Australian cloud for business software — built and run in Sydney since 2002

NetLeverage
NetLeverage

Netlogon event 5719 or Group Policy event 1129 at start-up: network not ready, or a broken trust?

Technical article
Troubleshooting

About this article: This article is reviewed against the cited sources. This is source research, not lab validation. Check the stated versions and your own requirements before making changes.

Netlogon event 5719 reads "This computer was not able to set up a secure session with a domain controller in domain [name]". It sounds serious, and it can be. But Microsoft documents two common cases where it is harmless, and one where it means the computer's trust with the domain really is broken. The work is telling them apart before changing anything.

What you see

One or more of these, usually straight after a restart:

  • Netlogon event 5719 in the System log.
  • Group Policy event 1129, or computer startup scripts that did not run, because Group Policy could not reach a domain controller when the computer started (Microsoft Learn, event 5719 or 1129).
  • On a Windows Server 2025 member server, a single 5719 with the code 0xC00000E5 (STATUS_INTERNAL_ERROR) each time the Netlogon service restarts (Microsoft Learn, 5719 STATUS_INTERNAL_ERROR).

What matters most is whether people can sign in with domain accounts once the computer is up.

Why it happens

Case 1: the network was not ready yet (the trust is fine)

Netlogon and the network start at about the same time. If the network takes longer to come up than Netlogon waits, the computer cannot find a domain controller and logs the failure. Once the network is ready, it tries again and succeeds. Microsoft lists these delays (Microsoft Learn, event 5719 or 1129):

  • Network adapters and switches whose link negotiation and address-uniqueness checks take longer than Netlogon waits, often seen with Gigabit adapters.
  • Network access checks (Network Access Protection, 802.1X authentication, or similar products) that hold the connection back until the computer passes.
  • An automatic DirectAccess connection that takes longer than Netlogon allows.
  • Slow DHCP, so the adapter gets its address late.

On Windows 8.1 and earlier, this shows as event 5719. On Windows 10 and later, 5719 is no longer logged in this situation. Netlogon.log records "cannot pick trusted DC… No IP addresses present, skipping No DC event log" instead, and the visible failures come from components that need a domain controller, mainly Group Policy start-up processing. Microsoft's words: "If you can log on to the domain without a problem, you can safely ignore event ID 5719."

Case 2: a Windows Server 2025 member talking to older domain controllers (harmless)

A Windows Server 2025 member server first tries to set up its secure channel with a newer Kerberos method. Domain controllers running Windows Server 2022 or 2019 do not support it, so the first attempt fails, a single 5719 with 0xC00000E5 is logged, and Windows falls back to the older Netlogon method, which works. Security-NetLogon event 9005 ("Secure channel setup has failed with Kerberos… Falling back to NetLogon") is logged alongside it (Microsoft Learn, 5719 STATUS_INTERNAL_ERROR). Windows Server 2025 members talking to Windows Server 2025 domain controllers do not log it.

Case 3: the trust relationship is broken (a real fault)

If the computer's own password and the one Active Directory holds for it no longer match, the secure channel cannot be set up at all. Microsoft names two common causes: the computer holds an older password than Active Directory, or Active Directory holds an older one than the computer, for example after a domain controller was restored to an earlier state or replication failed (Microsoft Learn, broken trust relationship). This one does not go away after the network comes up.

How to prove the cause

All of these checks are read-only, apart from turning Netlogon debug logging on and off.

1. Can a domain account sign in once the computer has started? If yes, and the events only appear at start-up, you are almost certainly in Case 1 or Case 2.

2. Test the secure channel.

nltest /sc_query:<your domain>

A healthy result names a trusted domain controller. ERROR_ACCESS_DENIED (status 5) is the broken-trust signature, especially together with these signs (Microsoft Learn, broken trust relationship):

  • the sign-in message "The trust relationship between this workstation and the primary domain failed", while local or cached accounts still work;
  • NETLOGON event 3210 ("This computer could not authenticate with \\DCName…").

3. Read the 5719 events and their codes.

wevtutil qe System "/q:*[System[(EventID=5719)]]" /f:text /c:50

A single 0xC00000E5 each time Netlogon restarts on a Windows Server 2025 member, with older domain controllers, is Case 2 (Microsoft Learn, 5719 STATUS_INTERNAL_ERROR). Repeated 5719s outside restarts, with sign-in failures, are not.

4. Turn on Netlogon debug logging to see the timing.

nltest /dbflag:0x2080ffff

Restart, then read %systemroot%\debug\NetLogon.log. In the start-up race, the log shows "Cannot find DC" and "cannot pick trusted DC" first, then the network adapter's address arriving afterwards. Microsoft's example shows getting an address taking more than 15 seconds. Turn logging off afterwards:

nltest /dbflag:0x0

5. Check what sits between the computer and the domain controllers. Note any 802.1X, network access control or DirectAccess in use, and whether the switch port negotiates quickly.

The fix

Case 1: give the network time, or make it come up faster

Start with the least invasive change and make one change at a time (Microsoft Learn, event 5719 or 1129):

  1. Fix the network side first. Install the current driver for the network adapter, or enable PortFast on the switch ports that computers connect to.
  2. Network access checks. Adjust firewall or IPsec policies so a computer that has just passed the access check can reach domain controllers straight away.
  3. Let Group Policy wait for the network. In Group Policy, under Computer Configuration > Administrative Templates > System > Group Policy, set Specify startup policy processing wait time (corporate LAN or WLAN) or Specify workplace connectivity wait time for policy processing (external connections). Base the value on how long the computer actually takes to get a working address.
  4. Registry options Microsoft documents. Back up the registry before using any of them:
    • GpNetworkStartTimeoutPolicyValue (REG_DWORD, decimal 60) under HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Winlogon, to delay Group Policy start-up processing. Increase it if startup scripts still do not run.
    • ExpectedDialupDelay (REG_DWORD, seconds, 0 to 600) under HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\Netlogon\Parameters. Microsoft notes this only helps if the computer already has an IP address, for example while it is held on a quarantine network.
    • NegativeCachePeriod under the same Netlogon key, so Netlogon retries domain controllers sooner after a failure. Microsoft suggests a low value such as three seconds, or 0 on a LAN. Event 5719 is still logged, but it stops causing other problems.
    • ArpRetryCount set to 1 (range 0 to 3, default 3) under HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\TcpIp\Parameters, to shorten the address-uniqueness wait.
  5. Check for a wrongly disabled IPv6. If DisabledComponents is set to 0xfffffff, delete it or change it to the intended 0xff. The wrong value adds a five-second delay at start-up. Microsoft does not recommend disabling IPv6 at all.

Microsoft also lists a Kerberos MaxPacketSize change, but it applies only to Windows XP and Windows Server 2003 and earlier. Leave it alone on current versions.

Case 2: leave it

Microsoft's guidance is not to try any remediation if the secure channel is established and domain operations carry on normally. The event stops once enough Windows Server 2025 domain controllers are in place. Microsoft documents an optional temporary registry setting (UseKerberosForSecureChannels = 0 under HKLM\SYSTEM\CurrentControlSet\Services\NetLogon\Parameters) that suppresses the event during the transition. It must be removed once the newer domain controllers are deployed (Microsoft Learn, 5719 STATUS_INTERNAL_ERROR).

Case 3: repair the trust

Work out which side holds the newer password and follow Microsoft's matching guidance (Microsoft Learn, broken trust relationship). Also check that these registry values hold the short computer name, not the full domain name:

Reg query HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\ComputerName\ComputerName /v ComputerName
Reg query HKEY_LOCAL_MACHINE\SYSTEM\ControlSet001\Services\Tcpip\Parameters /v hostname

Rejoining the computer to the domain is a valid fix, Microsoft notes. But if the trust keeps breaking, find the cause first, such as a restored domain controller or failing replication.

How to confirm the fix

  • After a restart, a domain account signs in normally.
  • nltest /sc_query:<your domain> names a trusted domain controller with a success status.
  • For Case 1: computer startup scripts run and Group Policy start-up processing succeeds, with no new 1129. On older systems, no new 5719 at boot.
  • For Case 2: no new 5719 events other than the single one at each Netlogon restart.

What to watch afterwards

  • Network changes. New switches, new 802.1X or network access control rules, and new adapter drivers can bring the start-up race back.
  • Repeat trust failures. One broken trust after a restore is explainable. Several suggest a restored domain controller or replication problem that needs fixing at the source.
  • Mixed Windows Server versions. As Windows Server 2025 domain controllers are added, check that the Case 2 events stop, and remove any temporary registry suppression.

Risks and limits

  • Do not rejoin a computer to the domain just because 5719 appeared at boot. If domain sign-in works, the trust is fine.
  • Registry changes affect every start-up. Microsoft warns that editing the registry incorrectly can cause serious problems: back it up first and change one value at a time.
  • Suppressing the Case 2 event with UseKerberosForSecureChannels hides only that event. If the error recurs outside Netlogon restarts or comes with sign-in failures, Microsoft says to investigate further, not suppress.

Sources

Next step

Most 5719 events at boot are timing, not trust, and the fix is in the network rather than the domain. If you want start-up failures sorted into the harmless and the real, and the real ones fixed at the cause, NetLeverage has looked after business servers since 2002. Talk to us through managed support with the output of nltest /sc_query from an affected computer.