What is Oracle Data Guard?
In the world of high-availability databases, Oracle Data Guard is the ultimate insurance policy. If your primary database is the engine of your business, Data Guard is the "backup engine" that is already synchronized and ready to take over the moment the first one fails.
It is a set of services that create, maintain, and monitor one or more Standby Databases to protect your data from failures, disasters, and human error.
The core of Data Guard is the movement of Redo data (the record of every change made to the database).
Capture: The Primary Database generates redo data as transactions happen.
Ship: Data Guard transmits that redo data across the network to the Standby Database.
Apply: The Standby Database receives the redo and "replays" it, keeping the standby in sync with the primary.
Data Guard offers different "flavors" of standby databases depending on your needs:
Physical Standby: A bit-for-bit identical copy of the primary database. It stays in "Recovery Mode" and applies changes at the block level. This is the most robust for disaster recovery.
Logical Standby: It contains the same data but can have a different physical structure (like extra indexes). It transforms redo into SQL statements and executes them. (Used less often today).
Snapshot Standby: A physical standby that you "pause" and open for read-write testing. When you're done, you "flash back" and it resyncs with the primary.
In a standard Data Guard setup, the standby is "mounted" but not "open"—it’s just sitting there waiting for a disaster.
Active Data Guard is a licensed option that allows you to open the standby in Read-Only mode while it is still applying changes.
Offload Reporting: Run your heavy "End of Month" reports on the standby instead of slowing down the primary.
Backups: Run your RMAN backups against the standby to save CPU on the primary.
Data Guard allows you to choose how strictly the two databases must be synced:
| Mode | Priority | Behavior |
| Maximum Performance | Speed | The primary commits transactions without waiting for the standby to acknowledge. (Standard default). |
| Maximum Availability | Balance | The primary waits for the standby to confirm receipt. If the network fails, it keeps running in "un-sync" mode. |
| Maximum Protection | Zero Data Loss | The primary will shut down if it cannot confirm that the redo has reached the standby. Zero risk, but high stakes. |
This is the most important distinction in Data Guard:
Switchover (Planned): You swap roles for maintenance. The Primary becomes the Standby, and the Standby becomes the Primary. There is zero data loss.
Failover (Unplanned): The primary server is on fire. You force the Standby to become the new Primary. Depending on your protection mode, there might be a tiny amount of data loss.
Disaster Recovery (DR): Survive a data center outage.
Data Protection: Protect against corruption (Data Guard validates redo blocks before applying them).
High Availability: Perform hardware upgrades on the standby, then "switchover" to keep uptime high.
Reporting: Offload read-only traffic (with Active Data Guard).
Peer Tip: Don't confuse Data Guard with Real Application Clusters (RAC).
RAC protects against server failure (multiple servers hitting one database).
Data Guard protects against database/site failure (multiple copies of the data).