Thanks for your reply.

@sebastian-roth said in [Can I deploy a simple buildroot image with FOGProject?]

Why don’t you want to go the normal route of capturing and deploying an image?

Because this is the last step in a continuous integration system. Having to copy the image to a machine and back again would be an extra unneccesary step since we already have the image ready to go. It could also create problems, for example, in that the size of the image could change which we don’t want to do.

Currently there is no official way of converting existing OS images mostly because capturing an image is the most straightforward way to go about it.

Maybe, I just don’t see that as more straightforward than dd’ing a known image to a device across a network. I could PXE boot a minimal image to do this but I thought since FOGProject already has an imaging tool, fixing my image to match the tool would be easier and have more smarts (like broadcast).

Hi,

I’ve been trying to work out how to use FOGProject as the imaging endpoint of our continuous integration system.

We currently build buildroot images for all our devices which are then manually loaded onto the disk (and auto-expand the filesystem, setting config from the network on first boot). I have managed to get the clients to load the PXE boot menu and run Ubuntu ISOs but when I try to load our custom buildroot image it seems to want to load the whole image into memory and run it rather than run an imager and image the disk from the network.

I have searched a lot but just keep coming across how to build FOS (which is buildroot-derived, too) rather than image the target disk.

Is there somewhere that lists the steps to convert a buildroot output (image or partitions) ready for FOGProject to deploy without going through a write/register/upload cycle?

Cheers,
Tudor.