This seems to be  most common on Windows 2008 and i guess it’s because of the UAC and the fact that opening the Control Panel isn’t running in administrative mode.

To work around this you need to run the msiexec command on the correct installation GUID from an administrative command prompt.

Besides running through the registry to find the GUID, one of the easier ways is this:

1. Open an administrative command prompt.
2. run wmic product
3. Locate your product by its name, the GUID (looks a bit like this {25097770-2B1F-49F6-AB9D-1C708B96262A}) directly after that is the one you want. Copy it.
4. run msiexec /i <PASTEYOURGUIDHERE>
5. Modify the agent as pleased

That’s pretty much it. Good luck.

Check the file permissions! And if that doesn’t help?

Check the share permissions! Yes, if it is a shared folder. And if that doesn’t help?

Check the file ownership! Great! But then what?

Well, the file could be in use, and then you would have to shut the locking process down and perhaps kick a user out. In a really bad scenario it could also be a symptom of a broken filesystem, a reserved filename (like “lpt1” or “PRN”) or even an invalid name (silly things like a space in the beginning or the end of a filename).
Another possible reason could actually be that the path to the file or folder is too long. You won’t actually get an error telling you that the filepath exceeds the 255 characters Windows can handle but a simple “Acces Denied”.

There are some, more or less tedious, work-arounds for the problem. Like renaming, starting from the root, all the directories to shorter ones or using the old DOS (8.3, like “dokume~1.doc”) names that windows can auto-generate for you. Personally, I have two favourite ways of handling this.

1. Map the parent-directory of the file/folder you are trying to access/delete as a network drive and access your files that way.
   This is particularly useful if the folder you are trying to access a DFS-share or perhaps a share on the central fileserver filepaths like “\servername01Central ProjectsCentral ServicesIT DepartmentDevelop Methods for Automatically Deploying New Central Servers2.2.1 Auto-Deploying SQL-Server 2005 ClusterDocumentsPreparationsWhitepapersSQL Server 2005 Failover Clustering White Paper.doc”
2. Create a new share to a folder further down the hierarchy. This works locally too if you are logged on to, say, SRV01, you create a new share on “D:FilesharesCentral ProjectsCentral ServicesIT DepartmentDevelop Methods for Automatically Deploying New Central Servers” called “Autodeploymethods” and access it from “\SRV01Autodeploymethods”. That way the filepath doesn’t exceed 255 characters.

Now. When designing fileservers, you really should think about how deep the filepaths may get. This is especially true on DFS-shares since you might have to deal with the full FQDN too, and not only the actual folder structure. Many big corporations I know uses “codes” for departments and assign a project ID (quite simply a number or maybe an abbreviation) to each project and uses theese for the fileshares too. Another scenario that could lead to similar problems are intranet sites where users can create and manage their own subsites and where filenames and folders are not stored in a database.

I have only seen this phenomena on Windows systems so far, and I’ve actually used a linux Live-CD on occasion when admin access is denied.

Read More:
http://support.microsoft.com/kb/320081