Guide for Image Authors

Only instructors registered with the classroom administrator (ICT department) may become image authors.

During registration, you will receive the current classroom password and classroom instructor password (used for image upload and image scheduling).

Boot Menu

The boot menu may contain many entries, but only a few are reserved for image authors.

Upload Image

This option is protected by the classroom password.

Select “Upload image” from the boot menu.
After the application starts, log in using the classroom instructor username and password, enter the name of the new image (the application verifies that the name is not already in use), and provide an image description.
Both of these fields are displayed when selecting an image.
Fill in the remaining image parameters (the most important setting is whether students are allowed to know the administrator password; such images must not be installed outside the networking and experimental classrooms).
After confirming the creation of the new image, it will be uploaded to the server.
When the upload is complete, the application waits for user confirmation and therefore does not automatically shut down the computer.

Frequently Asked Questions

How do I prepare an image?

The Upload Image function assumes either a DOS or GPT disk label and archives everything from the first sector to the end of the selected partition, plus an additional data block (currently 64 MiB). This allows a valid swap partition to be stored after the primary partition because only its header needs to be copied.

In the future, we may switch to copying the exact size based on the size of the last stored partition. Swap UUIDs and labels may then be initialized according to a predefined scheme. From a long-term perspective, it is safe to initialize swap space with your own values during the system’s first boot.

If you want a small image and a more complex disk layout, place the operating system on the first partition and include scripts that format or otherwise prepare additional filesystems during the first startup of the installed operating system.

The image is compressed before being stored, so before creating a new image it is highly recommended to overwrite the used partitions with zeros, for example using the dd utility.

If you use an extended partition in a DOS partition table, that partition cannot be selected as the copy limit. Copying can only end at a logical partition located inside it.

How do I prepare and reduce the size of a Windows image?

Remove all unnecessary language packs.
Uninstall all unnecessary software.
Disable pagefile.sys in the virtual memory settings.
Disable hiberfil.sys using powercfg.exe -h off.
Defragment the disk using dfrgui.

Zero free space using sdelete -z drive:.

SDelete - Secure Delete v1.51
Copyright (C) 1999-2005 Mark Russinovich Sysinternals
www.sysinternals.com
usage: delete [-p passes] [-z|-c] [drive letter]
-c Zero free space (good for virtual disk optimization)

How do I delete an image?

Contact the classroom administrator.

My image disappeared, but I never requested its deletion.

Images that have not been used for at least one year (> 366 days) become candidates for deletion. They are not deleted automatically but only when disk space is needed or when the image selection menu becomes excessively long.

My image disappeared, but I never requested its deletion, and it was new and actively used.

It was most likely a Unix image, and remote SSH access from the classroom server was not working, preventing remote administration (shutdown/reboot) of the workstation. If sshd is running and access from the server is allowed (server IP address), then the uploaded SSH key is probably missing from /root/.ssh/authorized_keys.