extcosmosdb

type CosmosDbFailoverState ¶

type CosmosDbFailoverState struct {
	SubscriptionId    string
	ResourceGroupName string
	AccountName       string
	// PromotedRegion is the region we will promote to priority 0. Captured for log messaging.
	PromotedRegion string
	// NewPolicies is the full priority order to submit at Start (computed from the current order
	// at Prepare, with PromotedRegion swapped to priority 0).
	NewPolicies []failoverPolicy
}

CosmosDbFailoverState carries the computed new failover priorities from Prepare to Start. This attack is one-shot (no Stop / no rollback): Azure's FailoverPriorityChange is async and takes 5-15 min per call, so a transient experiment-style "set & restore" semantics is misleading. Users who want to switch back can run the experiment again against the (now-secondary) region.