Why is the MRP process not running in Oracle Data Guard?

The MRP process stops in Oracle Data Guard for five main reasons: manual cancellation during maintenance, missing Standby Redo Logs preventing real-time apply, Data Guard Broker setting apply state to OFF, protection mode conflict causing MRP abort, or no auto-start after standby instance crash. Each cause has a different fix command.

How do I restart the MRP process in Oracle Data Guard 19c?

Run on the standby: ALTER DATABASE RECOVER MANAGED STANDBY DATABASE USING CURRENT LOGFILE DISCONNECT FROM SESSION; Then verify with: SELECT process, status FROM v$managed_standby WHERE process='MRP0'; If using Data Guard Broker, run in DGMGRL: EDIT DATABASE standby_db SET STATE='APPLY-ON';

What is the difference between MRP not running and ORA-16766?

They describe the same condition from different perspectives. ORA-16766 is the error code Oracle reports in DGMGRL SHOW CONFIGURATION output. MRP not running is the underlying technical reason for that error. Both mean the Managed Recovery Process on the standby has stopped and redo is no longer being applied.

Why does MRP stop immediately after I restart it?

Recurring MRP stop after restart points to three causes: Data Guard Broker has apply set to OFF and overrides your manual restart (fix via DGMGRL), missing Standby Redo Logs cause MRP to fail on start (create SRLs first), or Maximum Protection mode with an unstable standby connection causes repeated abort (lower to Maximum Availability).

How do I prevent MRP from stopping after a standby reboot?

Create an AFTER STARTUP database trigger on the standby that checks the database role and automatically runs ALTER DATABASE RECOVER MANAGED STANDBY DATABASE USING CURRENT LOGFILE DISCONNECT. Alternatively, enable Data Guard Broker with dg_broker_start=TRUE and set the database state to APPLY-ON, which makes the broker auto-restart MRP after any restart.

MRP Process Not Running in Data Guard: Fix It Step-by-Step (Oracle 19c)

MRP Process Not Running in Data Guard? Fix It Step-by-Step (Oracle 19c)

5 Root Causes, DGMGRL Diagnosis and Exact Fix Commands for Every Scenario

📅 April 2026

👤 Chetan Yadav, Senior Oracle & Cloud DBA, Oracle ACE Apprentice

⏱️ 12,14 min read

⏱️ Estimated Reading Time: 12,14 minutes

MRP Not Running, ORA-16766, DGMGRL Fix, SRL Creation, Broker State, Auto-Start Trigger, Oracle 19c

mrp process not running data guard oracle 19c server infrastructure troubleshooting guide

⚙️ Environment Referenced in This Article

Oracle Database: 19.18.0.0.0 Enterprise Edition • Primary: 2-Node RAC, 4.8 TB OLTP, 2,800 TPS
Standby: Physical Standby with Active Data Guard enabled
Protection Mode: Maximum Availability (SYNC/AFFIRM) • Broker: Data Guard Broker enabled

The monitoring alert arrived at 2:48 AM: "Standby apply lag crossing 90 minutes." I connected to DGMGRL immediately. SHOW CONFIGURATION confirmed it: the MRP process was not running on the standby. Every transaction committed on the primary for the past 90 minutes was sitting unprocessed in Standby Redo Logs, and the gap was growing by the second.

In my 15 years managing Oracle production environments, a stopped MRP process is one of the most common Data Guard incidents I have resolved. It is not complicated once you know which of the five root causes you are dealing with. The problem is that each cause has a completely different fix, and applying the wrong one wastes critical time.

This guide gives you the exact decision path, the diagnostic commands to identify your specific cause, and the precise fix for each scenario. In most cases the MRP process not running in Data Guard is resolved in under 5 minutes.

Chetan Yadav

Pages

Monday, May 11, 2026

MRP Process Not Running in Data Guard: Fix in Oracle 19c

MRP Process Not Running in Data Guard? Fix It Step-by-Step (Oracle 19c)