ADA-C01 Question 552
Single answerIdentify the appropriate failover or fallback procedureA company uses Snowflake Business Critical Edition with account replication and a configured failover group between AWS us-east-1 (primary) and AWS us-west-2 (secondary). During a regional outage in us-east-1, analysts must resume querying replicated production data in the secondary account as quickly as possible. The administrator also wants the environment to be ready for applications to reconnect with minimal changes. Which action should the administrator take first in the secondary account?
- A
Refresh the failover group in the secondary account, then resume all suspended warehouses.
- B
Promote the secondary failover group to primary, then redirect clients to the secondary account endpoint.
- C
Convert the replicated databases to writable databases using database failover, then recreate users manually.
- D
Use Time Travel to restore the production databases in the secondary account, then update replication schedules.
Show answer and explanation
Correct answer: B
Explanation
For Snowflake disaster recovery across regions or cloud regions, the appropriate procedure when the primary region is unavailable is to fail over the configured secondary failover group. Failover groups are intended to replicate supported account objects, such as databases and certain account-level objects, so that a secondary account can be promoted during an outage. Once the secondary failover group is promoted to primary, clients can be redirected to the secondary account to resume operations. This is different from fallback-style recovery methods such as Time Travel, which recover previous object states but do not provide regional disaster recovery. Best practice is to preconfigure connection management and client redirection processes so application changes are minimal during failover. Refer to Snowflake documentation on Account Replication, Failover Groups, and Failover/Failback procedures for the exact operational steps and supported object behavior.
- A. Incorrect.
Incorrect. A failover group is refreshed from the primary before failover, but during an actual primary-region outage the key recovery action in the secondary is to perform failover by promoting the secondary failover group to primary. Resuming warehouses may be necessary afterward for compute, but it is not the first or defining failover step. Also, replicated objects remain read-only until failover occurs.
- B. Correct.
Correct. In a cross-region disaster scenario using account replication and failover groups, the administrator should fail over by promoting the secondary failover group to primary. This makes supported replicated objects in the secondary account active and writable where applicable, allowing operations to continue there. After failover, applications and users can be redirected to the secondary account URL/connection endpoint. This is the standard Snowflake failover procedure for a regional outage.
- C. Incorrect.
Incorrect. Snowflake does not use a separate 'database failover' process to individually convert replicated databases in this scenario when failover groups are already configured. Failover groups are designed to replicate and fail over a broader set of account objects together. Recreating users manually is also unnecessary if users/roles are included in the failover group and have replicated successfully.
- D. Incorrect.
Incorrect. Time Travel is for recovering historical versions of data within a database, schema, or table in the same account context; it is not the mechanism for regional disaster recovery between replicated accounts. In this scenario, account replication with failover groups is the appropriate DR capability, not restoring objects with Time Travel.