@george1421, great stuff! How couldn’t I see this before?
lol!!!
Well, I reinstalled it, and already applied what Elliot said. Let’s see if it will take things easy. Thanks!!

The intent of the following tutorial is to show you how its possible to use FOG to pxe boot a live environment with FOG: https://forums.fogproject.org/topic/10944/using-fog-to-pxe-boot-into-your-favorite-installer-images
Again creating a disk-less boot environment is not the intent of FOG, its just a proof of concept.

@ylber you can get the code from from working_dev branch in git repository.

@agray said in Issues with copying software configs?:

here are 13 people that will be using that PC. All accounts where set up, including IBM Notes, on the PC before Cloning.

Was this system sysprep’d? I know Lotus Notes (been a Domino admin on and off for almost 10 years). LN creates is config file in the user’s home %appdata% directory. Its possible that sysprep is cleaning out these directories where LN lives. I know post install you can just copy the notes data directory over from one computer to the next without any problems or needing to reactivate the account. With Notes 7, we were running the notes data directory from the user’s home drive without issue. That allowed users to roam between computers without needing to mess with local directories. Since then I understand LN now supports roaming profiles, but I digress.

@Wayne-Workman Interesting, I found our problem was the initial setup of the Ubuntu server that hosts FOG was set to allow auto-updates. This will come in handy for future installs, thanks!

@anhphuong all assigned to that host. i agree that being able to select what snapins you want at the time of deploy would be nice, but the logic is tricky and there hasn’t been a lot of demand for it.

@meobeou Hi, i am using it and it‘s easy to use, google for create pushbullet api key, create it in your pb account and paste it into the fogs pb config. After that it should work.

Bild Text

Bild Text

Easy 🙂

Regards X23

@sampsont I guess you could come up with a solution where you add our FOS linux system to your clients and boot from it. I don’t know anyone who’s done this but I reckon it’ can be done. Surely Ubuntu 12.04 had grub installed. So you’d copy bzImage and init.xz to the clients and modify grub.conf to boot exactly that as first boot option. The problem I see is that you only have one go (old Ubuntu and FOS being wiped from disk as soon as imaging starts) and if things do wrong…?!

And you’d need to prepare all the clients so why not just using clusterssh and do a normal dist-upgrade 12.04 - 14.04 - 16.04 - 18.04? Probably not trouble free either but not much involved to try it out with the first batch.

What kind of hardware do you have? As George said, some manufacturers like Dell and HP might provide command line tools that let you change boot options.

Would be real interesting to see if FOG would image properly while wiping itself from the disk. Should work as it is loaded to RAM but I never tried that before.

@Wayne-Workman Thanks for the response. I ended up renumbering the image name to include 01. through 14 and that did sort them. I appreciate the quick thinking and that fixes my issue!

@vishalthejigsaw What’s the end-goal here? Snapins? And you should think about SnapinPacks. Put all your stuff into one SnapinPack and then run that on a schedule. Documentation here: https://wiki.fogproject.org/wiki/index.php?title=SnapinPacks

With fog you would set the image type to “multiple disk resizeable” and pick “all partitions”.

Thanks kafluke! This was just what I was looking for!

For anyone else looking:
To change selinux to premissive permanently in CentOS 7 edit /etc/selinux/config. Change “SELINUX=permissive” the write out.

@ConJon said in Boot to Image:

We only image one kind of machine. I am wondering if it’s possible to skip everything on the host side and when we boot to the NIC, it skips every menu and starts the only image we use. Anyone know of any way to do that.

Hi, what you want sounds dangerous, what if a client boots via network by accident? You will wipe the hdd/ssd and deploy a new image.

A good idea would be something like waiting for admin to approve. Like IBM Tovoli does.

Regards X23

Thanks for the info. I’ve been leaning towards Cent so I’ll start there with a test migration.

I’m not understanding your question. Or at least not understanding what you mean or are intended to obtain.

FOG, itself, knows nothing about your Drives used/free space. We use a block based imaging tool called partclone. Partclone can detect the used vs. free/empty space.

All Fog does is use utilities that are readily available to do all the work. FOG just automates the usage and mathematics of these things to help with resizing and expanding.

I’m sure this isn’t the answer your looking for, but to understand what I mean you’d have to look at the source code. https://github.com/fogproject/fos

Specifically the bin and usr/share/fog locations.
https://github.com/FOGProject/fos/tree/master/Buildroot/board/FOG/FOS/rootfs_overlay

@china-boy said in Fogproject can only clone C disk:

Sorry, what I want to say is that the captured image should be installed on other user machines in bulk, but other disk characters other than C disk cannot be erased.

Hope you are still around. We’ve kind of lost track of this as you haven’t posted the picture of disk management of the target PC before deploy.

Without knowing the details I guess that you are trying to simply update 😄 after you have initially cloned the PC’s full disk. So steps I would recommend: In your FOG image definition set Partition to “Partition 2 only - (4)” and deploy the image to the target machine. This way only partition 2 (hope this is 😄 in your case) will be deployed and the other partitions won’t be touched.

@Wayne-Workman Amazing !! thanks a lot

@Wayne-Workman Okay i’ll give that shot on monday

@pberberian @Sbergeron I am sorry to bring up this kind of oldish topic again. Lately I have been working on a couple of things and stumbled upon something that makes me think that ARM support never made it into FOG completely (exit type calling ARM binaries that don’t exist in FOG). So I am wondering if you are still interested or maybe got it worked out yourself? I am more than happy to add what is needed to fully support ARM clients but on the other hand I don’t fancy maintaining the ARM support if it’s not used at all.

@The-Dealman As George said, take a look at the logs to see the current status of replication going. It replicates image by image but kind of parallel to all storage nodes.The actual replication of one image to one storage node is started as a background job and so it is mostly going all in parallel.