As originally reported here – a recent update has caused on prem exchange servers to become unstable and potentially requiring regular reboots to continue operating. At the time there was nothing about this particular problem available on line – so diagnosing the problem took extended amounts of time and effort while there was radio silence from Support.
The official response line is here from Microsoft with the generic symptoms
“Symptoms that can be seen are various, but here are some:
- Exchange Server OWA may fail with an error MapiExceptionMdbOffline: Unable to make connection to the server. (hr=0x80004005, ec=1142) with event ID 1006 logged in the Application Log.
- Exchange email queues might grow during the day, with email crawling to a halt.
- Various issues related to DAG members not being able to reach quorum have been reported too.
- Exchange server reboot might resolve the issues temporarily.
- Exchange Server 2016, 2019 or Subscription Edition (SE) might be impacted.”
The main criticism though is that it took so long for Microsoft to acknowledge there was a problem and to offer suggestions for dealing with. Indeed there were some comments around the suggestion that Microsoft had trouble replicating the problem due to a lack on on prem installations to test.
The workaround when it landed was
- Do not uninstall October 2026 Windows Server updates.
- From an elevated Command Prompt on an Exchange Server, run:
reg add "HKLM\SOFTWARE\Microsoft\Rpc" /v EpmapSavedContextCap /t REG_DWORD /d 65536 /f A reboot is recommended and monitoring to see if the issue is resolved.