Merge pull request #251244 from axelgMS/patch-51

Update planned-maintenance.md

Author: prmerger-automator[bot]

articles/aks/planned-maintenance.md

@@ -308,6 +308,14 @@ az aks maintenanceconfiguration delete -g myResourceGroup --cluster-name myAKSCl
 
   Yes, you can run all three configurations i.e `default`, `aksManagedAutoUpgradeSchedule`, `aksManagedNodeOSUpgradeSchedule`simultaneously.  In case the windows overlap AKS decides the running order. 
 
+* I configured a maintenance window, but upgrade didn't happen - why?
+
+  AKS auto-upgrade needs a certain amount of time to take the maintenance window into consideration. We recommend at least 6 hours between the creation/update of the maintenance configuration, and when it's scheduled to start.
+
+* AKS auto-upgrade didn't upgrade all my agent pools - or one of the pools was upgraded outside of the maintenance window?
+
+  If an agent pool fails to upgrade (eg. because of Pod Disruption Budgets preventing it to upgrade) or is in a Failed state, then it might be upgraded later outside of the maintenance window. This scenario is called "catch-up upgrade" and avoids letting Agent pools with a different version than the AKS control plane.
+
 *  Are there any best practices for the maintenance configurations?
 
   We recommend setting the [Node OS security updates][node-image-auto-upgrade] schedule to a weekly cadence if you're using `NodeImage` channel since a new node image gets shipped every week and daily if you opt in for `SecurityPatch` channel to receive daily security updates. Set the [auto-upgrade][auto-upgrade] schedule to a monthly cadence to stay on top of the kubernetes N-2 [support policy][aks-support-policy].