World of EUC

Hi <@ULKDZHKFY>, why is the “Scaled-On” tag created by the script and what it does? I didn’t found anything in the documentation. 
Thanks

Hi  scaled on means that the machine has been scaled on during the day by the script. These tags should be removed when machines are shutdown.

Hey Ashley, if you've not told it to force a logoff it should just wait for users to be off. If you are messaging users it will force them off after a given interval. The logs should have it all in there though.

Hi Salmen, you need to add the value right now. It could be overridden quite easily if necessary.

Thanks Lee, got this setup and gonna see if it works after setting up similar to what you posted.

<@ULKDZHKFY> would the script still put a log into the PowerScale Logs folder if it's ran as a Scheduled Task.

Yes indeed it would. Sounds like you're having issues with the scheduled task. Is it identical to mine?

condistions and settings page are both identical.

Looks okay to me, run this in a powershell administrative  console

Get-Childitem "Path to Powershell Folder" -Recurse | Unblock-File    -verbose

It may have been marked by windows using streams and therefore not be able to run. Try your scheduled task after this.

If this does not do the trick. Open a powershell prompt and from there call the decision making ps1 script to see what happens.

sorry, i'll run the mentioned script.  I thought you wanted to see the script run from a console

I ran the unblock and also put a screen shot of manually calling the script from PS Console.

-ExecutionPolicy Unrestricted -File "C:\PowerScale\Decision Making.ps1"

C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe

I'm rebuilding the Task. Here's a copy and paste of what I put in the Program/Script and Arguments paths.

Well done for sticking with it. Can you confirm you got it working?

No problems at all. I will do it tomorrow.

Hi Simon, it can be done relatively simply. You just need a script based action to do this?

Hi Leee, yes exactly i would like to simplify the work of our support, at the moment they have to connect to the DDC put a TS in maintainance and add the exclude tag and often they forget to remove the exclude tag after they finished doing maintainance. would be really helpfull to do with the controlup console.

Okay, All the command are actually already in the script. In order for them to work though, you have to have the Citrix PowerShell modules installed where the script runs. Let me tag a friend in who has a lot of experience with ControlUp Scripted actions. <@ULJM9DQF2> Can you lend a hand here. ControlUp scripted actions on a machine and performing Citrix SDK commands.

I'm working on it - I don't usually develop from forum requests but this is "interesting" and "useful" (and easy)

image.png

Manipulate Citrix tag.ps1

Manipulate Citrix Virtual Apps and Desktops Tags.xml

Here you go - use at your own risk. Import the xml into the CU console. It needs to be set to run on a delivery controller (or machine with CVAD PowerShell cmdlets where "Delivery Controller" parameter needs to be set to a delivery controller).

The operations can be one of : set, unset, list or delete.
The script is used entirely at your own risk although I have tested it in my lab with 2106.

Thank you very much Guy, i will do some testing and give you a feedback

Thank you Guy, I wasn't expecting you to just do it :wink:

I wouldn't normally but I can see multiple use cases for tag manipulation as I don't think it's great in the Citrix console, primarily as it was functionality tacked on later in the life of 7.x

Much appreciated Guy. I had the PowerShell, you've saved me the effort of integrating it into a scripted action. Essentially, to isolate a machine it needs to be placed into maintenance and a tag needs to be added for PowerScale to now ignore that machine from calculations. This was a request from Simon to be able to manage this process from within ControlUp.

Hey friends. Anyone tried setting up powerscale recently? I am running into odd problems. When running Decision Making.ps1 At C:\PowerScale\Decision Making.ps1:1171 char:3
+ &lt;&lt;&lt;&lt;&lt;&lt;&lt; HEAD
+   ~
Missing file specification after redirection operator.
At C:\PowerScale\Decision Making.ps1:1171 char:2
+ &lt;&lt;&lt;&lt;&lt;&lt;&lt; HEAD

reverting to previous version on github fixes the issue.

is anyone using PowerScale in production? Would be interested to hear how others are setting up.

I can indeed, I have people using it in production. There are a few bugs with the newer functionality that I'm still working out but it's solid on power managing machines.

Might have some ghosts in that code release <@ULKDZHKFY> happens to me with vscode sometimes 

Its such a beast of a script now its really time consuming to test. I've got it in my to do list. Too busy in the run up to Christmas though.

I might fix and sub a pull over the weekend if I have time - I had this happen to me when I submitted a change for bisf once, poor Matthias had to fix my crud - check line 1171 for an example, there be a few in there 

Cheers James, much appreciated chap. Obviously, log that time, you know where.

I've got some new features and bug fixes on my machine, I just haven't published it yet.

<@U0125S0736V> you got these two super smart guys on it now.

<@ULKDZHKFY> <@UKJTJD4KA> Thanks guys, super generous of you to take time to work on this.  I am looking at the $Metrics variable to see if its possible to point it at a delivery group instead of the whole farm. Havent gone through entire code yet. so if you have an idea on specific lines it would be really cool.

Hey Dima, it only captures performance from the machines in the delivery group. It doesn't do the entire farm, just what you setup to monitor.

If you look at the code Get-BrokerMachine you will see it's filtered based on what you configure.

So this part is filtered somewhere in decision making script? I havent dug too deeply in, but looking on the surface I cant find the filter part in the script.

Okay, I see. You've seen the word farm in there and thats made you think its the entrie farm. I understand.

This is the code where I grab all the machines:

There is then a further line that filters down performance monitoring targets to machine that are not in maintenance mode:

It sure does, it works with Studio so it'll work in any environment.

So, if understand, it monitors the ressources, and start / stop the ec2/ aws? 
If there are prerequisits for aws, could you give me infos?

There are no prereqs for AWS. All the actions to shutdown machines etc... are managed through the Citrix Broker Infrastructure so you can just install it.

But, in order to use your script, i need mcs managing the deployment / ec2 according to the docs

Your machines need to be able to be power managed by studio. That's the only requirement really.

currently, ddc on prem + vda on aws (network linked of course)

Okay and what is the provisioning method for the VDAs? Are they power managed via Studio?

yeah that is the question, and it seems mcs for aws provisionning is the way

As long as you can power down/up machine from studio powerscale will work.

There is no option to build more no, that's relatively complex to handle so I decided not to add that as a feature.

With regards to your buffer, thats your out of hours machines. You can use performance scaling to fire-up more machines as they come along.

Will the out of hours machine be up all the time though?

what I was thinking is using performance metrics to create a buffer based on number of users. However this is VDI scenario and I cant specify less then one per machine if that makes sense.

Yes, the OOH machines will be up all the time. In that scenario, use load index. Set your policies in Studio so your load index is triggered by session count. Do some testing and see how that works out.

I did not think about that. Will give it a shot. I do understand that the tool is geared mostly to RDSH scenario. For the most part normal power policies work for VDIs. I just need more granular option that weekday weekend and your script is awesome at that.

It was aimed at both but the primary development piece was indeed for RDSH. I'm glad you've got it running. Nice one Dima.