2

im having problems getting my DC (PDC Role as it is the only DC in the network) to sync time with an external ntp server. My Server is running Windows Server 2016 x64 fully patched. The Windows Firewall is completely disabled just to rule out possible isues. EDIT: It is not a VM, just a normal physical Supermicro 2U Server.

I keep getting the following message:

C:\Documents and Settings\Administrator>w32tm /resync /rediscover
Sending resync command to local computer...
The computer did not resync because no time data was available.

I tried configuring NTP the following ways:

Using w32tm

PS C:\Users\Administrator> w32tm.exe /config /manualpeerlist:”192.168.1.10 1.europe.pool.ntp.org 2.europe.pool.ntp.org 3.europe.pool.ntp.org” /syncfromflags:manual /reliable:YES /upda

te Der Befehl wurde erfolgreich ausgeführt. PS C:\Users\Administrator> w32tm.exe /config /update Der Befehl wurde erfolgreich ausgeführt. PS C:\Users\Administrator> Restart-Service w32time

Using GPO's using these guides (tried them both)

EDIT: After any changes to the GPO's I did issue gpupdate /force to enforce the GPO


no matter which way I try to implement it I always get the following results:

Checking no other application is blocking port 123

C:\Users\Administrator>netstat -o -a -n | find ":123"
UDP    0.0.0.0:123            *:*                                    1248
UDP    [::]:123               *:*                                    1248

Verifying that connection to the ntp server is working using /stripchart

C:\Users\Administrator>w32tm /stripchart /computer:0.europe.pool.ntp.org /samples:5 /dataonly
0.europe.pool.ntp.org wird verfolgt [195.46.37.22:123].
5 Proben werden gesammelt.
Es ist 26.01.2018 11:55:40.
11:55:40, -52.4480780s
11:55:42, -52.4482309s
11:55:44, -52.4482593s
11:55:46, -52.4484657s
11:55:49, -52.4482562s

w32tm configuration

C:\Users\Administrator>w32tm /query /configuration
[Konfiguration]

EventLogFlags: 2 (Lokal)
AnnounceFlags: 5 (Lokal)
TimeJumpAuditOffset: 28800 (Lokal)
MinPollInterval: 6 (Lokal)
MaxPollInterval: 10 (Lokal)
MaxNegPhaseCorrection: 172800 (Lokal)
MaxPosPhaseCorrection: 172800 (Lokal)
MaxAllowedPhaseOffset: 300 (Lokal)

FrequencyCorrectRate: 4 (Lokal)
PollAdjustFactor: 5 (Lokal)
LargePhaseOffset: 50000000 (Lokal)
SpikeWatchPeriod: 900 (Lokal)
LocalClockDispersion: 10 (Lokal)
HoldPeriod: 5 (Lokal)
PhaseCorrectRate: 7 (Lokal)
UpdateInterval: 100 (Lokal)


[Zeitanbieter]

NtpClient (Lokal)
DllName: C:\Windows\SYSTEM32\w32time.DLL (Lokal)
Enabled: 1 (Lokal)
InputProvider: 1 (Lokal)
AllowNonstandardModeCombinations: 1 (Lokal)
ResolvePeerBackoffMinutes: 15 (Lokal)
ResolvePeerBackoffMaxTimes: 7 (Lokal)
CompatibilityFlags: 2147483648 (Lokal)
EventLogFlags: 1 (Lokal)
LargeSampleSkew: 3 (Lokal)
SpecialPollInterval: 3600 (Lokal)
Type: NTP (Lokal)
NtpServer: 0.192.168.1.10 1.europe.pool.ntp.org 2.europe.pool.ntp.org 3.europe.pool.ntp.org (Lokal)

NtpServer (Lokal)
DllName: C:\Windows\SYSTEM32\w32time.DLL (Lokal)
Enabled: 1 (Lokal)
InputProvider: 0 (Lokal)
AllowNonstandardModeCombinations: 1 (Lokal)

w32tm status

C:\Users\Administrator>w32tm /query /status
Sprungindikator: 0(keine Warnung)
Stratum: 1 (Primärreferenz - synchron. über Funkuhr)
Präzision: -6 (15.625ms pro Tick)
Stammverzögerung: 0.0000000s
Stammabweichung: 10.0000000s
Referenz-ID: 0x4C4F434C (Quellname:  "LOCL")
Letzte erfolgr. Synchronisierungszeit: 26.01.2018 11:32:13
Quelle: Local CMOS Clock
Abrufintervall: 6 (64s)

The source will always remain at Local CMOS Clock The following commands did not change anything:

  • w32tm /resync
  • w32tm /resync /rediscover
  • w32tm /resync /rediscover /nowait

In between "Tries" I did the following to reset my settings:

net stop w32time
w32tm /unregister
w32tm /register
net start w32time

Basicly no matter what I do I cannot get my DC to sync time with any NTP server (even tried my router using 192.168.1.10 because it runs an integrated NTP-Server but no luck as well)

Any ideas on how to fully reset NTP settings completely other than using w32tm /unregister && w32tm /register and rebooting?

EDIT: Server Configuration seems fine, when connected to my router (ntp-server) it works just fine.

The Switch is a HPE OfficeConnect Switch 1850-24G 2XGT (JL170A)

Any idea why the switch is "blocking" ntp traffic despite the fact that w32tm /stripchart is working fine?

regards

9
  • 1
    How long did you wait after committing the changes? I discovered that Windows Server takes forever to update its timesettings sometimes. I'd delete the list, set w32tm /config /syncfromflags:manual /update /reliable:yes /manualpeerlist:"0.de.pool.ntp.org" and check HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\W32Time and subkeys if something is wrong there. And reset w32time service of course.
    – Lenniey
    Jan 26, 2018 at 12:17
  • I waited a few minutes but also trying to force an resync with w32tm /resync /rediscover /nowait anda fter giving up last night (having the GPO in place) nothing changed overnight...
    – Alkahna
    Jan 26, 2018 at 12:34
  • Ah yes, forgot that: disable the GPOs and force a gpupdate via gpupdate /force. Try it locally first. Also check if some other GPO overwrites your settings gpresult /R and gpresult /H gpresult.html
    – Lenniey
    Jan 26, 2018 at 12:38
  • edited my post. I issued gpupdate /force after every change to the GPO. I generated the GPO result via gpresult /H gpresult.html like you suggested. There is no sign of any ntp settings left as I now (again) try to configure NTP via w32tm commands
    – Alkahna
    Jan 26, 2018 at 13:01
  • You can also try to just delete / reset all ntp settings in registry and start over (as last resort). But be careful!
    – Lenniey
    Jan 26, 2018 at 13:03

2 Answers 2

3

I found the Problem!

The server is connected to an HPE OfficeConnect 1850-24G 2XGT (JL170A).

This switch has a 'Security' feature named "Prevent UDP Blat Attack" after disabling said feature NTP works like a charm with both ways of configuration (cmd or GPO)

For reference, I found an article (different switch model though) describing the problem: http://www.gadjev.com/2014/06/23/auto-dos-feature-on-hp-v1810-switches-blocks-legitimate-network-traffic/

2
  • Wow. That was a weird one. Nice catch.
    – joeqwerty
    Jan 27, 2018 at 22:48
  • Yeah spent way too much time on looking at the server rather than other network components. Hopefully this will help others to help troubleshoot their problems :-)
    – Alkahna
    Jan 28, 2018 at 0:30
2

I suspect that after resetting the Windows Time service that you didn't reconfigure it. Resetting it sets it back to default. You'll need to reconfigure it to your specific needs, as such:

w32tm.exe /config /manualpeerlist:time.windows.com /syncfromflags:manual /reliable:YES /update

using whichever time sources you want to use.

4
  • after every reset I did configure it again to test if the new configuration would work.
    – Alkahna
    Jan 26, 2018 at 16:36
  • OK. I'm at a loss then.
    – joeqwerty
    Jan 26, 2018 at 16:37
  • This does work, but not immediately. My server clock was 5 minutes behind, I ran your command in command prompt, it said it was successful. The time didn't change but when I clicked the clock (bottom right in desktop view) I noticed it was ticking very quickly. When it caught up to the actual time it started ticking at a normal pace.
    – JJJ
    Apr 13, 2019 at 19:51
  • @JJJ This is expected behavior. The system clock tick duration is quartered or halved (depending on the size of the gap between current and correct time) which allows the clock to catch up gradually and avoid the side effects of a sudden large jump in time. Aug 6, 2019 at 18:13

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .