Optimized Refresh and RPO Assurance for failover groups

Optimized Refresh and RPO Assurance make failover group replication more efficient, scaling with the rate of change rather than the number of objects. Optimized Refresh processes only metadata changes, while RPO Assurance manages continuous refreshes with SLA backing.

Use cases

Imagine you have hundreds of databases and schemas to replicate and need predictable refresh durations and lower data loss risk. With these features, you can:

• Lower refresh duration by scaling with the rate of change.
• Automate continuous refreshes to keep secondary accounts in sync.
• Ensure RPO alignment backed by an SLA on eligible pairs.

How to use

To enable Optimized Refresh on a new failover group on the primary account:

CREATE FAILOVER GROUP myfg
  OBJECT_TYPES = DATABASES
  ALLOWED_DATABASES = db1
  ALLOWED_ACCOUNTS = myorg.myaccount2
  REPLICATION_SCHEDULE = '10 MINUTE'
  OPTIMIZED_REFRESH = TRUE;

What happens in this code:

1. Creates a new failover group named myfg.
2. Sets the object type to databases and restricts the target account.
3. Establishes a replication schedule of 10 minutes.
4. Enables the optimized refresh property.

To enable RPO Assurance on an existing primary failover group:

ALTER FAILOVER GROUP myfg SET RPO_ASSURANCE = TRUE;

What happens in this code:

1. Alters the existing failover group on the primary account.
2. Enables RPO Assurance, removing any concurrent manual schedules.
3. Initiates Snowflake-managed continuous replication refreshes.

Limitations and requirements

• Generally available (GA).
• Requires Snowflake Business Critical Edition or higher.
• Optimized Refresh requires a REPLICATION_SCHEDULE set to an interval of no more than 6 hours.
• RPO Assurance and manual replication schedules are mutually exclusive.
• Existing failover groups continue running as Replication Classic by default.