, jobs).
Requested: Standard_B1ms (Burstable), PostgreSQL 16, 32 GB, no HA
"code": "CapacityNotAvailable"
"message": "Capacity is not available in this region/zone.
Please retry after some time."
Attempts: two with an Azure-selected availability zone, one with an
explicit availabilityZone = 1. Each ran 20-25 minutes before reaching
terminal state 'Failed'.
Diagnostics already done:
- "az postgres flexible-server list-skus --location westus2" advertises
Standard_B1ms and Standard_B2s in zones 1, 2 and 3.
- The subscription-scoped capabilities endpoint
(.../locations/westus2/capabilities?api-version=2025-08-01) shows no
restriction in effect: supportedFeatures contains
{"name":"OfferRestricted","status":"Disabled"}, and all three tiers
plus the full SKU list are present and unmarked.
- That same payload reports ZoneRedundantHa: Disabled,
ZoneRedundantHaAndGeoBackup: Disabled and FastProvisioning: Disabled
for West US 2.
- Microsoft.DBforPostgreSQL is Registered; flexibleServers exposes
api-version 2025-08-01 and newer.
So this appears to be genuine regional capacity exhaustion rather than an
offer restriction, quota limit or permissions issue. Is there a capacity
constraint affecting PostgreSQL Flexible Server provisioning in West US 2,
and what is the recommended path forward?
I cannot file an Azure support request (no Support Request Contributor
role on the subscription), which is why I am asking here.