RTO for Cosmos DB Service-managed failover (regional outage)

Question

In a two-region account with one write region, what is the RTO for a service-managed failover during a regional outage?

Answer 1

Hi, @metalheart Welcome to the Microsoft Q&A platform, and thanks for using Azure Services.
If I understand correctly you want to know the RTO in case of service-managed failover for the write region.

Expected and maximum RPOs and RTOs depend on the kind of outage that Cosmos DB is experiencing. For instance, an outage of a single node will have a different expected RTO and RPO than a whole region outage.
Please refer to this doc for a better understanding High availability in Azure Cosmos DB | Microsoft Learn

The time to complete the failover is essentially dependent on the number of physical partitions for the account. So an account that has 1000 physical partitions will take longer than an account with 500 partitions. The volume of data stored in the partitions does not matter and it's purely a function of the physical partition count.

That said RTO is also dependent on the time taken to trigger the failover, which is subjective and this is why we don't have published SLAs on the RTOs for Service-Managed failover.

I hope this information helps, please let me know if you are looking for additional information.

Regards
Geetha