Merge pull request #226939 from Heidilohr/work-scaling-faq

Added new question to autoscale FAQ.

Author: prmerger-automator[bot]

articles/virtual-desktop/autoscale-faq.yml

@@ -4,7 +4,7 @@ metadata:
   description: Frequently asked questions and best practices for Azure Virtual Desktop.
   author: Heidilohr
   ms.topic: faq
-  ms.date: 04/14/2022
+  ms.date: 02/13/2023
   ms.author: helohr 
   ms.custom: devx-track-azurepowershell
   manager: femila
@@ -48,22 +48,22 @@ sections:
       - question: |
          If I set drain mode on session hosts, will autoscale change my configured setting?
         answer: |
-         Yes, autoscale will still turn VMs in drain mode on or off, no matter who put it in drain mode. Autoscale overrides drain mode on all VMs included in scaling, so if you want to exclude a VM from scaling actions, you must use exclusion tags.
+         Yes, autoscale still turns VMs in drain mode on or off, no matter who put it in drain mode. Autoscale overrides drain mode on all VMs included in scaling, so if you want to exclude a VM from scaling actions, you must use exclusion tags.
       
       - question: |
          How often does autoscale monitor the session hosts and perform scaling evaluations?
         answer: |
-         Autoscale monitors for when users sign in or out of their session hosts and categorizes this activity as session change events. Session change events trigger a scaling evaluation that creates logs. If there aren't any session change events or the event service has an outage, autoscale then checks to see if it missed any events. When several session change events happen within a short time period, then the feature will batch the scaling evaluations. This batching allows autoscale to process large numbers of events quickly without overloading the system.
+         Autoscale monitors for when users sign in or out of their session hosts and categorizes this activity as session change events. Session change events trigger a scaling evaluation that creates logs. If there aren't any session change events or the event service has an outage, autoscale then checks to see if it missed any events. When several session change events happen within a short time period, the feature batches the scaling evaluations. This batching allows autoscale to process large numbers of events quickly without overloading the system.
 
       - question: |
          Can force sign-out happen in any phase of the day?
         answer: |
-         No. If you've enabled autoscale, you can only force users to sign out during the ramp-down phase. This means that if, during ramp-down, a session host is put in drain mode in preparation for it to be shut down but not all users get signed out before the phase changes to off-peak, the remaining user sessions will not be forced to sign out from their session. This is because during off-peak hours, users aren't forced to sign out of their sessions. Instead, autoscale will wait until all users have signed out before deallocating the VM. For example, if the ramp-down phase is 15 minutes long, and the wait time before signing out users and shutting down VMs is 20 minutes long, the schedule will shift to the off-peak phase and the user sessions will not be forced to sign out.
+         No. If you've enabled autoscale, you can only force users to sign out during the ramp-down phase. If you put a session host in drain mode during ramp-down to prepare it to be shut down but not all users sign out before the phase changes to off-peak, the remaining user sessions won't be forced to sign out from their session. The reason users aren't signed out is because autoscale doesn't force users to sign out of their sessions during off-peak hours. Instead, autoscale waits until all users have signed out before deallocating the VM. For example, if the ramp-down phase is 15 minutes long, and the wait time before signing out users and shutting down VMs is 20 minutes long, the schedule shifts to the off-peak phase and the user sessions won't be forced to sign out.
       
       - question: |
          If I configure autoscale to force users to sign out during ramp-down, will it also sign out users with active sessions?
         answer: |
-         Yes. Idle, disconnected, and active sessions will be forced to sign out if the users don't sign out during the ramp-down phase wait time.
+         Yes. Idle, disconnected, and active sessions are forced to sign out if the users don't sign out during the ramp-down phase wait time.
 
       - question: |
          If an active session is forced to sign out, but the user tries to reconnect, is there a way to prevent them from starting a new session on a session host that autoscale is about to turn off?
@@ -83,7 +83,7 @@ sections:
       - question: |
          Why would I want to configure the load balancing algorithm differently during different phases of the scaling plan schedule?
         answer: |
-         When you set up your scaling plan schedule, you can specify different load balancing algorithms for different phases of the day. For example, during the ramp-up and peak phases, you can use the breadth-first load balancing algorithm. This algorithm will ensure you have an even distribution of user sessions during the first two phases of the day, which will optimize performance. Likewise, during the ramp-down and off-peak phases, you can use the depth-first load balancing algorithm to help the autsocale feature consolidate user sessions until it reaches the minimal possible number of session hosts in the host pool.
+         When you set up your scaling plan schedule, you can specify different load balancing algorithms for different phases of the day. For example, during the ramp-up and peak phases, you can use the breadth-first load balancing algorithm. This algorithm ensures you have an even distribution of user sessions during the first two phases of the day, which optimizes performance. Likewise, during the ramp-down and off-peak phases, you can use the depth-first load balancing algorithm to help the autsocale feature consolidate user sessions until it reaches the minimal possible number of session hosts in the host pool.
 
       - question: |
          Can you use Azure CLI to configure autoscale?
@@ -93,9 +93,14 @@ sections:
       - question: |
          Does autoscale handle scaling session hosts in secondary regions if the session hosts in the primary region have an outage?
         answer: |
-         No. Customers will need to set up their own disaster recovery strategy to manage outages. Autoscale only handles scaling existing VMs within the region they're created in.
+         No. Customers need to set up their own disaster recovery strategy to manage outages. Autoscale only handles scaling existing VMs within the region they're created in.
          
       - question: |
          Does autoscale consider availability zones during scaling operations if session hosts are created in multiple zones within a region?
         answer: |
          No. Autoscale does not know what availability zone a VM was created in and so may not perform scaling operations across all zones equally. 
+
+      - question: |
+         Can autoscale turn off all VMs in a host pool and still work properly?
+        answer: |
+         Autoscale can turn off all VMs in a host pool if the minimum percentage of hosts is set to 0% and there are no active user sessions in the host pool. However, if the last session host that’s been turned on in the host pool still has active user sessions on it, those users won’t be forced to sign out. The final VM won’t turn off until after those users sign out.