Link to home
Start Free TrialLog in
Avatar of Paul Wagner
Paul WagnerFlag for United States of America

asked on

Set up Localized vDisks in PVS Servers

I originally asked a question here on best practices for storing/presenting the VHD file in a PVS environment.

Can you give me a little direction on how to configure things in the PVS console for storing a local vdisk on each PVS server?

Where I'm at now:
- successfully run a Robocopy script and verified that the VHD files are mirrored between the two PVS servers.
- created two new Stores (one local to PVS01 and one local to PVS02).
User generated image- added the vdisk from each PVS server to the vDisk Pool.
User generated image
Now, I'm a little uncertain.
Did I do the right thing by adding both identical images from each PVS server to the vdisk pool?
In Device Collections, what vdisk would I assign to each VM? By selecting one vdisk vs. the other, it doesn't seem like there is a way for the PVS servers to balance the connections if a server goes down.


To update the vdisk with a maintenance version, do I:
- add maint. version to PVS01 vdisk in console
- update/seal image and shut down
- promote maint. version to test and verify
- put vdisk into production
- robocopy the vdisk to PVS02
ASKER CERTIFIED SOLUTION
Avatar of Carl Webster
Carl Webster
Flag of United States of America image

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of Paul Wagner

ASKER

Remove the one from PVS02. The for your target devices, you would only assign the one vDisk.

Done. I now only have the PVS01 vdisk added to the vdisk pool.
What about the PVS stores? Was I right to add a store for each vdisk path?

Thanks for the info.
I'm still unclear about the load balance portion. If I only have the PVS01 vdisk in the pool, how will the PVS02 vdisk take over if PVS01 goes down?
SOLUTION
Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Each PVS server has its own Store.
PVS01 is the "master" PVS server and the "master" Store.
Once the vDisk is ready for production, you use your Robocopy script to make sure the same vDisks and PVP files exist on PVS02 that exist on PVS01.

I'm tracking on that. No problems here.

Each vDisk is configured for Load Balancing (right-click, Load Balancing) and should be set to be served from all PVS servers.
This is what I am unclear on. If the vdisk is sourced from PVS01, then how does it know to switch to the PVS02 store if PVS01 is down? Is that just a proprietary thing I don't need to worry about?
User generated image
Yep, as long as the vDisk(s) exist in the exact same location on all PVS servers and the vDisk(s) are configured for "use the load balancing algorithm", then all the PVS servers will automatically load balance and re-home connections if a PVS server becomes unavailable.
What is the expected behavior when a server goes down? How long will it take to switch over to PVS02?

I just restarted PVS01 while connected as a test user and, the ICA session stayed connected, but there was no response from trying to open new web pages or programs. As soon as PVS01 came back up, all of the things I tried to do popped up all at once.
In all my experience with PVS, it is instantaneous and the user never knows anything happened.
Run this script and email me the output. Do not post it here as it will have private info in it.

http://carlwebster.com/downloads/download-info/pvs-assessment-script/
Ran it and sent to you.
I also noticed that Replication only shows for one server. Is it supposed to be like that?
User generated image
Hmmm, the report you sent shows all the vdisks are replicated to all servers.
SOLUTION
Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Thanks for the help, Carl! You're a good dude.