There are now a few ways of using a GPU with View but the GPU needs to be supported one way or another, Titan Black is not one of the supported ones.
// Linjo
There are now a few ways of using a GPU with View but the GPU needs to be supported one way or another, Titan Black is not one of the supported ones.
// Linjo
Not yet.
Is it that we cannot change clipboard redirection at present on connection server by opening file in notepad?
Those files only define the rules for use in Group Policy Objects. Changing the file itself won't do anything useful, you must import it to use it - see Group Policy: Inside ADM and ADMX Templates for Group Policy
Yes thats coerect we need to add template to active directory and apply policy to VDI OU suing GPO.
But I will like to know at what stage I need to edit the policy configurae redirection policy for VDI- is it first edit in connection server and then add to active directory or first add to active directory and then edit that policy.
Also how do I edit it if I suppose it has to be added first in connection server and then to be added to Active directory.
I don't understand what you're asking - you never edit the template file itself, you edit the policy that becomes available after you have imported it. The policy is applicable to sessions on the desktop VMs, applying it on the connection server itself will not make any difference.
Thank you for your response.
If I will use Location based printing have a a lot of disadvantages:
1. the parent machine must have all printer drivers in it.
2. the printer will now use thin print, so no compression for the print jobs.
3. i'm using floating pool so every time users logs on the printer will be installed from scratch!
are you sure is a good idea ?
Is there others solution to this ?
It's really only one or the other I'm afraid with the built-in printing solution - either client redirection or LBP and each have a trade-off. Ultimately the data that's being sent to the printer has to be in the form the printer understands, so it's about limiting the last step to only local network. So the best improvement you can get is to not have your print server and printers separated over a WAN link, have a print server local to each site. Then the thinprint solution would mean compressed data to the client and actual print data to the print server (close to the client) and on to the printer (close to the client). There are customers that do use LBP with floating pools that work within the requirements - pre-installing the printers on the template means that it's not really an install from scratch at login time.
Hey all,
I'm looking to accomplish the following: Have an RDSH session that I can access from within a View desktop in a View 6.0 environment, in such a way that the shortcut takes the user straight through to the app with no authentication prompts. When I create a vanilla shortcut to an RDSH app in the View client I get something that looks like this using the URI:
view://view.domain.local/cn=wordpad,ou=applications,dc=vdi,dc=vmware,dc=int
I've been doing some reading on the View Client documentation specifically around URI parameters. It doesn't appear that we can use what was the equivalent of the command line function -logInAsCurrentUser, also can't even specify password if I wanted to do user/pass for a single shortcut. Anyone know of another way to make this work such as a command line to launch an RDSH app rather than a desktop?
https://www.vmware.com/pdf/horizon-view/horizon-client-windows-document.pdf
Hi to all!
I'm in a customer designing a new Horizon View 6 architecture.
The customer has two datacenters, managed by the same vCenter.
The VDI resource pool has 10 esxi hosts, 5 from each datacenter.
So, I have the idea to separate replica and OS disks in different datastores, to read the replica disk in a flash datastore, and the virtual machines (OS diks) in another datastores with SAS.
But the problem I see is that I have only one datastore (in datacenter 1 or datacenter 2) with the replica disk. So, If I lose this datastore (the datacenter 1 goes down for example), I'll lose access to all machines in that pool, because the machines will not be able to read that file.
In this scenario, would you recommend? I've been thinking the best answer would be two datastores (one per datacenter) and NOT separate replica disk, so Composer will create one replica disk in each datastore. If I lose one of the datacenter, I'll lose only 50% of the virtual desktops, right?
Any recommendations?
Thanks in advance
Sounds to me like you need some AD/GPO help, not View help.
How to design View DR is a pretty complicated task not to be taken lightly. While I like your spirit, I'll just sum up my experience briefly.
Multiple active/active environments have the fastest failover with the least amount of IT induced confusion. While it is possible to replicate desktops whether linked clone or full, it adds a lot of undue complexity. Not knowing much about your environment, I'd recommend from a general experience separate connection brokers and golden images, abstract and replicate profile data apart from the desktops. It doubles the work for gold image updates, but then at least you have a simple way to fail over: change DNS and increase the count of desktops on the DR side. Examine the use of GSLB if required.
Here's another good resource if you decide to go the replication route with SRM.
Will this affect the run command on connect and reconnect? Is it okay to use this in production?
This won't affect CommandsToRunOnConnect, no. I have no reason to believe it would have a negative impact on desktop systems but it's not officially supported as a workaround right now - please raise a support request to get this solved for you. Did you have a sample app I can try out myself that shows this problem?
Unfortunately, this is a closed source proprietary app. I might be able to put you in touch with the developers.
Hi all,
(I am in the process of logging an SR for this, but MyVMware isn't completing that so just posting here to cover the bases)
Infrastructure: View 6.0 on vSphere 5.5.0; Windows 8.1 virtual desktops with View Agent 6.
Quick summary:
If a user disconnects their PCoIP session leaving IE open with an HTML5 based page in the foreground, CPU reaches and maintains 100% utilization. When the user re-connects, utilization drops back to normal levels.
Microsoft support performed traces on the IE DCOMP process and the findings are below.
Leaving IE minimized when a user disconnects is fine. Firefox/Chrome also do not exhibit this, only IE.
Microsoft support response:
"The trace shows IE is waiting for the batch ID confirmation.
+0x050 m_currentBatchId : 0x24bb57dc
+0x054 m_lastConfirmedBatchId : 0x20e335b0
+0x058 m_lastSurfaceProcessedBatchId : 0x20e335b0
The current batch ID keeps increasing, but the last confirmed batch id is never updated until the user Gary’s session resumed. The confirmed ID jumped from 0x20e335b0 to 0x24bb57da all of a sudden. After that, the IE UI thread jumped out the loop and resume to normal.
As a conclusion, we have some important findings:
• The CPU spike is observed after user disconnected the session, the monitoring tool will give out the warning.
• When logon using a different user, the CPU spike persists.
• When the disconnected session resumes, IE CPU drops to normal.
• However, when you disconnect the session with IE window displays, the CPU soon increased to 100%.
• We also did a test by minimize the IE window before disconnecting the session, and then the CPU stays normal and the issue doesn’t recur.
The dump we collected today shows the exact same issue as the previous dumps. DCOMP client is waiting for the confirmed ID to be updated. It seems when the VDI session is disconnected, DWM stops processing the content frame until the user session is active again."
Is there anything that can be done at the Horizon Agent side (via policy perhaps) that could mitigate this?
Thanks,
NickB
Hi,
We have a Horizon View 6. Connecting from PCoIP clients (Wyse/Samsung) throws "View Connection Server communication error" whereas there is no problem connecting from PC/iPad/iPhone with Horizon View Client or web browser.
On PCoIP client I have tried checking "do not verify server identity certificates" and also tried putting 443 or 4172 or 80 on the port for the View Connection Server but they throws same error.
Any help would be much appreciated, thank you.
What firmware are you running on the clients? If its really old this could be a possible explanation.
Latest is 4.5.
// Linjo
Thanks for the response elgwhoppo!
I like your solution, It requires more storage but I think this may not be a problem for the customer
I'll take a look to SRM solution, but It seems that requires more than two vCenter Servers. The customer consolidated the vCenters he had into one last month.So the SRM solution may not like it
I'm going to propose this solution and I'll let you know the feedback
Thanks!!
They are on latest 4.5.1.
This isn't a View fix, I don't think there's anything we can do about this Microsoft bug directly, but have you tried turning off hardware rendering in IE as a workaround? Internet Options->Advanced->Accelerated graphics->Use software rendering.