Feature pages are design documents that developers have created while collaborating on oVirt.

Most of them are outdated, but provide historical design context.

They are not user documentation and should not be treated as such.

Documentation is available here.

VM Screenshot


Allow users to take screenshot of a running VM


  • Name: Saif Abu Saleh (sabusale)


Users need a way to see the current state of their VMs for debug purposes, when they don’t have console access


Not supported


VDSM will call Libvirt screenshot API [2]

Libvirt will call QEMU to capture the screenshot

then VDSM will return ppm file content as base64 string


Backend will call VDSM to capture screenshot and will receive base64 string screenshot and then base64 string will be converted to byte array and returned


Rest API will call Backend to capture the screenshot

Rest API call is synchronous operation

Rest API implementation is intended for debugging only

API call Example:


      POST /vms/1234/screenshot
      Accept: application/xml

Response is a byte array of the screenshot


Image format issues

ppm is not a standard screenshot format, and users may have issues to see the screenshot, and it is not straight forward to display it in HTML[4], unlike e.g. png base64

issue[5] opened to QEMU to add screen dump as PNG

other option is to convert the ppm content to other standard format like png ways to convert it:

  • Convert it in VDSM using PIL image library: example
  • In Engine use java API to convert it using ImageMagick

Future work

  • Add the functionality to web UI
  • Libvirt screenshot API provides option to capture screenshot by screen number (in case there are multiple screens).

    for now by default 0 only used(first screen), add the option to take screenshot by screen number

  • Change generate screenshot API call to be asynchronous