Want to protect your cyber security and still get fast solutions? Ask a secure question today.Go Premium

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 342
  • Last Modified:

NTP server, one DC updates correctly another adds 4 minutes

I have two data centers in two different physical locations about 60 miles apart (in the same timezone).  I have them both acting as NTP servers for the respective domains.  They are syncing time to north-america.pool.ntp.org.  When I restart the w32time service both system logs indicate they are receiving good data from north-america.pool.ntp.org.

I then change the time manually so both servers have the same time and run w32tm /resync /nowait on each server.  One server corrects itself to the correct time.  The other server corrects itself to the correct time for just a instant then it suddenly goes 4 minutes into the future.

Every resync does the same thing.  They match for just a second then boom 4 minutes ahead.  Any suggestions on how I find out whats really going on here with this one server?

The server that works correctly is running Server 2003 SP2, the one that is visting the future is 2008R2.
0
-Darvin-
Asked:
-Darvin-
  • 3
1 Solution
 
becraigCommented:
Do you get any events in the system log when this occurs  ?
0
 
-Darvin-Author Commented:
I tried this again.  Manually changing the time and the behavior has changed.  As I said previously.  I would change the time, run the w32tm command it would change to the 'right' time and then forward.  Now, when I change the time I don't even get a chance to run the command before it puts itself back.  It's nearly instant.

System log shows this.

The system time has changed to ¿2013¿-¿02¿-¿08T16:00:00.000000000Z from ¿2013¿-¿02¿-¿08T16:09:30.964103500Z.

The system time has changed to ¿2013¿-¿02¿-¿08T16:00:00.000000000Z from ¿2013¿-¿02¿-¿08T16:00:00.000000000Z.

The system time has changed to ¿2013¿-¿02¿-¿08T16:09:34.765000000Z from ¿2013¿-¿02¿-¿08T16:00:03.796496900Z.

Very strange behavior.
0
 
-Darvin-Author Commented:
I found the solution.  Something I neglected to mention was that the domain controller is a hyper-v guest.  It works so well I often forget it's virtual.  The problem is that hyper-v has it's own time keeping mechanism that it passes along to the guests.  By disabling this feature in the integrated services area hyper-v stopped changing the time.  I don't know where it was getting the bad time from as the hyper-v server was taking time from the domain controller.  Perhaps it was in bios of the hyper-v server.  

In either case that solved the problem.
0
 
-Darvin-Author Commented:
This is the solution.
0

Featured Post

[Webinar] Database Backup and Recovery

Does your company store data on premises, off site, in the cloud, or a combination of these? If you answered “yes”, you need a data backup recovery plan that fits each and every platform. Watch now as as Percona teaches us how to build agile data backup recovery plan.

  • 3
Tackle projects and never again get stuck behind a technical roadblock.
Join Now