d246b84912
The user have no way to set anything before starting the template for the first time. Especially firewall settings, netvm etc. So to not expose the template to the outside world, disable networking there. |
||
---|---|---|
appmenus_generic | ||
prepared_images | ||
rpm | ||
rpmbuild | ||
tests | ||
.gitignore | ||
appvm.buildlog | ||
build_template_rpm | ||
builder_fix_filenames | ||
builder_setup | ||
cleanup_image | ||
create_template_list.sh | ||
functions-name.sh | ||
functions.sh | ||
Makefile | ||
mount_root.sh | ||
prepare_image | ||
qubeize_image | ||
README | ||
templates.spec | ||
umount_kill.sh | ||
version |
The Template Builder ====================== 1) First, create a clean image of Fedora Linux install. You can use the fedoraize_image script for this: fedorize_image <image> <file-with-packets-lists> NOTE: The fedorize_image script can be also used to update an image -- it automatically assumes the update mode if the <image> already exists. 2) Update symlinks in rpms_to_install/ using the create_symlinks_in_rpms_to_install_dir.sh script. The script reads the core, gui, and xen packages versions from version_{core,gui,xen} files and creates appropriate symlinks in rpms_to_install/ dir. Verify that all symlinks are correct. 3) Now, you can "qubeize" the image. This must be done as root. E.g.: # ./qubeize_image <img> <template-name> 4) Finally, build the Template RPM (do it as normal user, not root), e.g.: $./build_rpm <name> The resulting rpm is stored in rpm/noarch directory.