Jan 11 2022 05:49 PM
Jan 11 2022 05:49 PM
This is my first post here, ment as advice / troubleshoot tips for Exchange admins.
I see regularly Exchange admins having trouble with IIS 500 page errors. This topic is not rather a question, but some information I would like to share with admins that run into an IIS500 Error (reason: Maintenance) on Exchange OWA and ECP after installing .NET Updates. The page of ECP (and OWA) looks like the follow page below.
Please note that IIS 500 can also be triggered by other errors, this is for specific the situation written below. If you have recreated the virtual directories, you find out that the problem is still there.
The example here is based on Exchange 2019 / Windows Server 2019, but I have also seen this behavior on Exchange Server 2016 and Exchange Server 2013 on Windows Server 2016 and 2012R2.
Rule one: Give Exchange servers some time booting up, don't try to immediately logon to ECP/OWA when the logon screen beach appears on the console. Check the CPU up-time, and I recommend waiting at least wait for 5 to 10 minutes uptime (CPU time) and verify the server CPU usage has normalized.
IIS 500 common error checks
Although IIS 500 is quite a common error, there are some occasions where you should figure out if you have installed Windows / .NET Updates on the server you are working on. Please examine the server eventlog for details in the “Windows Logs > Setup” section.
If so? You should first verify that all Exchange Services are running correctly that have the startup type to start automatically. In environments we manager we regularly see the following services not starting after a reboot of the server:
You will also find:
What should you do?
First, start the Exchange Services that aren’t running and have startup type set to automatically. Start with back-end services like MS Exchange Replication and give it some time and lastly start services that are for log searching or auditing. When observing your performance monitor you will probably see a surge in resource usage.
In the Eventlog > Application you will see some Exchange things to come to life again. After a few minutes, when you login to ECP you will see the environment again (note: the first time someone logs in to a virtual directory the web-pool needs to start, this might take a few seconds).
Restarted the server and it’s broken again!
If you restart the server again, you will probably run into the same situation again. Why? The root cause is not known to me, but it has to do something with .Net Framework updates. What you need to do is basically waiting. So, don’t reboot the server if you have your pages working again by starting the services, etc. Rebooted the server? You can do the actions written above again.
.Net Optimization Proces
For it to permanently work after reboots you have to wait for a process to start working on the .Net optimization. When the server is “idle”, usually 20-30 minutes after booting up the process “.Net Optimization Task” starts running on your Exchange server. This usually takes up to 20-40 minutes. When it’s done you will see most of the time (not always) notifications in eventlog:
Application > .NET Runtime Optimization Service > EventID 1130.
.NET Runtime Optimization Service (4.0.30319.0) - Installed from repository: mscorlib
When the task is completed and you restart your server again, it should be running fine again. You don’t have to reboot the server If you have started the services manually. The key is they don’t want to start automatically unless the “.Net Optimization Task” has done it’s thing.
What actually happens, I don’t know, maybe some of the Exchange Team members can give an answer, but this is what I’ve found out in many environments (test production, etc). It doesn’t matter if you install updates with Antimalware software switched on or off.
I hope this will help some users having trouble and starting to fix thing’s that actually aren’t broken, but just need some special attention.