Bug 736526

Summary: Virtio Win - not possible to "Change an existing device to use the para-virtualized drivers"
Product: Red Hat Enterprise Linux 6 Reporter: John Brier <jbrier>
Component: doc-Virtualization_Host_Configuration_and_Guest_Installation_GuiAssignee: Laura Bailey <lbailey>
Status: CLOSED NEXTRELEASE QA Contact: ecs-bugs
Severity: medium Docs Contact:
Priority: medium    
Version: 6.3CC: emcnabb, jskeoch, lnovich, rhod, tburke, vrozenfe
Target Milestone: rcKeywords: Documentation
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2012-03-07 04:29:55 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Deadline: 2012-01-27   

Comment 6 Laura Novich 2012-01-24 08:41:43 UTC
I would ask Vadim Rozenfeld for more info

Comment 7 Laura Novich 2012-01-24 08:44:40 UTC
Also the bug doesn't contain enough information to see exactly where the problem is w/in the existing documentation.

I am wondering if the title should be changed, to reflect that it is a Windows installation?  I changed it, but if I am incorrect, pls change it back....

Comment 8 Vadim Rozenfeld 2012-01-24 20:56:09 UTC
Sure, it is not possible to update IDE device with viostor driver. They are just incompatible, and Windows installer will refuse to do it, no matter how hard we try. If we are taking this scenario, then you are absolutely right - the docs are really misleading. 
But I believe, they are talking about different situation: 
- a system (VM) has no viostor driver pre-installed; 
- a new virtio-blk device has been recently added to the system;
- on startup Windows recognized a new storage device plugged in;
- Windows PnP manager failed to find a device driver, which matches the new HW
  and installed a stub "SCSI controller" driver and put it under "Other devices"
  category;
In this case we do "update" driver. But we update a stub with a real driver.
Actually the same situation happens with the rest of out drivers - net, balloon and serial. Windows recognizes PCI class IDs and puts devices under predefined category, according to these ids.

So, we are not updating in-box drivers, which are running on top of emulated devices with our para-virtualized drivers. We update default (stub) drivers,
installed by Windows, with our functional device drivers.

Best,
Vadim.