Responder ServiceHealthMSExchangeReplForceReboot causes Exchange Server to restart with Stop Error

Original KB number:   2969070

Symptoms

Exchange Server 2016 or Exchange Server 2013 restarts frequently and displays the following Stop error on a blue screen.

bugcheck
Bugcheck code 000000EF
Arguments fffffa801a5f9980 0000000000000000 0000000000000000 0000000000000000

The restarts are initiated by the Managed Availability responder ServiceHealthMSExchangeReplForceReboot.

The recovery action event is observed in RecoveryActionResults:

Log Name: Microsoft-Exchange-ManagedAvailability/RecoveryActionResults
Source: Microsoft-Exchange-ManagedAvailability
Date: DateTime
Event ID: 500
Task Category: Recovery
Level: Information
Keywords:
User: SYSTEM
Computer: Exch1.contoso.com
Description:
Recovery Action Started. (ActionId=ForceReboot, ResourceName=Exch1, Requester=ServiceHealthMSExchangeReplForceReboot, InstanceId=140428.105409.22538.004, ExpectedToFinishAt=2014-04-29T04:59:09.2253815Z

And associated Probe failure message:

Log Name: Microsoft-Exchange-ActiveMonitoring/ProbeResult
Source: Microsoft-Exchange-ActiveMonitoring
Date: DateTime
Event ID: 2
Task Category: Probe result
Level: Error
Keywords:
User: SYSTEM
Computer: Exch1.contoso.com
Description:
Probe result (Name=ServiceHealthMSExchangeReplEndpointProbe/TCP/MSExchangeRepl)
<Error>Check 'Microsoft.Exchange.Monitoring.TcpListenerCheck' thrown an Exception! Exception - Microsoft.Exchange.Monitoring.ReplicationCheckHighPriorityFailedException: High priority check TcpListener failed. Error: The health test of the TCP listener for the Microsoft Exchange Replication service on server 'Exch1' failed. This server cannot participate in replication until the error is resolved. Error: Couldn't determine the IP address for server 'Exch1' because DNS didn't return any information.
at Microsoft.Exchange.Monitoring.ReplicationCheck.FailInternal()
at Microsoft.Exchange.Monitoring.ReplicationCheck.Fail(LocalizedString error)
at Microsoft.Exchange.Monitoring.TcpListenerCheck.InternalRun()
at Microsoft.Exchange.Monitoring.ReplicationCheck.Run()
at Microsoft.Exchange.Monitoring.ActiveMonitoring.HighAvailability.Probes.ReplicationHealthChecksProbeBase.RunReplicationCheck(Type checkType)
Check 'Microsoft.Exchange.Monitoring.TcpListenerCheck' did not Pass!
Detail Message - The health test of the TCP listener for the Microsoft Exchange Replication service on server 'Exch1' failed. This server cannot participate in replication until the error is resolved. Error: Couldn't determine the IP address for server 'Exch1' because DNS didn't return any information.
</Error>

Cause

Misconfigured network adapter on Exchange Server causes failure in Domain Name System (DNS) name resolution.

Resolution

Make sure that the Register this connection's addresses in DNS property is selected on the network adapter:

Review the guidelines provided in the following article and make sure that the network adapter that represents the MAPI and Replication networks are configured accordingly.

Planning for high availability and site resilience

More information

Use the following PowerShell command on the affected Exchange server to ensure the server restarts are indeed caused by the ServiceHealthMSExchangeReplForceReboot responder.

Get-WinEvent -LogName Microsoft-Exchange-ManagedAvailability/* | where {$_.Message -like "*ServiceHealthMSExchangeReplForceReboot*"} | ft Message,TimeCreated

Or, you can search the crimson channel log named Recovery Action Logs and Remote Action Logs for word ForceReboot to find the responder that causes the restart.