The Microsoft Application Virtualization Blog : Troubleshoot

App-V “Application is not working as expected” Troubleshooting Check List

J.C. Hornbeck — Wed, 25 Apr 2012 19:12:35 GMT

Hello everyone, John Behneman here again. The purpose of the article is to highlight some very simple OSD file edits/tweaks that can potentially fix a large number of most common issues we see here in product support. Chances are you can save yourself a lot of time and money if you try these simple troubleshooting techniques before calling us for support.

As a test you should always edit the OSD file in the App-V client cache on the App-V client, not on the original package until you know which parameter is causing the problem. It is always a good policy to backup the OSD cache file before you start editing so you can return it to its original configuration should none of the edits resolve your issue. If you corrupt the cached OSD file and you did not make a backup, you can always redeploy the package or do a DC refresh if you are using an App-V Management Server.

NOTE Before doing anything else, be sure you have the latest version of the App-V client. That alone can fix a lot of issues.

How to launch App-V Application OSD in Notepad from the Client Cache

To find the cached OSD for an App-V application, launch the Application Virtualization Client Admin console, highlight the Application node of in the left hand pane, find the desired application in the right hand pane and double-click on it. Find the Local OSD File: parameter on the General Tab and copy the parameter. Type Notepad<space> in the Start -> Run and then paste the Local OSD File: parameter after the space in Notepad so you have something similar to following string:

notepad C:\ProgramData\Microsoft\Application Virtualization Client\SoftGrid Client\OSD Cache\3277aa21-2d5c-4bc4-a396-8f20c7953fec.osd

Now, click OK and Notepad will open the cached OSD of the application you wish to modify so you can add tags or modify existing tag in the appropriate section of the OSD. The following is a graphical representation of the process:

Figure 1 - Copy the Local OSD File parameter.

Figure 2 - Type notepad plus a space and then paste the Local OSD File parameter.

Figure 3 -The Cached OSD is presented for editing.

Editing tags 1-5 in the OSD

Figure 4 - Snippet of an App-V OSD file

Test items 1,3,4 in Figure 4 should be tested on every Application that exhibits a problem when virtualize with App-V, items 2 & 5 should be done only in specific situations to be explained in more detail below.

Test item 1:

Change

<LOCAL_INTERACTION_ALLOWED>FALSE</LOCAL_INTERACTION_ALLOWED>

<LOCAL_INTERACTION_ALLOWED>TRUE</LOCAL_INTERACTION_ALLOWED>

Test

For more information:

A Look Under the Covers - The LOCAL_INTERACTION_ALLOWED Tag : http://blogs.technet.com/b/appv/archive/2007/09/20/a-look-under-the-covers-the-local-interaction-allowed-tag.aspx

Test item 2:

If an App-V application fails to shutdown properly or causes anomalies with the OS when the application is shutdown, try setting the following value to TRUE.

Change

Test

For more information:

An application may not shut down completely when you exit the application in Microsoft SoftGrid : http://support.microsoft.com/kb/931589

Test item 3:

Change

Test

For more information:

When launching an application virtualized using Microsoft App-V, the application hangs for 10 minutes and then terminates : http://support.microsoft.com/kb/2195120

A known good App-V virtual application package may fail with error 19-00001003 after adding a pre-launch script : http://blogs.technet.com/b/appv/archive/2011/10/20/a-known-good-app-v-virtual-application-package-may-fail-with-error-19-00001003-after-adding-a-pre-launch-script.aspx

Test item 4:

On rare occasions the working directory parameter can cause problems, as a test try removing the value altogether.

Change

<WORKINGDIR>%CSIDL_PROGRAM_FILES%\Microsoft Host Integration Server 2010\system\<WORKINGDIR>

For more information:

<WORKINGDIR> Examples - Specifying a Working Directory : http://blogs.technet.com/b/virtualworld/archive/2007/11/07/workingdir-examples-specifying-a-working-directory.aspx

Microsoft Application Virtualization 4.6 SP1 Sequencing Guide : http://download.microsoft.com/download/F/7/8/F784A197-73BE-48FF-83DA-4102C05A6D44/App-V/App-V%204.6%20Service%20Pack%201%20Sequencing%20Guide.docx

Test item 5:

Typically the problem these tags cause is that short-cuts for a particular application may not be available to users that have been granted access to the application. You may see the following entry in the sftlog.txt on the App-V client:

The app manager could not create an application from '\\softgriddc\content\DefaultApp.osd' (rc 07708A44-00000007)
Removing these entries altogether will deploy this application to any Operating System, even to Operating Systems for which the application may not be designed. In this situation the cached OSD will not exist on the client and you must edit the OSD from the content share on the App-V Management server or in the original package before deployment.

Change

</VM>
<OS VALUE="Win7"/>
<OS VALUE="Win764"/>
</IMPLEMENTATION>

</VM>
</IMPLEMENTATION>

For more information:

Error xxxxxx44-00001007 in the SFTLOG.TXT file : http://blogs.technet.com/b/appv/archive/2008/10/29/error-xxxxxx44-00001007-in-the-sftlog-txt-file.aspx

Updated OS VALUE XML Tag Reference and Supported Client Versions : http://blogs.technet.com/b/appv/archive/2009/10/29/updated-os-value-xml-tag-reference-and-supported-client-versions.aspx

If your package is still not working properly after the aforementioned tweaks, grab a cup of coffee and startup Process Monitor in the bubble.

NOTE For more information on how to use Process Monitor please see How to run Process Monitor (ProcMon) inside the App-V virtual environment - http://blogs.technet.com/b/appv/archive/2012/04/24/how-to-run-process-monitor-procmon-inside-the-app-v-virtual-environment.aspx

Please be sure to return all the OSD tag values back to the original settings, I would advise doing this after each test item so you know exactly what fixed the issue. It is possible that you may need to tweak more than one value to get some applications working. If you are editing the OSD in the App-V Client Cache, and you deploy your App-V Applications using an App-V Management Server you can always do a DC refresh to return the OSD to its original configuration.

DC Refresh:

For suffering through all the steps above here is an additional couple of nuggets of information.

Bonus Test 1:

There are 2 types of packages a MNT, which is the preferred method, where the application is installed to directory on the Q: drive during sequencing, and VFS where the application is installed to the C: drive.

Some applications have to be installed to the C: drive to work correctly when virtualized. If the application fails to launch correctly when you sequence it to Q: (MNT), please re-sequence the application by installing it to the C: drive. In some situations the inverse is true, I have had situations where sequencing an application to the Q: drive fixed a problem. The bottom line is you may need to experiment to determine which install type (MNT or VFS) works best for the application. For more information please see the following references:

For more information:

The VFS versus MNT Install : http://blogs.technet.com/b/appv/archive/2007/07/16/the-vfs-versus-mnt-install.aspx

Best practices to use for sequencing in Microsoft App-V (SoftGrid) : http://support.microsoft.com/kb/932137/en-us

Bonus Test 2:

By default “Enforce Security Descriptors” is enabled on the Sequencer, sometimes this feature can cause problems, try disabling it by unchecking it. To do this start the sequencer, choose “Modify an Existing Virtual Application Package”, then “Edit Package” and browse to the *.sprj file in the package and select it and click Open then Edit. This will load the following screen, select the Deployment tab and uncheck “Enforce Security Descriptors”, now go to File and Save and it will update the package. You will need to redeploy the package for testing.

Figure 6 – Sequencer Package edit console

For more information:

Improving Security During App-V Sequencing : http://technet.microsoft.com/en-us/library/ee662362.aspx

Bonus Test 3:

If you suspect that System-Guard (protective App-V bubble) is causing your issue you can use the following registry tweak to determine for certain if it is the cause, please only do this as a test, leaving this in production is not recommended. To gain a complete understanding of this tweak please read the following blog post before implementing it.

Allowing Better Interaction to App-V Virtual Applications : http://blogs.technet.com/b/deploymentguys/archive/2011/07/05/allowing-better-interaction-to-app-v-virtual-applications.aspx

Find the following registry hive and add a string value of N+1 where N is the highest number listed in the ObjExclusions registry hive, in most cases the next value will be 94 (on my machine the highest number is 93) and give it a data value of *. After making this change either reboot the test machine or restart the Application Virtualization Client service.

For X86:

HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\SoftGrid\4.5\SystemGuard\ObjExclusions

For X64:

HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\SoftGrid\4.5\SystemGuard\ObjExclusions

Example:

Figures 7 & 8 - Registry

For more information:

Error message when you try to sequence Abra Suite in Microsoft SoftGrid: "The procedure entry point GetFileVersionInfoA could not be located in the dynamic link library VERSION.DLL" : http://support.microsoft.com/kb/931979

Error: HP Universal Printing | Invalid Address when using the Hewlett-Packard Universal Printer Drivers with Microsoft Application Virtualization : http://blogs.technet.com/b/appv/archive/2011/10/13/error-hp-universal-printing-invalid-address-when-using-the-hewlett-packard-universal-printer-drivers-with-microsoft-application-virtualization.aspx

I want to thank Alvin Morales, Mark Stanfill and J.C. Hornbeck for their assistance and guidance during the process of creating this post.

John Behneman | System Center Support Engineer

Get the latest System Center news on Facebook and Twitter:

The Forefront Server Protection blog: http://blogs.technet.com/b/fss/
The Forefront Endpoint Security blog : http://blogs.technet.com/b/clientsecurity/
The Forefront Identity Manager blog : http://blogs.msdn.com/b/ms-identity-support/
The Forefront TMG blog: http://blogs.technet.com/b/isablog/
The Forefront UAG blog: http://blogs.technet.com/b/edgeaccessblog/

How to run Process Monitor (ProcMon) inside the App-V virtual environment

J.C. Hornbeck — Tue, 24 Apr 2012 17:33:57 GMT

Hi everyone, Mark Stanfill here. Running Process Monitor (ProcMon) to troubleshoot Microsoft Application Virtualization (App-V) client issues such as missing files, access denied errors or other file and registry-based issues is sometimes necessary to successfully package an application. This post details a simple method to allow you access inside the App-V virtual environment (bubble) for troubleshooting purposes. The procedure consists of 4 main steps:

1. Create a shortcut to an application

2. Edit the shortcut to launch a command prompt inside the App-V bubble

3. Launch ProcMon or other troubleshooting tools

4. Optionally fix the issue

Create a shortcut to an application

Start by identifying an executable inside the package you are investigating. To do this, launch the Application Virtualization client through Start\Run\Sftcmc.msc. In the Application Virtualization client, select Applications and identify your problem application. Right-click on the application on the right-hand pane and choose New Shortcut…. Follow the wizard's prompts to create a shortcut. On the New Shortcut - Step 1 screen, change the name to something easy to identify such as Command Prompt in ApplicationX Bubble. On the New Shortcut - Step 2 screen, click The Desktop for the location to place the shortcut. Leave command line parameters blank for now.

Edit the shortcut to launch a command prompt inside the App-V bubble

Right-click on the newly created shortcut on your desktop and choose Properties. Edit the Target: text to include the string /exe cmd.exe between Sfttray.exe and /launch. For example, this command line:

C:\Program Files (x86)\Microsoft Application Virtualization Client\sfttray.exe" /launch "DefaultApp MFC Application 1.0.0.1

Becomes:

C:\Program Files (x86)\Microsoft Application Virtualization Client\sfttray.exe" /exe cmd.exe /launch "DefaultApp MFC Application 1.0.0.1

Click Apply and OK to save the shortcut.

Launch ProcMon or other troubleshooting tools

Double-click on the shortcut to launch the command prompt in the App-V bubble. From the command prompt, type the full path to Procmon.exe to launch it (for example c:\temp\procmon.exe). Procmon will have full access to changes occurring inside the bubble at this point.

Optionally fix the issue

At this point, troubleshooting will depend on the type of application failure you are experiencing. It is possible to fix individual issues such as missing registry keys through Regedit.exe, file permissions (via cacls.exe), etc. These changes only affect the local installation of the package. To fix the issue permanently, you will need to open the package for upgrade in the App-V Sequencer, resolve the issue, then save and publish the modified application.

FAQ

Q: Why do I need a command prompt? Why can't I use Windows Explorer?

A: It is not possible to load the Explorer shell directly from App-V. Access to files is limited to command-line tools. Other GUI tools may, however be launched from the command prompt. One trick to see a full Explorer view is to launch a tool that exposes a file…open dialog such as Notepad.exe and then use that to navigate. You can launch executables through the common file dialog via the right-click context menus.

Q: Why load a command prompt and not just specify /exe ProcMon.exe directly in the command line?

A: This technique will work, but it relies on either specifying the entire path to ProcMon.exe or else putting it in a folder in the user's path. We find that most troubleshooting involves multiple tools and troubleshooting steps inside the bubble, so keeping a command prompt open rather than running multiple commands is more efficient.

Q: I have System Center Configuration Manager (ConfigMgr) in my environment. My Target: text specifies VAppLauncher.exe and adding /exe cmd.exe doesn't seem to work.

A: Replace VAppLauncher.exe with the complete path to sfttray.exe (normally C:\Program Files (x86)\Microsoft Application Virtualization Client\sfttray.exe). VAppLauncher does not support the /exe switch.

Q: How can I verify that a command prompt is inside or outside the bubble?

A: The easiest way is to type set SFT_MNT in the command prompt. This variable will not be set outside the bubble. Inside the bubble, you should see a response like SFT_MNT=Q:.

Q: Do I need to specify /externalcapture in the ProcMon command line?

A: /Externalcapture allows Procmon to see in to the App-V bubble without first launching a command prompt in the bubble. If all you need to do is to gather diagnostic information (i.e. you are not going to attempt to fix the issue and don’t need to run any other commands against the package), ProcMon.exe /externalcapture can be ran directly from Explorer to capture data from applications running under App-V.

Mark Stanfill | Senior Support Escalation Engineer

Get the latest System Center news on Facebook and Twitter:

Troubleshooting 04-0000180C / 6156 Errors in Microsoft Application Virtualization

J.C. Hornbeck — Tue, 27 Mar 2012 18:44:17 GMT

Here’s a new Knowledge Base article we published today. This one talks about troubleshooting an issue where attempting to upgrade a package in Microsoft Application Virtualization (App-V) fails with error 04-0000180C:

=====

Symptoms

Attempting to upgrade a package in Microsoft Application Virtualization (App-V) fails with error 04-0000180C:

C:\Program Files (x86)\Microsoft Application Virtualization Client>sftmime.com add package:" APP_V1.1" /manifest "C:\\Program Files (x86)\\SMS_CCM\\Cache\\ABC00076.2.S-1-5-21-1111111111-1111111111-1111111111-500\\APP_V1.1_manifest.xml" /overrideUrl "FILE://C:\\Program Files (x86)\\SMS_CCM\\Cache\\ABC00076.2.S-1-5-21-1111111111-1111111111-1111111111-500\\ABC00076.sft"

The Application Virtualization Client could not complete

the operation.

The operation failed because the GUID in the package

manifest matches a package name other than the one

specified. Verify the package name, and then try again.

Error code: 4605F3-18305104-0000180C

For System Center Configuration Manager (CfgMgr) clients, the VirtualApp.log will show entries similar to the following, exiting with error code 6156 (SWRC_PKG_NAMEDOESNOTMATCH):

<![LOG[Executing command line: "C:\Program Files\Microsoft SoftGrid\Microsoft SoftGrid Application Virtualization for Desktops\SFTMIME.COM" ADD PACKAGE:"Package1" /MANIFEST "C:\\WINDOWS\\system32\\CCM\\Cache\\ABC00500.3.S-1-5-21-1111111111-1111111111-1111111111-2672\\Package1.xml" /OverrideUrl "FILE://C:\\WINDOWS\\system32\\CCM\\Cache\\ABC00500.3.S-1-5-21-1111111111-1111111111-1111111111-2672\\ABC00500.sft"]LOG]!><time="16:37:46.837+-120" date="08-08-2008" component="VAppRegistration" context="" type="1" thread="3120" file="registrationhandler.cpp:1031">
<![LOG[The command completed with exit code 6156.]LOG]!><time="16:37:47.368+-120" date="08-08-2008" component="VAppRegistration" context="" type="1" thread="3120" file="registrationhandler.cpp:1140">

Cause

This error indicates that there is a mismatch between the package name of the original package and that of the upgraded package. This results in the App-V client incorrectly determining that there is a duplicate GUID, since the package GUID in both manifests will be the same. To successfully upgrade a package, both of the following conditions must be true:

1. The package GUID must be the same (the Version tag and VERSIONGUID tag will change).

2. The package name must be exactly the same. This can be verified by opening the original and upgrade package's manifest XML in a text editor to compare the NAME and GUID tags.

The NAME tag is derived from the Package Name field on the General Properties tab of the package. The App-V Sequencer allows this field to be changed after sequencing a package but before saving the files. To avoid this error, do not edit this field.

In a second variation, the Microsoft System Center Configuration Manager (ConfigMgr) Virtual Application Package Update Wizard allows for the Name: field to be changed on its General page. Editing the Name: or Version: fields of the New Package's values will result in the errors above.

Resolution

To avoid the 04-0000180C and 6156 errors listed above, do not change the package name in either the App-V Sequencer or the ConfigMgr Virtual Application Package Update Wizard. If the package manifest file is correct, simply re-run the Virtual Application Package Update Wizard without changing any values. If the package name has been changed in the Sequencer, re-sequence the application and do not change this value.

=====

For the most current version of this article please see the following:

2688524 : Troubleshooting 04-0000180C / 6156 Errors in Microsoft Application Virtualization

J.C. Hornbeck | System Center & Security Knowledge Engineer

Get the latest System Center news on Facebook and Twitter:

How to troubleshoot applications failing to stream from an App-V management server

J.C. Hornbeck — Wed, 07 Sep 2011 19:13:00 GMT

Here’s a new App-V KB article we published today. If you ever need to troubleshoot issues where a client fails to stream a virtualized app from the App-V server then there’s no better place to start than right here:

=====

Summary

This article covers how to troubleshoot virtualized applications in Microsoft Application Virtualization (App-V) that fail to stream from the management server.

When an application fails to stream on an App-V client, the application will fail to launch with the following error:

The Application Virtualization Client could not launch application name.

The error message will also include an error description and code like the examples below:

No connection could be made because the target machine actively refused it.
Error code: xxxxxxx-xxxxxx2A-0000274D
The package requested could not be found in the system data store or the files associated with this package could not be found on the server. Report the following error code to your System Administrator.
Error code: xxxxxxx-xxxxxx0A-20000194
No such host is known.
Error Code: xxxxxxx-xxxxxx2A-00002AF9
The specified Application Virtualization Server could not be accessed.
Try again in a few minutes. If the problem persists, report the following error code to your System Administrator.
Error Code: xxxxxxx-xxxxxx0A-10000002

In the Sftlog.txt file, the following error will be logged:

[08/24/2011 15:32:56:618 JGSW ERR] {hap=5:app=Appname:tid=16C:usr=Administrator}
The Application Virtualization Client could not connect to stream URL 'rtsp://appv-svr:554/Application/Application.sft' (rc 19D07F2A-0000274D, original rc 19D07F2A-0000274D).

Note: The error code in the Sftlog.txt will vary.

More Information

The first step in troubleshooting an application failing to stream is to determine if the issue is isolated to a single application or all applications.
Once the scope of the applications affected is determined, perform the steps below that are appropriate for your scenario.

How to troubleshoot a single application that fails to stream

Step 1: Review the Sftlog.txt file on the App-V client

On the App-V client, review the Sftlog.txt file on the App-V client. This log file may include additional information that wasn’t included in the error message.

The default location for the Sftlog.txt is: %systemdrive%\ProgramData\Microsoft\Application Virtualization Client

Step 2: Review the application .osd file on the App-V Management Server

1. On the App-V Management Server, open the application .osd file and scroll down to the following line:

<CODEBASE HREF="rtsp://servername:554/ApplicationDirectory/Application.sft"

2. Verify the protocol, sever name, port and path to the SFT file are correct.

Default ports for each protocol type:

RTSP=554
RTSPS=322
HTTP=80
HTTPS=443

3. If changes were made to the application .osd file, save the changes and then open the Application Virtualization Client MMC snap-in on the App-V client and refresh the Publishing Server.

4. Launch the application on the App-V client to see if the error continues to occur.

Note: If the application OSD file is using the %SFT_SOFTGRIDSERVER% environment variable for the server name, verify the environment variable is configured on the App-V client by performing the steps documented in the following Microsoft knowledge base article:

http://support.microsoft.com/kb/2271342

Step 3: Delete the application from the cache on the App-V client

1. On the App-V client, open the Application Virtualization Client snap-in that’s located under Administrative Tools.
2. Click Applications.
3. Right-click the application that is failing to stream and click Delete.
4. Click Yes when you receive the confirmation dialog box.
5. Once the application is deleted, refresh the publishing server to republish the application.

To refresh the publishing server, perform one of the following methods:

Method 1

1. Open the Application Virtualization Client snap-in.

2. Click Publishing Servers.

3. Right-click the publishing server and choose Refresh Server.

Method 2

1. Right-click the App-V icon in the notification area and choose Refresh Applications.

2 Launch the application on the App-V client to see if the error continues to occur.

How to troubleshoot all applications failing to stream

Step 1: Review the Sftlog.txt file on the App-V client

On the App-V client, review the Sftlog.txt file on the App-V client. This log file may include additional information that wasn’t included in the error message.

The default location for the Sftlog.txt is %systemdrive%\ProgramData\Microsoft\Application Virtualization Client.

Step 2: Verify the App-V client can access the content directory

On the App-V client, click Start, in the Search or Run line, type the UNC path of the content share (Example: \\appv-svr\content) and then press ENTER.

If the client fails to connect to the content share, verify the UNC path is correct and verify the NTFS and Share permissions on the content directory are correct by performing the steps below.

On the server hosting the content directory, verify the following NTFS and Share permissions are configured on the content directory:

· App-V Users = Read
· App-V Administrators = Read and Write
· Network Service = Read and Write

The default location for the content directory is: %systemdrive%\Program Files (x86)\Microsoft System Center App Virt Management Server\App Virt Management Server\content.

Step 3: Verify the path to the content directory on the App-V Management Server

To verify this, perform the following steps on the App-V Management Server:

1. In Administrative Tools, open the Application Virtualization Management Console.
2. Right-click the server name and then click System Options.
3. Verify the Default Content Path is pointing to the content directory location.

Note: The content location should be referenced by UNC path (Example: \\appv-svr\content).

4. Click OK to close the System Options window.
5. Close the Application Virtualization Management Console.
6. Open Regedit.
7. Navigate to the following key:

32-bit systems: HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\SoftGrid\4.5\Server

64-bit systems: HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\SoftGrid\4.5\Server

8. Verify the SOFTGRID_CONTENT_DIR registry value is pointing to the content directory location.

Note: The content location should be referenced by UNC path if the content share is a DFS share (Example: \\appv-svr\content).

9. If the SOFTGRID_CONTENT_DIR registry value was modified, restart the Application Virtualization Management Server service or restart the server.

Step 4: Verify the Application Virtualization Management Server service is started on the App-V Management Server

To verify this, perform the following steps on the App-V Management Server:

1. In Administrative Tools, open the Services MMC snap-in.
2. Locate the Application Virtualization Management Server service.
3. Verify that the service is Started.
4. If the service is not started, right-click Application Virtualization Management Server, and then click Start.
5. If the service fails to start, search the Microsoft Knowledge Base for the error message that is reported.

Step 5: Verify the App-V client can telnet to the App-V Management Server and port.

To verify this, perform the following steps on the App-V client:

1. At a command prompt, type telnet ServerName Port, and then press ENTER.

For example, type the following command and then press ENTER:

telnet appv-svr 554

2. If the connection succeeds, the window is blank. Press ENTER two times and you will receive the following message:

RTSP/1.0 400 Bad Request
Server: Microsoft Application Virtualization Server/x.x.x.xxxxx [Win32; Windows NT x.x ]
Date: xxx, xx xxx xxxx xx:xx:xx xxx

If the connection is unsuccessful, you will receive the following message:

Could not open connection to the host, on port 554: Connect failed

If the Application Virtualization Management Server service is started but the client cannot telnet to the server, verify that port traffic between the client and the server is not restricted by a firewall or by other software. For more information, contact the network administrator.

Step 6: Review the application .osd files on the App-V Management Server

1. On the App-V Management Server, open the application .osd file and scroll down to the following line:

<CODEBASE HREF="rtsp://servername:554/ApplicationDirectory/Application.sft"

2. Verify that the protocol, sever name, port and path to the SFT file are correct.

Default ports for each protocol type:

RTSP= 554
RTSPS=322
HTTP=80
HTTPS=443

3. If changes were made to the application .osd file, save the changes and then open the Application Virtualization Client MMC snap-in on the App-V client and refresh the Publishing Server.

4. Repeat steps 1-3 for all applications that fail to stream.

Note: If the application OSD file is using the %SFT_SOFTGRIDSERVER% environment variable for the server name, verify the environment variable is configured on the App-V Client by performing the steps documented in the following Microsoft knowledge base article:

http://support.microsoft.com/kb/2271342

Step 7: Clear the cache on the App-V client

If steps 2-6 have confirmed that the App-V client can communicate with the App-V Management Server and the settings are configured properly, it’s possible that the application is failing to stream due to a corrupted cache file on the App-V client.

Note: Clearing the cache on the App-V client will delete all application data from the cache file. This may cause application load times to increase the first time an application is launched after the cache is cleared.

To clear the cache on the App-V client, perform the following steps:

1. Open Regedit.
2. Navigate to the following key:

32-bit systems: HKEY_LOCAL_MACHINE\Software\Microsoft\Softgrid\4.5\Client\AppFS
64-bit systems: HKEY_LOCAL_MACHINE\Software\Wow6432Node\Microsoft\SoftGrid\4.5\Client\AppFS

3. Double-click the value name State and change the value data to 0.
4. Restart the App-V client computer.

Additional Resources

Microsoft Showcase video which covers how to fix common App-V configuration issues: http://www.microsoft.com/showcase/en/us/details/b2e8800e-e84f-4d98-aaec-68f1af00ab08

App-V TechCenter on TechNet: http://technet.microsoft.com/en-us/appvirtualization/default.aspx

Query Words

44-00001004 0a-00000193 0a-10000001 0a-0000e02b 0a-200001f4 64-00000003 2A-80090322 08-10000003 0a-0000E005 0A-0000E0A3

=====

For the most current version of this article please see the following:

2615139: How to troubleshoot applications failing to stream from an App-V management server

J.C. Hornbeck | System Center Knowledge Engineer

How to troubleshoot failures when refreshing the publishing server on an App-V client

J.C. Hornbeck — Wed, 07 Sep 2011 18:12:29 GMT

Here’s a new App-V KB article we published this today. If you ever need to troubleshoot client refresh issues with an App-V server then there’s no better place to start than right here:

=====

Summary

This article covers how to troubleshoot failures when refreshing a publishing server on a Microsoft Application Virtualization (App-V) client. Refreshing the publishing server is also known as refreshing the desktop configuration.
When an App-V client fails to refresh the publishing server, the following event will be logged in the Application event log:

Log Name: Application
Source: Application Virtualization Client
Event ID: 3131
Level: Error
Description:
{tid=8E0:usr=UserAccount}
Failure on Desktop Configuration Server request to URL {rtsp://appv-svr:554/} with header {Host: appv-svr
Content-Type: text/xml
AppV-Op: Refresh
} (rc 19D07F2A-0000274D).

If a refresh fails using the Application Virtualization Client snap-in, the following error occurs in addition to the Event ID 3131 in the application event log:

The Application Virtualization Client could not update publishing information from the server ServerName.
No connection could be made because the target machine actively refused it.
Error code: xxxxxxx-xxxxxx2A-0000274D

Note: The error description and code will vary.
Note: The frequency of a publishing server refresh is controlled by the provider policy on the App-V Management Server. The default policy setting is to refresh the publishing server when a user logs in. You can also manually refresh the publishing server on an App-V client by performing on the following methods:

Method 1
1. Open the Application Virtualization Client snap-in.
2. Click Publishing Servers.
3. Right-click the publishing server and choose Refresh Server.

Method 2
1. Right-click the App-V icon in the notification area and choose Refresh Applications.

More Information

Step 1: Review the Sftlog.txt file on the App-V client

On the App-V client, review the Sftlog.txt file on the App-V client. This log file may include additional information that wasn’t included in the Event ID 3131 in the Application event log.
The default location for the Sftlog.txt is: %systemdrive%\ProgramData\Microsoft\Application Virtualization Client

Step 2: Verify the publishing server settings are correct on the App-V client

To verify this, perform the following steps on the App-V client:

1. In Administrative Tools, open the Application Virtualization Client MMC snap-in.
2. Click Publishing Servers.
3. Right-click the server reporting the error and then click Properties.
4. Verify the publishing server Type.
5. Verify the server name that is specified in the Host Name box.
6. Verify the port that is specified in the Port box.

Step 3: Verify the Application Virtualization Management Server service is started on the App-V Management Server

To verify this, perform the following steps on the App-V Management Server:

1. In Administrative Tools, open the Services MMC snap-in.
2. Locate the Application Virtualization Management Server service.
3. Verify the service is Started.
4. If the service is not started, right-click Application Virtualization Management Server, and then click Start.
5. If the service fails to start, search the Microsoft Knowledge Base for the error message that is reported.

Step 4: Verify the user groups assigned to the provider policy on the App-V Management Server

To verify this, perform the following steps on the App-V Management Server:

1. In Administrative Tools, open the Application Virtualization Management Console.
2. In the navigation pane, expand the server name object, and then click Provider Policies.
3. In the details pane, double-click the required policy. For example, double-click Default Provider.
4. Click the Group Assignment tab.
5. Verify that the users are a member of one of the user groups listed or add a user group that’s missing.

Step 5: Verify the App-V client can telnet to the App-V Management Server and port.

To verify this, perform the following steps on the App-V client:

1. At a command prompt, type telnet ServerName Port, and then press ENTER.

For example, type the following command:

telnet appv-svr 554

2. If the connection succeeds, the window is blank. Press ENTER two times and you will receive the following message:

RTSP/1.0 400 Bad Request
Server: Microsoft Application Virtualization Server/x.x.x.xxxxx [Win32; Windows NT x.x ]
Date: xxx, xx xxx xxxx xx:xx:xx xxx

If the connection is unsuccessful, you receive the following message:

Could not open connection to the host, on port 554: Connect failed

How to troubleshoot Event ID 3017 when refreshing a publishing server

If the App-V client is unable to access the content directory or application .osd file, the publishing server refresh will succeed but the application shortcut (or icon) will not be published to the user and the following event will be logged in the application event log:

Log Name: Application
Source: Application Virtualization Client
Event ID: 3017
Level: Warning
Description:
{tid=9E8:usr=UserAccount}
Could not load OSD file \\appv-svr\content\AppDirectory\Appname.osd

To resolve the Event ID 3017, perform the steps below:

Step 1: Verify the App-V client can access the content directory

On the App-V client, click Start, in the Search or Run line, type the UNC path of the content share (Example: \\appv-svr\content) and then press ENTER.

If the client fails to connect to the content share, verify the UNC path is correct and verify the NTFS and Share permissions on the content directory are correct by performing the steps below.

On the server hosting the content directory, verify the following NTFS and Share permissions are configured on the content directory:

· App-V Users = Read
· App-V Administrators = Read and Write
· Network Service = Read and Write

The default location for the content directory is: %systemdrive%\Program Files (x86)\Microsoft System Center App Virt Management Server\App Virt Management Server\content.

Step 2: Verify the path to the application OSD file on the App-V Management Server

To verify this, perform the following steps on the App-V Management Server:

1. In Administrative Tools, open the Application Virtualization Management Console.
2. In the navigation pane, expand the server name object, and then click Applications.
3. Right-click the application that is generating the Event ID 3017 and select Properties.
4. Verify the OSD path is correct.

Note: The OSD path should be the UNC path to the OSD file (Example: \\appv-svr\content\AppDirectory\Appname.osd)

5. Click OK to close the application properties.

Additional Resources

App-V TechCenter on TechNet: http://technet.microsoft.com/en-us/appvirtualization/default.aspx

Query Words

2A-0000274D 04-00000917 64-00000005 2A-00002745 0A-10000005 0A-10000009 0A-00002002 0A-20000194 64-00000002

=====

For the most current version of this article please see the following:

2615201: How to troubleshoot failures when refreshing the publishing server on an App-V client

J.C. Hornbeck | System Center Knowledge Engineer

New KB: Recommended antivirus or antimalware exclusions when troubleshooting Application Virtualization (App-V) client issues

J.C. Hornbeck — Wed, 06 Jul 2011 18:21:19 GMT

Just a quick note to let you know about a new Knowledge Base article we just published. This one discusses the AV exclusions you should implement when troubleshooting App-V client issues:

=====

Summary

The purpose of this article is to provide information about directories that should be excluded on Microsoft Application Virtualization (App-V) clients if an issue is suspected to be caused by antivirus or antimalware software.

Note We recommend that you temporarily apply these procedures to evaluate a system. If your system performance or stability is improved by the recommendations that are made in this article, contact your antivirus software vendor for instructions or for an updated version of the antivirus software.

The following errors have been observed on App-V clients due to antivirus and antimalware scanning:

The Application Virtualization file system has been dismounted due to a severe error. Report the following error code to your System Administrator.
Error code: xxxxxx-xxxxxx0A-0000E0A3

An unexpected error occurred. Please report the following error code to your System Administrator.
Error code: xxxxxx-xxxxxx0A-0000E005

Note These errors have been associated with antivirus scanning but can be caused by other issues.

More Information

Warning This workaround may make a computer or a network more vulnerable to attack by malicious users or by malicious software such as viruses. We do not recommend this workaround but are providing this information so that you can implement this workaround at your own discretion. Use this workaround at your own risk. If your system performance or stability is improved by the recommendations that are made in this article, contact your antivirus software vendor for instructions or for an updated version of the antivirus software.

To determine if an issue is caused by antivirus or antimalware scanning, the following directories should be temporarily excluded on an App-V client:

Windows Vista, Windows Server 2008 or later
· %USERPROFILE%\AppData\Local\SoftGrid Client
· %USERPROFILE%\AppData\Roaming\SoftGrid Client
· %PROGRAMDATA%\Microsoft\Application Virtualization Client\SoftGrid Client

Windows XP or Windows Server 2003
· %USERPROFILE%\Application Data\SoftGrid Client
· %ALLUSERSPROFILE%\Application Data\Microsoft\Application Virtualization Client\
· %ALLUSERSPROFILE%\Documents\SoftGrid Client

=====

For the latest information please see the following KB article:

KB2576031: Recommended antivirus or antimalware exclusions when troubleshooting Application Virtualization (App-V) client issues

J.C. Hornbeck | System Center Knowledge Engineer

App-V Fix: Streaming RTSP may fail with the Citrix Netscaler Load Balancer

J.C. Hornbeck — Wed, 08 Jun 2011 15:11:44 GMT

Hi everyone, Steve Thomas here and today I wanted to talk about a recent issue where streaming may fail when using a Netscaler load balancer. Load balancers are common when using App-V in an RTSP streaming mode (one that leverages either the management server or the streaming server), however on some third-party load balancers special adjustments may be necessary. For example, streaming RTSP using the RTSP application rules may fail with the Citrix Netscaler Load Balancer.

When this occurs, the Netscaler log (nslog) will show the following entries (or similar):

5736 0 6 rtsp_err_session_creation_failed
5737 0 6 rtsp_err_parse_fail

According to the RTSP RFC, http://www.ietf.org/rfc/rfc2326.txt (Section 12.39 Transport) a client port is an expected parameter. This parameter provides the unicast RTP/RTCP port pair on which the client has chosen to receive media data and control information. is not complying to the standards for the RTSP setup messages.

In this setup message the Netscaler Load Balancer RTSP VIP is expecting a client port after the RTP/SDCP/TCP options in the Transport field but it is not being sent. Network traces will confirm this:

SETUP RTSP://SERVER:554/DIRECTORY/PACKAGE.sft RTSP/1.0
Transport: RTP/SDCP/TCP;unicast,RTP/SDCP/TCP;interleaved=0-1;unicast
SoftricityDirectoryID: dbec6b02-ea1e-46d7-bad1-21fb31c1c09b
SoftricityFileVersion: 4
SoftricityVersionGUID: 1c03b6db-012f-4062-ac0c-f4483f518d23
Connection: Keep-Alive
User-Agent: Microsoft Application Virtualization Desktop Client/v4.6.0.20200 [Win32; Windows NT]
Client-GUID: 00000000-0000-0000-0000-000000000000
Softricity-LogID: 5179
Bandwidth: 100000000
CSeq: 5
RTSP/1.0 200 OK
Server: Microsoft Application Virtualization Server/4.5.3.19480 [Win32; Windows NT 6.1]
Date: Mon, 02 Aug 2010 18:31:50 GMT
CSeq: 5
Session: 758460554
Transport: RTP/SDCP/TCP;unicast;server_port=49162-49163

In addition, the end users will get the following error when attempting to stream using the virtual address associated with the Netscaler RTSP VIP:

The Application Virtualization Server has returned an Internal System error. Report the following error code to your System Administrator.
Error code: 4604EE8-1690140A-200001CD

While the RTSP VIP rules included with the Citrix Netscaler comply with RFC 2326, Microsoft's RTSP implementation with App-V does not. This standard was developed for real time streaming media, and it was adapted for streaming applications.

Netscaler users will have to use the following workaround on the Netscaler configuration:

1.)  RTSP will have to be supported at layer 4 instead of layer 7 ( the application layer).  You can set up a layer 4 VIP on the Netscaler like the following example:

add service appv1 appv1-corp TCP *
add service appv2 appv2-corp TCP *
add lb vserver Appv_Streaming TCP x.x.x.x * -persistenceType SOURCEIP
bind lb vserver

NOTE: x.x.x.x denotes the IP address.

Hope this helps,

Steve Thomas | Senior Support Escalation Engineer

How to fix the “application took too long to be ready to interact with the user” 04-000041E error

J.C. Hornbeck — Tue, 17 May 2011 14:26:18 GMT

Hi everyone, John Behneman here, I’m a Support Engineer on the System Center Virtualization team and I wanted to show you how to quickly and easily fix an issue you might see with some Java based applications when sequenced (virtualized) with Microsoft App-V.

So let’s say you sequence a Java based application like Libre Office, for example. When you try to launch the application, the application does initially try to launch but you notice the launch indicator shows 100% and just sits there for around 10 minutes or so:

Finally after approximately 10 minutes the launch fails and the application terminates with the following error message:

The Application Virtualization Client could not launch <appName>

The application took too long to be ready to interact with the user, possibly because the system was too busy, Try again in a few minutes. If the problem persists, report the following error code to your System Administrator.

Error code: 460281F-0B01F504-00000041E

What’s happening here is that when an App-V enabled application is started, the App-V client tries to detect when the application is ready to interact with the user. With some applications (for example certain java based applications) this detection may fail. After 10 minutes a timeout is triggered and the App-V client assumes that something went wrong during the launch and terminates the application.

The easiest way around this if you see it with your app is to disable the detection feature. This detection feature can be disabled by changing the SUBSYSTEM tag from windows to console in the package OSD files. For example:

Change:
<VM VALUE="Win32">
<SUBSYSTEM VALUE="windows"/>
</VM>

To:
<VM VALUE="Win32">
<SUBSYSTEM VALUE="console"/>
</VM>

If you are deploying your packages using Virtual Application Server you will need to initiate a DC refresh on your App-V client to update the OSD cache and presto this issue should be resolved.

Note that we published this same information today in a new KB article:

KB2195120 - When launching an application virtualized using Microsoft App-V, the application hangs for 10 minutes and then terminates

Enjoy!

John Behneman | System Center Support Engineer

Problem Notification and Automated Best Practices in the App-V 4.6 SP1 Sequencer (Part 2)

J.C. Hornbeck — Tue, 26 Apr 2011 13:18:00 GMT

Yesterday we released Part 1 in this series, and now in Part 2 we will look at the Automated Best Practices throughout the new App-V 4.6 SP1 Sequencer. In Part 1, Problem Notification features were addressed, which are also an important part of Automated Best Practices in the App-V 4.6 SP1 Sequencer.

If you read the App-V Sequencing Guide then you knew about the Best Practices for Sequencing. If you had not read it, well, you weren't going to get much help from the Sequencer itself prior to App-V 4.6 SP1. Now, the Sequencer helps keep your Sequence within Best Practices through the new workflow, pop-up notifications, and help links.

Workflow and New UI

The new UI design makes it clear what actions are available during any specific phase. The significant redesign of the UI and workflow keeps a Sequencing Engineer on the path of Best Practices by dividing Sequencing into more phases while presenting clearly defined choices along the way.

Automatic Package Root

A major limitation and customer complaint of previous versions of SoftGrid and App-V was the 8.3 character naming limitation of the Package Root. The Package Root is the folder created on Q: where the sequenced applications installs, and any Virtual File System mappings are redirected.

In this version of the Sequencer, the 8.3 character limitation is gone. In addition, the Package Root folder is created automatically on the Q: drive to match the given Package Name.

Pop-Up Notifications

Pop-Up notifications will appear when a Sequencing Engineer may need an extra reminder of the correct action to take in a particular phase. One example is the reminder to install to the Q drive once the application's installer launches.

Additionally, the pop-up makes another important appearance when choosing which OS to allow the application to deploy. If the application is sequenced on Windows 7, for example, and a lower OS is selected then this pop-up would appear:

Contextual Help

At the bottom of the screen for each phase of Sequencing is a contextual help link. This link takes you directly to the section of the Help file that is relevant to the Phase of sequencing. The Help file provides more detail about that particular phase of Sequencing and the choices available.

Saving Packages

A common mistake in earlier versions of the Sequencer was to for a Sequencing Engineer to make it all the way to the "tabs" interface of the Package Editor at the end of sequencing, and then revert the Sequencer without saving the package. It was not clear to some users that the package was not saved at this point and a lot of time was lost in having to resequence.

Now, the Sequencer by default has saving the package as an option along the way. The package can be saved with the default settings prior to going to the Package Editor or continue on to the Package Editor without saving.

The Sequencer saves the package with the correct naming and file structure to the desktop by default.

Overall, between the Automated Best Practices and the Problem Notification features, Sequencing Engineers should fine the new Sequencer much easier to use with a much higher success rate for Sequencing.

Steve Bucci | Support Escalation Engineer

Problem Notification and Automated Best Practices in the App-V 4.6 SP1 Sequencer (Part 1)

J.C. Hornbeck — Mon, 25 Apr 2011 17:45:00 GMT

In previous versions of the Microsoft Application Virtualization (App-V) Sequencer, there was little to no built-in guidance from the Sequencer around potential problems or following best practices. With the new Microsoft App-V 4.6 SP1 Sequencer, automated best practices and problem notification throughout the sequencing process help ensure a higher rate of successful sequencing.

In part 1 of this series we will explore the new Problem Notification features of the App-V 4.6 SP1 Sequencer.

In addition to notifying the Sequencing Engineers of potential problems, the Sequencer also provides guidance to resolve the issues detected.

Report.xml

Beginning with the Prepare Computer phase, the Sequencer records any problem notifications in the file "Report.xml." The Sequencer writes to this file in the Prepare Computer, Installation Report, and Completion Report phases and is included in the final package by default. All Problem Notifications, their details, and the recommended actions are included in the Report.xml. If a package does not work as expected after delivery to the client, refer back to this file to begin troubleshooting.

The Completion Report is the final Report.xml that contains all the information written during the Prepare Computer phase, Installation Report, and any final issues detected after configuring the software.

Prepare Computer detects:

Pending reboots
VM not reverted
Services (Defender, Indexing, Defrag, SCCM)
Applications (MMC, Anti-Virus, etc)

Installation Report records and detects:

Excluded files
Drivers
COM+
System Differences
SxS Conflicts
Shell Extensions

Prepare Computer Phase

In the Prepare Computer phase, the Sequencer notifies the Engineer of any running programs or services that may interfere with Sequencing or corrupt the resulting package. Detectors are run for services such as Windows Defender and Windows Search. Additionally, Detectors are run for programs like Anti-Virus.

The Sequencer will attempt to stop some services on its own, and list any that it cannot on the Prepare Computer phase screen. In the SFT-SEQ-LOG.TXT file, services that are stopped by the Sequencer will appear as below:

[10/15/2010 16:46:38 VRB CORE] successfully stopped service: Defragsvc
[10/15/2010 16:46:40 VRB CORE] successfully stopped service: wsearch

Anything detected in the Prepare Computer phase is listed with additional details and/or guidance to stop the listed item. Security related programs, such as Windows Defender or Microsoft Forefront are not stopped automatically and require user interaction.

Clicking the REFRESH button will run another check for running programs or services. If there is nothing running that will interfere with successful sequencing, then instead of a list of programs and services, the message "Your system is ready to create a virtual package, no issues were detected. Click Next to Continue."

Below is an example from "Report.xml " for the Problem Notification of Windows Defender detection.

</ISSUE>
- <ISSUE CODE="196609" SEVERITY="WRN" MODULE="Front-End Component">
<DESCRIPTION SHORT="Windows Defender™ is active." FULL="Windows defender can interfere with creation of a package by accessing files that need to be added to the virtual package or adding extraneous data to the virtual package." />
<RESOLUTION SHORT="Stop this before continuing." FULL="Stop Windows Defender by opening Services from the Start Menu, clicking 'Windows Defender', and clicking 'Stop'." />
<MOREINFO DESCRIPTION="The process name associated with this service is 'WinDefend.exe'." />
</ISSUE>

The logging of the detectors and results are preceded with the <AUTO BP> tag in the SFT-SEQ-LOG.TXT file.

Installation Report

The Installation Report is provided after the Configure Software phase is complete. Information provided in the Installation Report will also be included in report made during the Completion phase.

Files Excluded from Package

The most common message from the Installation Report is the informational message that files were excluded from the package. Double-clicking on this message will provide greater detail on the files excluded from the package. Listed paths usually match paths set in TOOLS>OPTIONS>EXCLUSION ITEMS.

The resolution for this message is always verify package functionality. If a package does not function correctly, there may be excluded paths or files that that may need to be included. This can be modified before another sequencing attempt in TOOLS>OPTIONS>EXCLUSION ITEMS.

The details of the "Files excluded from package" message, as well as any other messages that appear here will be in REPORT.XML

Below is an example from "Report.xml " for Excluded Files.

The Problem Notification and resolution is similar for other issues listed in the bullet points near the top of this article. Overall, the new Sequencer greatly enhances the ability to detect and resolve sequencing related issues. Stay tuned for part 2 where we take a look at the Automated Best Practices throughout the new App-V 4.6 SP1 Sequencer.

Steve Bucci | Support Escalation Engineer

Tips and tricks for troubleshooting App-V and SQL

J.C. Hornbeck — Mon, 21 Mar 2011 15:34:06 GMT

This guide is intended to provide some general troubleshooting tips and tricks in regards to issues you might experience with App-V Management Servers 4.5 and SQL 2005/2008.

When the App-V Management Server notices issues towards the SQL Datastore, in certain scenarios we can see problems such as the “Application Virtualization Management Server” Service (AppVirtServer) failing to start, for example App-V Management Server upgrade failures etc.

When you are experiencing issues, you could use the following tips and troubleshooting tricks to pinpoint the problem.

If you suspect issues regarding SQL connectivity or between the management server and the SQL data store, we have identified some common troubleshooting tips that we use in assisting customers;

1. First, start to look at the App-V Server log (sft-server.log) and check for errors.

When we have a serious issue, this log will point towards SQL but it might not reveal much info. You might see an entry like:

41494 "Failed to establish a connection to the data source ()."

Note that in the context of the App-V server starting up, 41494 messages can be expected and are benign. Only when you see 41494 messages in the context of trying to launch applications can it be considered part of a serious issue.

2. Verify that the App-V Administrators group has a login in SQL. This should be configured during the installation of the App-V Management Server, but in case of a migration, it should be created manually. Make sure this login has SFTadmin and SFTeveryone roles added.

3. Verify that a SQL Security Login is present for the App-V Management Server machine account.

This login is usually created during the installation of the App-V Management Server against the database, but if it is not present, it should be created. If the SQL GUI does not allow the creation of a login for a machine account, use the below SQL statement:

CREATE LOGIN [Domain\machine_name$] FROM WINDOWS
Go

Then verify the newly created login:

SELECT * FROM sys.server_principals WHERE name LIKE 'Domain\machine_name%'
Go

4. Use SQL Server Profiler to find a cause for the failure. A SQL Profiler trace can tell us a lot about what’s going on between the Management Server and the datastore, but if we select too many events, we will be looking for a needle in a haystack. In most situations, the following settings will be useful:

i. Start SQL Server Profiler

ii. Decide whether you want to run the trace or save it as a *.trc. You can also save the trace to a table on one of the existing databases so that you can query the trace output.

iii. On the ‘Events Selection’ tab, check the boxes “Show all events” and “Show all columns”.

iv. Browse to “Security Audit”, right-click and select “Select Event Category”.

v. Browse to “Errors and Warnings”, right-click and select “Select Event Category”.

vi. If necessary, you can also select “Stored Procedures” and “TSQL”, but this will cause your trace to grow large.

This would be sufficient in most cases for connection troubleshooting. When you are troubleshooting on a SQL server that has a lot of activity, you can filter the trace for ApplicationName, “Microsoft Application Virtualization”, or LoginName, domain name\machine name$. You can use SQL Profiler to read a trace, or you can then import it to a SQL table so that you can query it easily.

5. Test that the port is clear all the way through the firewalls using the following steps:

Use port query on the App-V Management Server:

http://www.microsoft.com/downloads/details.aspx?FamilyID=8355e537-1ea6-4569-aabb-f248f4bd91d0&displaylang=en

Run this on the App-V Management Server and use the same DNS name for your SQL server. The SQL Server name can be found on the App-V Management Server at:

HKLM\Software\Microsoft\SoftGrid\4.5\Server\SQLServerName

The App-V database name can be found on the App-V Management Server at:

HKLM\Software\Microsoft\SoftGrid\4.5\Server\SQLDatabaseName

Test to see if port 1433 is listening.

If the tool returns anything else other than listening means we are being blocked somewhere.

If the above works, it means the traffic can get to SQL.

6. If your App-V Service fails to start, test if it does run under a Service Account. This way you can exclude certain aspects and you have a workaround for the time being.

i. Create a service account in AD Users and Computers.

ii. Add this account to the administrator group used for the App-V Administrators.

iii. Run Services.msc, then right-click on “Application Virtualization Management Server” and select Properties.

iv. Click on the Log on tab.

v. Change the NT Authority\Network Service to the service account created in Step 1 and enter the password. Click OK.

vi. Click OK to acknowledge the Services dialog box indicating that the new logon will not be in effect until the service is restarted.

vii. Right-click on Application Virtualization Management Server and select Restart.

viii. Assuming the Support Tools are installed on the App-V Management Server, go to Start, All Programs, Windows Support Tools, Command Prompt.

ix. Enter the following 2 commands:

setspn -A SoftGrid/<FQDN of your machine> <YourDOMAIN>\<YourServiceAccountName>
setspn –A SoftGrid/<NetBIOS name of your machine> <YourDOMAIN>\<YourServiceAccountName>

x. Grant access to the Content share to the Service Account you created.

xi. Add the Service Account to Logon as A Service rights in the Local Security Policy.

xii. Give the Service Account Modify permissions to the Microsoft System Center App Virt Streaming Server\Logs directory.

How to configure the App-V Management Server Service to run as a Service Account

http://blogs.technet.com/b/appv/archive/2008/08/21/how-to-configure-the-app-v-management-server-service-to-run-as-a-service-account.aspx

7. Make sure that when you are migrating and upgrading, that you use the correct order. When you are running SoftGrid against a SQL 2005 database, you should first upgrade to App-V 4.5 before you migrate your database to SQL 2008. SoftGrid 4.1 is not supported for use with SQL 2008.

For more info, please see:

Exploring the SoftGrid Platform v4.1

http://technet.microsoft.com/en-us/library/bb608285.aspx

Here we can see the following:

“SoftGrid supports MS SQL 2000 with Service Pack 4 and MS SQL 2005.”

8. Determining the port the App-V database is running on. You can use the SQL Log File Viewer to check the SQL Server log to determine on what port the App-V database listens on. You can do this using the following steps:

i. Open SQL Server Management Studio.

ii. Browse to “Management” and then “SQL Server Logs”.

iii. Right-click the “Current” log and choose “View SQL Server Log”

iv. Then, you will see the log in a new window.

v. Look for an entry called “Server is listening on…”. This will tell you the port we are currently using.

Madelinde Walraven | Justin Zarb

How to troubleshoot situations where GPO settings are not applying to packages in App-V 4.5 and later

J.C. Hornbeck — Wed, 02 Mar 2011 14:29:00 GMT

Hi, Steve Thomas here. Back in April of 2009, J.C. Hornbeck posted a blog on the changes that 4.5 brought with regards to how group policies affect registry settings:

http://blogs.technet.com/b/appv/archive/2009/04/23/some-insight-into-how-softgrid-and-app-v-4-5-handle-group-policies.aspx

Since the policies are no longer automatically applied to the virtual registry inside the package, additional adjustments will need to be made to packages to ensure that registry policies are applied properly from the local native registry. What we will need to do is adjust the package to not include any policy settings. We can do this at the key level but not at the value level so if you are ok with a policy or virtualized approach to these settings, we can certainly adjust the sequence to exclude these policy settings and that the GPO settings will always take precedence. For most common policies, those keys or values are not captured during the launch phase. However, we have found that there may be some exceptions where these values do get populated during sequencing. This is often due to external factors such as the design of some registry tattooing with some ADM templates or the application was not launched during sequencing.

In a recent case I worked, the customer could not seem to get the domain-supplied policies for Office Communicator to apply to the virtual package. As with any of these issues, the first thing we did was to isolate the registry locations of all potential policy settings. In the case of Communicator, we found them in the following key locations:

· HKEY_LOCAL_MACHINE\SOFTWARE\Policies\Microsoft\Communicator

· HKEY_CURRENT_USER\Software\Policies\Microsoft\Communicator

· HKEY_USERS\.DEFAULT\Software\Policies\Microsoft\Communicator

The issue with the problem package was that these settings were captured and either had the “Override Local key” or “Merge with Local Key” setting. We did not want either in this case. We wanted these keys excluded from the sequence so the values will defer to what is local (passed by policy.)

For example, as you can see in the example below, the Messenger policy sub-keys were actually caught during sequencing and the keys were set to “Override Local Key.”

What we did with the customer’s package was to delete those keys from the package. (I would advise deleting the entire Policies key if it exists.) If you wanted to re-sequence to ensure that these keys would not get captured then you would need to re-sequence Office Communicator with the SPRJ file modified to exclude those keys. These can be adjusted by adding these new ones (Under Tools – Options – Exclusion items.)

You will need to add in the following exclusions with a mapping type of VREG:

· REGISTRY\MACHINE\SOFTWARE\Policies\Microsoft\Communicator

· REGISTRY\USER\%SFT_SID%\Software\Policies\Microsoft\Communicator

· REGISTRY\USER\S-1-5-18\Software\Policies\Microsoft\Communicator

Then start the monitoring phase and sequence Communicator as normal.

For more information on policies and App-V see the following: http://blogs.technet.com/b/appv/archive/2009/04/23/some-insight-into-how-softgrid-and-app-v-4-5-handle-group-policies.aspx

Steve Thomas | Senior Support Escalation Engineer

Fix: Applications virtualized with Microsoft App-V disappear after enabling drive encryption

J.C. Hornbeck — Tue, 30 Nov 2010 14:45:13 GMT

Here's another quick tip on an issue you might run into where an application virtualized with Microsoft App-V may disappear and no longer become streamable after enabling drive encryption on App-V clients.

This is due to the drive encryption locking the associated OSD files. If this happens and if the OSD files are not available after several attempts by the App-V client to access them, the App-V client places the application into the IGNOREDAPPS list.

The resolution to this problem is a two step process:

1. Open the registry on the App-V client and navigate to the following location:

HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\SoftGrid\4.5\Client\IgnoredApps

Under this key you will find a list of applications that the App-V client marked to ignore. Delete all of the affected applications listed under the IgnoredApps key and restart the App-V Client Service.

2. Create the following exceptions for your drive encryption software:

-^%ENV:ALLUSERSPROFILE%\documents
-^%ENV:USERPROFILE%\Application Data\Softgrid Client
-^%ENV:USERPROFILE%\local settings\Application Data\Softgrid Client

Also keep in mind that if you are using a path other than the default for your Global Data Directory or your User Data Directory configured in the client you will have to exclude those paths as well.

Justin Luyt | Senior App-V Support Engineer

New video on how to fix common App-V configuration issues

J.C. Hornbeck — Wed, 30 Jun 2010 14:33:21 GMT

Our very own Steve Bucci, resident Support Escalation Engineer extraordinaire, recently posted another great video he put together that shows you how to troubleshoot and correct some common configuration mistakes by guiding you through the areas to double-check in a typical App-V deployment. Whether you're new to Microsoft Application Virtualization or a seasoned veteran, you'll definitely want to spend the 10 minutes to check out this short video:

For more information on troubleshooting some common App-V issues see this post by Microsoft App-V Support Escalation Engineer Steve Thomas.

J.C. Hornbeck | System Center Knowledge Engineer

How to collect the log files for Microsoft App-V

J.C. Hornbeck — Tue, 04 May 2010 13:38:39 GMT

Our very own Steve Bucci recently posted this great video that shows you how to collect the logs and information from your App-V environment. With almost any App-V support call we will ask you for this information but knowing where to find these logs will prove invaluable when troubleshooting any issue you might run into.

We also have some additional information on the App-V log files here:

Troubleshooting App-V with log files

KB931803 - Log options for SoftGrid Client 4.x

KB930871 - How to interpret the messages in the SoftGrid Virtual Application Server log

J.C. Hornbeck | System Center Knowledge Engineer

Resolution: You may be unable to uninstall the App-V 4.6 Desktop Client

J.C. Hornbeck — Thu, 22 Apr 2010 16:23:37 GMT

We recently uncovered this issue and while we're going to be publishing a Knowledge Base article on it very soon I thought it might be helpful to mention it here as well.

Issue:

Uninstallation of the Microsoft Application Virtualization 4.6 desktop client may fail with no specific error message given. If you collect a verbose MSI log you may see the following:

SoftGrid Action: IsRuntimePackagePolicyCfgValid Details: NET_DOTIMEOUTMINUTES(#0) is invalid

Also note that you can see this same issue when trying to upgrade a previous version of the App-V client to version 4.6.

Cause:

This is a known issue with the App-V client versions 4.5 and later.

Resolution:

To resolve this issue you will need to modify the registry entry for the Disconnected Operation Timeout in the registry.

1. For x86 clients, open Regedit and navigate to the following location:

HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Softgrid\4.5\Client\Network\

For 64-bit clients, open Regedit and navigate to the following location:

HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\SoftGrid\4.5\Client\Network

2. Change the DOTimeoutMinutes value to the default which is 129600 (Decimal) or 1fa40 (Hex).

3. Reboot the client.

Once rebooted, you should now be able to successfully uninstall the App-V client.

Steve Thomas | Senior Support Escalation Engineer

Troubleshooting Common RTSPS Issues with App-V

J.C. Hornbeck — Tue, 09 Mar 2010 15:05:55 GMT

The following are some of the common problems customers may run into when using RTSPS. These are based on specific errors and their likely causes. Please bear in mind that some of these errors may occur under other contexts so understand that the context here is strictly related to RTSP.

Common RTSPS Problem #1:

Users receive the following error on publishing refresh or when trying to stream an application via RTSP:

"The specified Application Virtualization Server has shut down the connection. Try again in a few minutes. If the problem persists, report the following error code to your System Administrator."

"Error Code: xxxxxx-xxxxxx0A-10000009"

This error can occur when one or more of the following conditions are true:

1.) Network Service Account (the account used to authenticate the App-V Server Service does not have access to path to the machine keys:

Location for Windows 2008/R2:

%ALLUSERSPROFILE%\Microsoft\Crypto\RSA\MachineKeys

Location for Windows Server 2003:

%ALLUSERSPROFILE%\Application Data\Microsoft\Crypto\RSA\MachineKeys

2.) The Certificate Store lacks the necessary permissions for the NETWORK SERVICE account. When this occurs there will also be an accompanying event in the Application Event Log:

Source: Application Virtualization
Event ID: 44955
Description: Certificate could not be loaded. Error code (-2146893043) Make sure the Network Service account has proper access to the certificate and its corresponding private key file.

When this occurs in Windows Server 2003 as the underlying operating system, you will need to modify the actual certificate store permissions using the WinHttpCertCfg command.

Additional information on using the WinHttpCertCfg.exe tools is available at the link below:

http://msdn.microsoft.com/en-us/library/aa384088(VS.85).aspx

The syntax you will need to use is:

winhttpcertcfg -g -c LOCAL_MACHINE\My -s Name_of_cert -a NetworkService

You can then verify that the security context was properly added by listing the permissions on the certificate using this syntax:

winhttpcertcfg -l -c LOCAL_MACHINE\My -s Name_of_cert

Windows Server 2008 makes the process of changing the ACLs on the private key much easier. The certificates MMC Snap-in can be used to manage private key permissions. To do so, follow these steps:

1. Create an MMC with the Certificates snap-in that targets the Local Machine certificate store.

2. Expand the MMC down to the certificate you need as shown in the image below and select Manage Private Keys.

3. Use the Security tab to add the Network Service account with Read access. These instructions are also available in the Security Operations Guide at the following link:

http://technet.microsoft.com/en-us/appvirtualization/cc843994.aspx

http://download.microsoft.com/download/f/7/8/f784a197-73be-48ff-83da-4102c05a6d44/AppV_Secuirty_Operations_Guide.docx

Common RTSPS Problem #2:

Users receive the following error on publishing refresh or when trying to stream an application via RTSP:

"No connection could be made because the target machine actively refused it."

"Error code xxxxxx-xxxxxx2A-0000274D"

This error can happen during publishing/refresh or during streaming. It usually occurs when one or more of the following conditions are true:

Potential issue with firewall on the server running the App-V Management server or streaming server. Either create and inbound rule that allows RTSPS Port 322 or an inbound rule that allows for the application/service sghwdsptr.exe (for the dispatcher) and sghwsvr.exe (for the core.)
Listener Port Configuration for server is configured incorrectly or the port is specified incorrectly in the OSD file.

Assuming your configuration is the default in 4.5 and later for RTSPS, watch for the following common URL errors:

a. <CODEBASE HREF="RTSP://APPVSRV:554/off2010RC1_mnt/off2010RC1_mnt.sft"

- Wrong Protocol

b. <CODEBASE HREF="RTSPS://APPVSRV:332/off2010RC1_mnt/off2010RC1_mnt.sft"

- Wrong Port (RTSPS default port changed with 4.5)

c. <CODEBASE HREF="RTSPS://APPVSRV:554/off2010RC1_mnt/off2010RC1_mnt.sft"

- Wrong Port (forgot to change port after changing protocol)

Common RTSPS Problem #3:

Users receive the following error on publishing refresh or when trying to stream an application via RTSP:

"The certificate chain was issued by an authority that is not trusted."
"Error code: xxxxxx-xxxxxx2A-80090325"

This error occurs when the cert being used on the App-V server either is not issued by a trusted authority, or if it is self-signed certificate, it has not been imported into the trusted root folder of the server's certificate store.

Viewing the certificate from within the Certificates MMC-Snap-in (for local computer) will show the certificate with a red X denoting that it is not trusted.

Self-signed certificates can be trusted by installing the certificate in the Trusted Root Certification Authorities Store. If you are using an online CA, you will need to collect certain data and verify the certificate before doing further PKI troubleshooting.

Load the Certificates MMC snap-in focused on the local computer. This requires Administrator credentials.

1.) Click Start, then Run, type mmc.exe, and press Enter.

2.) Click the File Menu and select Add/Remove Snap-in.

3.) Click the Add button. Select Certificates from the list of snap-ins, and click Add.

4.) Select Computer Account, and click Next.

5.) Select Local Computer, and click Finish.

6.) Click Close, and then click Ok.

7.) Expand Certificates (Local Computer).

8.) Expand the Personal folder, and select Certificates.

9.) Export the certificate. In the list of certificates, select the one you want to export.

10.) Right-click and select All Tasks and then Export…

11.) This will start the Certificate Export Wizard. Click Next.

12.) Select No, do not export the private key. Click Next.

13.) Select DER encoded binary X.509 (.cer) and click Next.

14.) Enter the location and file name for the file, and click Next. I usually use appvcert.cer

15.) Click Finish. Click Ok.

After you are finished exporting the .CER file, you can then run the following command against it to verify the certificate:

certutil -urlfetch -verify appvcert.cer >appvcert.txt

As you can see, I have redirected the output to this command to a file. To quickly search for status errors, use the findstr command to list the error status for each certificate in the chain.

Common items you may want to search for include:

0x800b0109
0x80092013
0x80072ee7
UNTRUSTED
failed
terminated
Expired

To determine if the certificate is published in Active Directory run the following command:

certutil -store "ldap:///CN=Certification Authorities,CN=Public Key Services,CN=Services,CN=Configuration,DC=ForestRootDomain?cACertificate?one?objectClass=certificationAuthority" SerialNumber

The serial number can be found by double-clicking the exported *.CER file and clicking on the "Details" tab.

Common RTSPS Problem #4:

Users receive the following error on publishing refresh or when trying to stream an application via RTSP:

"The target principal name is incorrect."
"Error code: xxxxxx-xxxxxx2A-80090322"

This is usually caused by the path to the server in the OSD mismatching the name specified for the certificate. It can also occur if the name specified for the publishing server via (RTSPS) does not match the name of the same machine specified in the OSD. For example: Name specified for the certificate is the FQDN but the OSD only references the NETBIOS name.

Common RTSPS Problem #5: Poor Streaming Performance

This was first addressed in Hotfix Package 5 for Microsoft Application Virtualization 4.5 Cumulative Update 1: August 2009 http://support.microsoft.com/?kbid=973873. It has also been included in Service Pack one for App-V 4.5.

Common RTSPS Problem #6

When using RTSPS and a 3rd party Certificate Authority some clients may receive the following error message during a refresh after a user login:

The revocation function was unable to check revocation because the revocation server was offline.
Error code: 450260-24C02F2A-80092013

Please see this blog post for the resolution:

http://blogs.technet.com/appv/archive/2008/09/18/softgrid-app-v-client-receives-an-xxxxxx-xxxxxx2a-80092013-error-during-refresh.aspx

General Tips on using self-signed certificates:

The use of certificates is much more streamlined for App-V version 4.5 and above. When constructing a certificate for use with RTSPS, you can leverage a web server template and share the certificate with an IIS service (especially if you are securing web management service access.)

If bypassing Internal CA's or 3rd-party certificates, you can make use of self-signed certificates. If your App-V Server is residing on a Windows 2003 Server, you can use the SelfSSL tool from the IIS Resource Kit Utility. SelfSSL (SelfSSL.exe) can help you generate and install a self-signed SSL certificate.

Information about the IIS Resource Kit can be found here: http://support.microsoft.com/kb/840671

The IIS 6.0 Resource Kit can be found here: http://www.microsoft.com/downloads/details.aspx?FamilyID=56fc92ee-a71a-4c73-b628-ade629c89499

Here are some syntax hints as well:

selfssl.exe /N:CN=app-v-ms /K:1024 /V:7 /S:1 /P:443

- Creates, Installs, and Binds the Certificate.

selfssl.exe /T

- Adds the certificate into the Trusted Root Store

You can then verify the certificate using the Certificates MMC Snap-in or CERTUTIL.EXE

Ensure the certificate you generate is for Server Authentication (or both Server and Client Auth.) Pay close attention to Name and Friendly Names

Windows Server 2008 has a built-in feature inside of the IIS Manager for creating self-signed certificates.

1.) Open IIS Manager and navigate to the server node (identified by your server name.)

2.) In Features view, double-click Server Certificates.

3.) In the Actions pane, click Create Self-Signed Certificate.

4.) On the Create Self-Signed Certificate page, type a friendly name for the certificate in the Specify a friendly name for the certificate box, and then click OK.

Hope this helps,

Steve Thomas | Senior Support Escalation Engineer

After upgrading a Microsoft App-V server, client machines may receive 0A-20000194 errors when trying to launch applications

J.C. Hornbeck — Tue, 02 Mar 2010 16:15:31 GMT

Here's an issue we see every once in a while and since I didn't see the details published anywhere I thought I'd go ahead and do that here.

Consider the following scenario:

1. You install the App-v 4.x server and set the packageroot (content directory) to an alternate location (not the default).
2. You delete the included Default Application, package and content from the disk.
3. You upgrade the server to a newer version of App-V.

After completing the steps above, when clients attempt to stream a virtualized application they receive one of the following errors:

A network error occurred error : XXXXXXXX-XXXXXXXX-20000194

The package requested could not be found in the system data store or the files associated with this package could not be found on the server. Report the following error code to your System Administrator.

Error code: xxxxxxx-xxxxxx0A-20000194

Cause:

This is caused by an invalid value for SOFTGRID_CONTENT_DIR in the registry.

Resolution:

To resolve this issue, complete the following steps:

1. Start RegEdit

2. Navigate to the following key:

32-bit Management Server: HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\SoftGrid\4.5\Server

64-bit Management Server: HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\SoftGrid\4.5\Server

3. Set the SOFTGRID_CONTENT_DIR value to the proper content directory location.

4. Restart the server.

More Information:

During the server upgrade, the installer first searches for the defaultapp.sft file location. After getting the defaultapp.sft file path, the installer then gets the content directory path from the SOFTGRID_CONTENT_DIR registry key. If the defaultapp.sft file path is not the same as SOFTGRID_CONTENT_DIR registry entry, the installer overrides the value of SOFTGRID_CONTENT_DIR. If there is no defaultapp.sft file (i.e. it has been deleted), the installer overrides the value of SOFTGRID_CONTENT_DIR anyway with the default installation content directory (PROGRAM_FILES\Microsoft System Center App Virt Management Server\App Virt Management Server\Content).

This is an issue we're aware of and it should be fixed in the next version of the server, but if you run into this in the mean time this is how you can fix it.

J.C. Hornbeck | System Center Knowledge Engineer

Solution: Licensing error when you run an Adobe Creative Suite 4 Application on a x64 App-V client platform

J.C. Hornbeck — Tue, 02 Feb 2010 14:35:34 GMT

When you virtualize one or more applications from Adobe's Creative Suite 4 (CS), such as Acrobat Reader or Adobe Photoshop, you may receive the following error message:

Licensing for this product has stopped working.
You cannot use this product at this time. You must repair the problem by uninstalling and then reinstalling this product or contacting your IT administrator or Adobe customer support for help.
Error:129:6

Note This error occurs within 15 minutes after you start a virtualized instance on the x64 Application Virtualization (App-V) client.

You may also receive the following error message:

This product has encountered a problem which requires that you restart your computer before it can be launched.
If you continue to see this message after restarting your computer, please contact either your IT administrator or Adobe technical support for help, and mention the error code shown at the bottom of this screen.
Error:148:3

This issue as well as the cause and resolution are discussed in the new App-V Knowledge Base article below:

KB979595 - Licensing error when you run an Adobe Creative Suite 4 Application on a x64 App-V client platform

J.C. Hornbeck | System Center Knowledge Engineer

Solution: You receive an access denied error message when you try to start the APP-V client console

J.C. Hornbeck — Thu, 07 Jan 2010 23:43:59 GMT

You may receive an access denied error message when you start the Microsoft Application Virtualization 4.5 (APP-V) client console by using the local administrator account. Additionally, you may receive an error message that resembles the following:

Error code: xxxxxx-xxxxxx33-00000005

If you're running into this issue then we just published a new Knowledge Base article you'll probably be interested in:

KB978554 - You receive an access denied error message when you try to start the APP-V client console

J.C. Hornbeck | System Center Knowledge Engineer

A C++ program compiled using a virtualized version of Visual Studio 2008 may not run if debugging is enabled

J.C. Hornbeck — Tue, 20 Oct 2009 14:56:53 GMT

If using a virtualized instance of Visual Studio (sequenced according to the instructions at http://technet.microsoft.com/en-us/appvirtualization/dd334515.aspx), when users build a C++ project in Visual Studio 2008 with debugging enabled the compiled program may not run and produce an error in the console window and/or the interface:

ERROR in Visual Studio Interface:

"Unable to start program 'path_to_program\program_name.exe'."

"This application has failed to start because the application configuration is incorrect. Review the manifest file for possible errors. Reinstalling the application may fix this problem. For more details, please see the application event log."

ERROR in console window:

"The system cannot execute the specified program."

This problem does not occur when the project is set to release mode.

This is caused by the debug run-time modules not being captured inside the virtual application package because of how the Application Virtualization side-by-side process works during sequencing. The debug runtimes are not used, so they are left out.

In order to work around this, the user needs to install the debug runtimes on the local machine. They can be built out of Visual Studio 2008 as follows:

1. From Visual Studio, navigate to "File" - "New" - "Project."

2. Select and expand "Other Project Types," then "Setup and Deployment."

3. In the Visual Studio Templates area, select "Setup Project." Click "OK."

4. In the Project window, Right-click "Setup1" and select "Add" - "Merge Module."

5. Select all debug runtimes desired from the list.

6. Click Open

7. Right-click Setup1 and select "Build."

8. Execute the project. It will install the modules you selected.

Hope this helps,

Steve Thomas | Senior Support Escalation Engineer

Opening Visio documents from a SharePoint site using Edit in Microsoft Visio may fail with App-V

J.C. Hornbeck — Mon, 29 Jun 2009 17:08:00 GMT

On an App-V 4.5 client that is running a sequenced (virtualized) version of Microsoft Visio 2003, when attempting to open a Visio document from a SharePoint site by clicking the drop-down and selecting "Edit in Microsoft Office Visio" it may fail with the following error:

The document could not be opened for editing. A Windows SharePoint Services compatible application could not be found to edit the document.

To resolve this issue complete the following steps:

1. On the local computer, install Windows SharePoint Services Support from the Office installation media. It is listed under Office Tools. In my testing, both the version from Office 2007 media and Office 2003 media worked.

2. Add the following registry keys to the local machine:

[HKEY_CLASSES_ROOT\.vsd]
@="Visio.Drawing.11"
"Content Type"="application/vnd.visio"
"SoftGridSource"="{9C0394B8-E606-49C1-BF51-3DB13E127A77}"

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\App
Paths\visio.exe]
@="\"C:\Program Files\Softricity\SoftGrid for Windows Desktops\sfttray.exe\"
/launch \"Microsoft Visio 2003 1.0""
"UseUrl"="1"
"SaveUrl"="1"

Note that in this example "Microsoft Visio 2003 1.0" is the name of the package used in this example. You will need to use the name of the package in each specific case.

[HKEY_CLASSES_ROOT\Visio.Drawing]
@="VISIO Drawing"

[HKEY_CLASSES_ROOT\Visio.Drawing\CLSID]
@="{00021A14-0000-0000-C000-000000000046}"

[HKEY_CLASSES_ROOT\Visio.Drawing\CurVer]
@="Visio.Drawing.11"

Alternatively, you could add a short-cut to IE into the Visio package and launch the SharePoint site from IE that has been launched within the virtual environment.

Michelle Foley | Support Escalation Engineer

What is the App-V IgnoredApps key and why should I be aware of it?

J.C. Hornbeck — Thu, 28 May 2009 16:37:12 GMT

The IgnoredApps key is used to mark an inconsistency between a locally-cached .OSD file which is either corrupt or not accessible for various reasons and its original location. There could be many reasons for this, including but not limited to:

Malformed XML
File corruption
Access Permissions erroneously placed on the cache

Location

The Key is located in the registry here:

HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Softgrid\4.5\Client\IgnoredApps

The value is a String Data type that correlates to the UNC or URL path to the OSD file from the Content Path, and will be the time stamp of the OSD file in the cache.

Examples of the various reasons for an application to appear here would be:

Unable to access or load OSD file from cache and cannot replace it with streamed or published copy due to access issues.
Unable to access or load AppFS storage from cache and cannot replace it with streamed or published copy due to access issues.
Unable to access or load icons from cache and cannot replace it with streamed or published copy due to access issues.

A common error message you may see that can result from this is:

The Application Virtualization Client could not launch the application you requested.
The specified application does not exist. Check the name you specified, and then try again.
Error Code: xxxxxx-xxxxxx04-00000A09

Why is this Key Here?

The Application Virtualization client keeps a list of OSD files that have been rejected in this key. The primary intent is to help optimize DC refresh (Publishing Refresh) by avoiding downloads of OSD files that cannot be used to create an application for the current user/machine, and have not changed since the last time they were downloaded (and hence, would still not be usable). When the OSD is downloaded on an attempt to create/add a new application, if the OSD is rejected for a reason that is not transient (unsupported OS, XML is invalid, etc.), an entry for it will be added to IgnoredApps. On every OSD download for a file that is not already cached, the list in the registry will be checked, and if an entry for the OSD path is found, the timestamps will be compared. If the timestamp is different, the IgnoredApps entry will be removed, the OSD file will be downloaded, and the parsing will continue (which may fail again, leading to a new entry on IgnoredApps). If the timestamp is the same, the OSD file will not be downloaded, and the attempt to create the application will fail.

The list is applied dynamically (checks are made in the registry each time they are applicable). The list is applied to any non-cached OSD path (local or remote).

You can manually add, remove, or modify entries, to prevent a download or force a download.

Removing these entries:

Even if the underlying issue has been corrected, you will still need to remove the entry(s) under this key manually. Once you corrected the underlying issues, remove the entries under the keys. You will need to restart the Application Virtualization Client service upon doing so.

Steve Thomas | Senior Support Escalation Engineer

Resolving a couple common SoftGrid/App-V Management Console issues

J.C. Hornbeck — Thu, 30 Apr 2009 22:57:00 GMT

I had a support call on this the other day and realized that there really isn’t a good blog post on it so I decided I might as well go ahead and correct that situation now. You may run into an issue where the console just randomly crashes, or it crashes and generates either a 0000C800 or 0000C81D error code. There are a couple different causes and workarounds for something like this but the first thing to check for is asymmetrical package icons and either remove them or modify them to be symmetrical. For more details on this issue and how to correct it see:

KB942687 - When you use the SoftGrid Management Console, it may crash

You can find a free Icon editor online to modify the icons to be symmetrical. One that I have used is “Icon Snatcher” but I’m sure there are others that work just as well:

Icon Snatcher 3.6.1

Another scenario where you might see random console crashes is when the machine has more the 2 gigabytes of physical memory and more than ~1000 icons exist for all the packages loaded on the App-V Management Console. If this is your issue then you may also see an error like:

Event ID: 5000
Source: .NET Framework 2.0 Error
Description: The description for Event ID ( 5000 ) in Source ( .NET Runtime 2.0 Error Reporting ) cannot be found. The local computer may not have the necessary registry information or message DLL files to display messages from a remote computer. You may be able to use the /AUXSOURCE= flag to retrieve this description; see Help and Support for details. The following information is part of the event:
clr20r3, mmc.exe, 5.2.3790.2617, 43c70d1c, mscorlib, 2.0.0.0, 4333ab80, 4052, 27, pszqoadhx1u5zahbhohghldgiy4qixhx, NIL.

Keep in mind that if you don’t see this error that doesn’t necessarily mean you’re not running into the 2GB issue.

Note: This is really an issue with MMC.EXE, not SoftGrid/App-V specifically.

To resolve this issue, or to simply test whether this is the issue you’re running into, use one of the methods below:

1. Use MSCONFIG | Boot.ini Tab | Advanced Options | /MAXMEM=2048 (check this box and set memory to 2048) click OK, OK again and reboot the server.

2. Remove the physical memory from the box so it has a maximum of 2048 Meg of memory

3. Install the SoftGrid/App-V Management Console on another machine that only has 2048 Meg of Memory and manage the servers from this Management console.

4. If you are running App-V version 4.5 you can add the LoadConsoleIcons Dword and set it to 0 to the following hive:

x86: HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\SoftGrid\4.5\Management Console
x64: HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\SoftGrid\4.5\Management Console
Name: LoadConsoleIcons
Type: REG_DWORD
Value: 0

Once you determine that this is your issue then you simply have to decide which of the methods above you’ll want to employ as a more permanent solution.

Hope this helps,

John Behneman | App-V and SCVMM Support Engineer

Potential issue upgrading a package sequenced on SoftGrid 4.2 with the App-V 4.5 Sequencer

J.C. Hornbeck — Tue, 03 Mar 2009 16:54:40 GMT

Here’s another issue we came across during the 4.5 beta that I thought deserved a mention here since it’s something a few of you may end up trying yourself. What happens is that when opening or upgrading a package sequenced on SoftGrid 4.2 with the App-V 4.5 Sequencer you may receive a Windows Installer error when you attempt to add a service pack or add another application to the package. The MSI errors encountered are usually similar to the following:

2755 Server returned unexpected error [2] attempting to install package [3].

This can occasionally happen because when upgrading a package from SoftGrid 4.2 you may end up with a mismatch of Windows Installer system files that were included by default in 4.2 and the Windows Installer libraries locally installed on your sequencing workstation.

To resolve this follow the steps below:

1. With the package open in the App-V4.5 sequencer, delete the following files from both “Files” and the “Virtual File System” tab by expanding the VFS\CSIDL_SYSTEM folder and highlighting and pressing the delete key on the following files:

cabinet.dll
msi.dll
msiexec.exe
msihnd.dll
msimsg.dlll

2. If you opened the package for Upgrade you will also need to delete these same files from the Q: drive.

Q:\Pckg.v01\VFS\CSIDL_SYSTEM

3. Once you complete the steps above, save the package and close the sequencer, reset the sequencer back to a clean state, then open this package up for an upgrade and complete your upgrade as you normally would.

That should do it. I must also extend a special thanks to Michelle Foley for doing all the initial work and for sending this one to me. Thanks Michelle!

J.C. Hornbeck | Manageability Knowledge Engineer