Close Menu
Oracle RAC

PRVG-01017 – How to Fix “NTP Configuration File” in Oracle RAC

HenriqueBy Updated:7 Mins Read

This post is also available in: Português (Portuguese (Brazil))

When running the runcluvfy.sh (Cluster Verification Utility) before installing Oracle Grid Infrastructure, it is common to encounter the following error:

Verifying Network Time Protocol (NTP) ...FAILED
PRVG-01017 : NTP configuration file "/etc/ntp.conf" is present on
             nodes "node1,node2" on which NTP daemon or service
             was not running

At first glance, the fix seems obvious: the file exists but the service isn’t running. But the real cause is usually something else – and renaming the file without understanding the context can create worse problems than the original warning.

In this article, I’ll explain why this error occurs, what the relationship is between NTP and Chrony, and how to resolve it correctly and safely.

In practice: In most cases on RHEL/OL 7+, this error appears because the system uses Chrony for time synchronization, but the file /etc/ntp.conf was left over from a previous configuration. The CVU finds the file and expects the ntpd service to be active – if it isn’t, it reports the error.

Why Error PRVG-01017 Occurs

The Cluster Verification Utility (CVU) performs a series of validations before installing the Grid Infrastructure. One of them is to check the time synchronization between the cluster nodes.

The logic of the ULC is:

  1. Check if the file /etc/ntp.conf exists
  2. If it exists, check that the ntpd service is running
  3. If the file exists but the service is not active, report PRVG-01017

The problem is that on modern Linux distributions(RHEL 7+, OL 7+, CentOS 7+), the default time synchronization service has changed from NTP (ntpd ) to Chrony (chronyd). Chrony uses its own configuration file (/etc/chrony.conf), but the file /etc/ntp.conf may have remained on the system as a result of a previous installation or OS upgrade.

Why Time Synchronization is Critical in RAC

Oracle Clusterware depends on precise synchronization between nodes. A clock skew between nodes can cause this:

  • CSS (Cluster Synchronization Services) failures
  • Node evictions – nodes being removed from the cluster
  • Data consistency problems in distributed transactions
  • Intermittent errors difficult to diagnose

That’s why the CVU must validate the time sync – and why it’s important to solve the root cause, not just silence the warning.

Diagnosis – NTP or Chrony?

Before correcting, identify which time sync service is active in your environment.

Check if Chrony is active

systemctl status chronyd

If it returns active (running), Chrony is the time sync service in your environment.

To confirm that it is synchronizing correctly:

chronyc tracking

Expected output (example):

Reference ID    : A1B2C3D4 (ntp-server.example.com)
Stratum         : 3
Ref time (UTC)  : Thu Feb 27 10:30:00 2026
System time     : 0.000012345 seconds fast of NTP time
Last offset     : +0.000003456 seconds

If Stratum is between 1-15 and System time is in the millisecond range, the time sync is working.

Check if NTP is active

systemctl status ntpd

If it returns inactive (dead) or not found, ntpd is not in use.

Check if /etc/ntp.conf exists

ls -la /etc/ntp.conf

If Chrony is active, NTP is inactive, and /etc/ntp.conf exists – that’s the cause of PRVG-01017.

Solutions

Solution 1: Remove the residual /etc/ntp.conf (when using Chrony)

This is the most common solution. If Chrony is up and running, /etc/ntp.conf is a residual file that can be safely removed.

Step 1 – Confirm that Chrony is active on ALL nodes:

# Executar em cada nó
systemctl status chronyd | grep Active

Expected result: Active: active (running)

Step 2 – Rename the file on ALL nodes (don’t delete, make a backup):

# Node 1
mv /etc/ntp.conf /etc/ntp.conf.bkp

# Node 2
mv /etc/ntp.conf /etc/ntp.conf.bkp

Important: Rename instead of deleting. If something goes wrong, you can quickly restore the original file.

Step 3 – Validate by running the CVU again:

./runcluvfy.sh stage -pre crsinst -n node1,node2 -verbose | grep -i ntp

Expected result:

Verifying Network Time Protocol (NTP) ...PASSED

Solution 2: Configure and start NTP (when NOT using Chrony)

If your environment doesn’t use Chrony and you need NTP, the solution is to configure and start the ntpd service.

Step 1 – Install NTP (if necessary):

yum install ntp -y

Step 2 – Configure /etc/ntp.conf:

# Verificar se o servidor NTP está configurado
cat /etc/ntp.conf | grep -i server

If there are no servers configured, add them:

# Exemplo com servidores NTP públicos
server 0.pool.ntp.org iburst
server 1.pool.ntp.org iburst
server 2.pool.ntp.org iburst

In corporate environments: Use your organization’s internal NTP servers, not public ones.

Step 3 – Disable Chrony (if installed) and enable NTP:

# Desativar Chrony
systemctl stop chronyd
systemctl disable chronyd

# Ativar NTP
systemctl enable ntpd
systemctl start ntpd

Step 4 – Check synchronization:

ntpq -p

Step 5 – Validate with the ULC:

./runcluvfy.sh stage -pre crsinst -n node1,node2 -verbose | grep -i ntp

Solution 3: Use CTSS (Oracle Cluster Time Synchronization Service)

If you don’t want to use NTP or Chrony, Oracle offers CTSS – its own time synchronization service that is part of the Grid Infrastructure.

For the CTSS to work:

  1. Remove or rename /etc/ntp.conf on all nodes
  2. Disable both ntpd and chronyd
  3. Grid Infrastructure will detect the absence of external time sync and activate CTSS automatically during installation

To check that CTSS is active after installation:

crsctl check ctss

Note: Oracle recommends using NTP or Chrony for production. CTSS is an alternative, but it depends on the clusterware being functional.

Quick Diagnostic Checklist

# 1. Qual serviço de time sync está ativo?
systemctl status chronyd
systemctl status ntpd

# 2. O /etc/ntp.conf existe?
ls -la /etc/ntp.conf

# 3. O Chrony está sincronizando corretamente?
chronyc tracking

# 4. Validar com o CVU
./runcluvfy.sh stage -pre crsinst -n node1,node2 -verbose | grep -i ntp

Conclusion

Error PRVG-01017 is a common warning during the pre-installation validation of Oracle RAC. In the vast majority of cases, the cause is a residual /etc/ntp.conf file in environments that have migrated to Chrony.

The fix is simple, but the critical point is: never remove the file without first confirming that a time sync service is active. Clock skew in RAC causes serious problems that are difficult to diagnose.

The correct diagnosis follows this order:

  1. Identify which service is active (Chrony, NTP or none)
  2. Make sure the time sync is working
  3. Remove the residual file (if applicable)
  4. Validate with runcluvfy.sh

References:

Share.

Related Posts

Add A Comment
Leave A Reply