Role Transitions in Oracle Data Guard

Seamlessly switch roles between primary and standby databases to ensure high availability and disaster recovery.

What is a Role Transition?

In Oracle Data Guard, a Role Transition is when the role of a Primary database and one of its Standby databases is swapped or promoted, depending on the situation:

• 🔁 Switchover → Planned, no data loss
• ❗ Failover → Unplanned, recovery after failure

1. Introduction to Role Transitions

Oracle Data Guard allows either:

• The Primary database to become a Standby, and
• A Standby to take over as the Primary.

This helps maintain zero or minimal downtime during:

• Maintenance
• Disaster Recovery
• Planned migration or upgrades

2. Preparing for a Role Transition

Before performing a role transition:

✅ Ensure:

• The standby is fully synchronized
• All configurations are correct
• Flashback Database is enabled (for reversibility)
• Monitoring tools (like dgmgrl or OEM) are in place

Use:

SELECT SWITCHOVER_STATUS FROM V$DATABASE;

If the result is TO STANDBY or TO PRIMARY, it’s ready.

3. Choosing the Target Standby for Role Transition

If you have multiple standby databases, choose one based on:

• ✔️ Sync level (real-time or lagging?)
• ✔️ Location (same region or DR site?)
• ✔️ Performance capacity
• ✔️ Recovery time objective (RTO)

👉 Use Data Guard Broker (dgmgrl) to get health and sync status of each standby.

🔁 4. Switchovers

A planned role transition — no data loss.

• Performed when both primary and standby are healthy
• Reversible
• Used during patching, upgrades, or testing DR setups

✨ Steps:

1. Verify readiness
2. Issue switchover command
3. Monitor the transition
4. Confirm new roles

DGMGRL> SWITCHOVER TO standby_db;

5. Failovers

An unplanned role transition — usually due to primary failure.

• Performed when the primary is not accessible
• May involve data loss (depending on protection mode)
• Cannot automatically go back to original state without reinstating

Steps:

1. Confirm primary is inaccessible
2. Promote the best standby
3. Reinstate the original primary later (if recoverable)

DGMGRL> FAILOVER TO standby_db;

⚙️ 6. Role Transition Triggers

Role transitions can be:

7. Transitions with Physical Standby

🔄 Switchover to Physical Standby:

• MRP must be stopped
• Redo apply must be in sync
• Very fast transition

❗ Failover to Physical Standby:

• Use ACTIVATE STANDBY DATABASE
• May require applying missing logs
• Reinstate primary if Flashback is enabled

8. Transitions with Logical Standby

🔄 Switchover to Logical Standby:

• SQL Apply must be stopped
• May require checking data compatibility
• More flexible, but slightly slower

❗ Failover to Logical Standby:

• Use ALTER DATABASE ACTIVATE LOGICAL STANDBY DATABASE
• Ensure standby is fully caught up

⚠️ Not all failovers are supported for Logical Standby if major DDL differences exist.

9. Using Flashback Database After Role Transition

Flashback allows you to reverse role transitions quickly:

Requirements:

• Flashback must be enabled
• Fast recovery area configured

✅ Summary Table