Region Requirements
Applies to every DynamoAI product. Cloud-agnostic; the commands that verify each requirement are in Cloud Reference.
Regions within one geography are not equivalent. The second region of a geography often carries fewer availability zones and a narrower set of GPU families than the first. Confirm each requirement below against both regions before committing to a pair. Several of them cannot be changed after provisioning.
Requirements
| Requirement | Why | Verify |
|---|---|---|
| Availability zones in both regions | The standby provides the availability of its own region. A region without zones cannot host zone-redundant datastores or spread replicas across zones | Cloud Reference |
| The GPU families the guardrail models require, in both regions | Guardrail serving is pinned to the GPU it was sized for. A region without that family needs a different instance type and a different card count for the same throughput | Cloud Reference |
| Quota for the standby GPU fleet | Capacity a region offers is still subject to your own quota, which is set per account and per region | Cloud Reference |
| A PostgreSQL tier that supports high availability and read replicas | The cheapest tier on most clouds supports neither, at any size | Datastores |
| A dedicated MongoDB cluster tier | Free and Flex clusters offer no point-in-time recovery and no multi-region option | Datastores |
| A global load balancer with health probing | Regional withdrawal is driven by probe results | Traffic Routing |
| A voting majority that survives the loss of one region | A replica set elects a primary only while a majority is reachable | Data Replication |
Availability zone support and GPU availability vary between regions of the same geography. Some regions offer no availability zones at all, and some offer a single GPU family. A pair chosen on the assumption that both regions are equivalent can pass design review and then fail to host the standby.
Decisions That Cannot Be Reversed
These are settled at provisioning time. Changing them later means rebuilding the resource.
| Component | Decision | Consequence |
|---|---|---|
| Node pools | Availability zones | Fixed per node pool. Changing zones requires a new pool and a workload migration |
| PostgreSQL | Compute tier | The cheapest tier can never gain high availability or read replicas. See Cloud Reference for the tier names |
| PostgreSQL | Geo-redundant backup | Can only be enabled when the server is created |
| Redis | High availability | Can be enabled on a running instance, but never disabled. Persistence depends on it |
| MongoDB | Region count and voter placement | Two regions cannot survive loss of the majority region without manual reconfiguration |
| Object store | Distributed mode | Single-instance mode cannot be converted in place. See Object Storage |
Verifying
Run the same three checks against both regions of the pair and compare the output: availability zones offered, GPU families offered, and current quota against limit. The commands differ per cloud.
A region that reports no availability zones cannot host the standby as designed.