If you dread the process of updating Exchange and the possibility that patches will break the servers, you can take control by putting best practices in place with three steps.
You probably want to avoid thinking about Exchange updates until they are knocking on the front door, but preparation outside of the regular schedule makes the process easier.
Microsoft releases cumulative Exchange updates each quarter to enhance functionality and security fixes every month on Patch Tuesday. If you get behind on updates, you only need the most recent update to get caught up.
Start with these steps to build your own method to update Exchange Servers following best practices.
Step 1: Create good habits around Exchange updates
Check your servers to get to know them. Understanding the way your servers normally run and recording a performance baseline will give you an accurate comparison of how a patch affects performance.
Plan a schedule to update Exchange regularly without disrupting applications. Keep the servers one or two versions behind at most. A regular cumulative update can also have a prerequisite -- such as a newer .NET Framework version -- that you must install, which can take additional time.
Before you apply any patches, learn how to restore Exchange from a backup to prepare for the possibility of a problematic update. Ensure you have the most recent information to revert back to by creating a server backup immediately before issuing patches and updates.
Step 2: Check for information on specific Exchange patches
Even when you do everything right, Exchange updates can cause problems. Sometimes patches break the system. Organizations can prepare by setting a standard procedure to handle a faulty patch.
Even when you do everything right, Exchange updates can cause problems.
First check the information Microsoft has published about the patches. Some Exchange MVPs use their early access to test patches and share their findings on their blogs. You may find that an MVP has already tested your production setup, but you should still allot yourself enough time to test Exchange patches. Microsoft and MVPs can still miss a bug that affects your servers, firmware or network. Look for compatibility with essential services and confirm everything runs properly after it's back online. Should patching go wrong, schedule enough time to recover the servers without disrupting users' work.
Step 3: Put safety first with maintenance mode
When you are ready to issue Exchange updates, put the deployment -- including redundant servers -- into maintenance mode. If you do not, you risk causing the database availability group (DAG) to fail over. Maintenance mode prevents interruptions to users from any maintenance you perform. The script drains the node and the server's transport queues.
In Exchange 2016, admins can set a period of time for the database to move back to its preferred node. The default time is one hour, but patching servers can take more time. You must reboot the server after applying the update. Without maintenance mode, the DAG can break and cease to function. You should run maintenance mode for any work you do, even just a server restart. After you complete the update, take the server out of maintenance mode and check the logs to ensure the server's health before moving on to another node that needs patching.
If you are at a loss about what to do, you can open a support case with Microsoft or get help from a consultant.
