MCS Persistent XenApp

[Christian Pedersen1709152341]

Hi guys,

 

In XenApp < 7.9 when you created the images from MCS the VM Was persistent, it was only on image update the VM was "reset" to the Image (I think - it was like that)

 

In 7.9 it seems like the VM is reset to zero at each reboot

 

Can i get the VM persistent on MCS - so the eventlog etc. is saved and its there after we reboot the server.

 

So its only if i do a "image" update of the XenApp server the server is "reset" to the last image update.

[Christoph Sinabell]

Hi,

 

I too can confirm, that regardless of the cache type or how you reboot the VM (Windows, Hypervisor or Studio) the cache is always wiped.

 

Regards

[CarlStalhood]

Yes.

[CarlStalhood]

Machines are either persistent, or non-persistent. There is no in between. If you created a non-persistent catalog (discard), then it will discard on reboot.

 

Persistent machines do not have an image update option.

[Ryan Goldstein]

It has never been that way.  You can either have a non-persistent catalog or a persistent one.

 

If you have a persistent desktops the only way to refresh the desktops is to delete them and create new ones.  This can automated.

 

If you have non-persistent they refresh at every reboot, there is a distinction between DDC reboot and vmware reboot.

[Christian Pedersen1709152341]

Okay wierd - i was pretty sure that xenapp servers was statefull until imageupdate was done

 

So you could do small updates to all xenapp servers on the fly and only do full updates on a pr month base with new updated images..

 

I was pretty sure that was possible in an older release where the linked disk from base vm was only reset at image update process (Only XenApp)

 

So we did not have to go through the image update process to implement small changes .. But guess im wrong..

[IT Support1709154654]

//wrong account.

see blow:

[Marco Zimmermann]

Hi,

 

before 7.9 you were able to reboot a MCS Server manually (via Start => Shutdown => Reboot) without wiping cache.

Rebooting via Studio wiped the cache (It wasn´t a reboot, it was a power off, power on).

Using Cache in Memory in 7.9 makes this impossible ;)

I´m not sure about 7.9 using hard disk cache only but that should be the same because of MCSIO.

 

See MCSIO:

https://www.citrix.com/blogs/2016/08/03/introducing-mcs-storage-optimisation/

 

"Changes in the architecture introduce a subtle, but important change to how MCSIO operates compared to standard MCS. In MCS, the delta disk cached all guest read/write operations and resides on the same storage as the master disk. In MCSIO, your temporary cache handles the I/O operations once handled by the delta disk. Now, you can separate persistent and temporary data."

 

 

 

Marco

[CarlStalhood]

If you have memory caching enabled, memory is wiped on every reboot. Disable memory caching and you'll revert to the older functionality.

[Marco Zimmermann]

That´s not correct Carl, even if you use hard disk cache only the cache is wiped at every reboot.

This happens when machine catalog is created with studio 7.9 or newer.

Old (migrated) machine catalog will act like pre 7.9

 

Marco

[CarlStalhood]

Depends on how you reboot. If you reboot inside the VM, it's not reverted if you don't use memory cache. If you reboot using Studio, it always reverts, no matter the cache type.

[Marco Zimmermann]

Just rebuild this in my lab with HDD cache only and it was also wiped when not using studio to reboot. VM was definitely not powered off during reboot.

[Bob Bair]

I'm running into this same thing. I often use one specific server to test updates. Now with 7.18, when I reboot the server it returns it back to it's original state w/o the installation. Is there a way around this?

 

Bob

[Bob Bair]

On 10/7/2016 at 1:17 PM, Carl Stalhood1709151912 said:

Depends on how you reboot. If you reboot inside the VM, it's not reverted if you don't use memory cache. If you reboot using Studio, it always reverts, no matter the cache type.

 

Carl,

I agree that's how it WAS. Not anymore.

[CarlStalhood]

Is Memory Caching enabled on your Catalog?

[Bob Bair]

12 minutes ago, Carl Stalhood1709151912 said:

Is Memory Caching enabled on your Catalog?

 

Carl,

If you are talking about the Configure a cache for temporary data on each machine, the default is checked. You are saying that by creating a new Machine Catalog with that unchecked it will stay persistent?

 

Bob

[Rahul Sukumar]

I am on 2009 delivering multi-session Server 2019 virtual apps. The settings I used to create the catalog are attached. I have a nightly reboot scheduled for the catalog. Each night, the VHDs are essentially deleted and recreated using the master image. I don't want this to happen for reasons others have mentioned here.

 

I also noticed that under the machine type column when viewing the list of machine catalogs, it says "User Data: Discard." I don't know if that has something to do with this but I am unable to change this setting, either during MC creation or afterwards. Any help is greatly appreciated!

 

 

 

[CarlStalhood]

MCS RDSH catalogs are always non-persistent. 

 

One option is to enable memory caching, specify a size for the cache disk, and then move your Event Logs and other persistent data to the cache disk.

 

Another option is to use your hypervisor to clone a template RDSH machine and then add the clones to a "Manual" Catalog in Studio so it doesn't use MCS.

[Rahul Sukumar]

Thank you Carl. How about using PVS instead? Will RDSH catalogs also be non-persistent?

[CarlStalhood]

Yes, PVS is also non-persistent for the C: drive. It also has the persistent cache disk option.

[Rahul Sukumar]

Thanks. Now I'm having a problem with machines created via MCS with VMM as the hypervisor are not able to live migrate to another host. Not sure what the issue is but staging a new 2009 + SCVMM 2019 has been a struggle for me since the time they announced it was supported!